Stability, water exchange, and anion binding studies on lanthanide(III) complexes with a macrocyclic ligand based on 1.7-diaza-12-crown-4: Extremely fast water exchange on the Gd3+ complex

Article abstract of DOI:10.1021/ic9011197

The picolinate-derivative ligand based on the 1,7-diaza-12-crown-4 platform (bp12c4 ^2-) forms stable Ln³⁺ complexes with stability constants increasing from the early to the middle lanthanides, then being relatively constant for the

Full text of DOI:10.1021/ic9011197

8878 Inorg. Chem. 2009, 48, 8878–8889
DOI: 10.1021/ic9011197
Stability, Water Exchange, and Anion Binding Studies on Lanthanide(III)
Complexes with a Macrocyclic Ligand Based on 1,7-Diaza-12-crown-4: Extremely
Fast Water Exchange on the Gd3+ Complex
†
‡
‡
‡
,‡
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
Zoltan Palinkas, Adrian Roca-Sabio, Marta Mato-Iglesias, David Esteban-Gomez, Carlos Platas-Iglesias,*
‡
‡
,†
ꢀ
ꢀ
´
Andres de Blas, Teresa Rodrıguez-Blas, and Eva Toth*
†
ꢀ
ꢀ
Centre de Biophysique Moleculaire, CNRS, rue Charles Sadron, 45071 Orleans, Cedex 2, France, and
‡
´
~
Departamento de Quımica Fundamental, Universidade da Coruna, Campus da Zapateira,
~
Alejandro de la Sota 1, 15008 A Coruna, Spain
Received June 10, 2009
The picolinate-derivative ligand based on the 1,7-diaza-12-crown-4 platform (bp12c42-) forms stable Ln3+ complexes
with stability constants increasing from the early to the middle lanthanides, then being relatively constant for the rest of
the series (logKLnL = 16.81(0.06), 18.82(0.01), and 18.08(0.05) for Ln = La, Gd, and Yb, respectively). The complex
formation is fast, allowing for direct potentiometric titrations to assess the stability constants. In the presence of Zn2+, the
dissociation of [Gd(bp12c4)]+ proceeds both via proton- and metal-assisted pathways, and in this respect, this system is
intermediate between DTPA-type and macrocyclic, DOTA-type chelates, for which the dissociation is predominated
by metal- or proton-assisted pathways, respectively. The Cu2+ exchange shows an unexpected pH dependency, with
the observed rate constants decreasing with increasing proton concentration. The rate of water exchange, assessed by
17O NMR, is extremely high on the [Gd(bp12c4)(H2O)q]+ complex (kex298 = (2.20 ( 0.15) ꢀ 108 s-1), and is in the
298
same order of magnitude as for the Gd3+ aqua ion (kex = 8.0 ꢀ 108 s-1). In aqueous solution, the
[Gd(bp12c4)(H2O)q]+ complex is present in hydration equilibrium between nine-coordinate, monohydrated, and ten-
coordinate, bishydrated species. We attribute the fast exchange to the hydration equilibrium and to the flexible nature of
the inner coordination sphere. The large negative value of the activation entropy (ΔSq=-35 ( 8 J mol-1 K-1) points to
an associative character for the water exchange and suggests that water exchange on the nine-coordinate,
monohydrated species is predominant in the overall exchange. Relaxometric and luminescence measurements on
the Gd3+ and Eu3+ analogues, respectively, indicate strong binding of endogenous anions such as citrate,
hydrogencarbonate, or phosphate to [Ln(bp12c4)]+ complexes (Kaff = 280 ( 20 M-1, 630 ( 50 M-1, and 250 (
20 M-1, respectively). In the ternary complexes, the inner sphere water molecules are fully replaced by the
corresponding anion. Anion binding is favored by the positive charge of the [Ln(bp12c4)]+ complexes and the adjacent
position of the two inner sphere water molecules. To obtain information about the structure of the ternary complexes, the
[Gd(bp12c4)(HCO3)] and [Gd(bp12c4)(H2PO4)] systems were investigated by means of density functional theory
calculations (B3LYP model). They show that anion coordination provokes an important lengthening of the distances
between the donor atoms and the lanthanide ion. The coordination of phosphate induces a more important distortion of
the metal coordination environment than the coordination of hydrogencarbonate, in accordance with a higher binding
constant for HCO3- and a more important steric demand of phosphate.
Introduction
mainly promoted by the successful biomedical application of
lanthanide chelates in both diagnostics2 and therapy.3 With
In the past two decades there has been a renaissance of
lanthanide coordination chemistry in aqueous solution,1
around 200 million doses injected so far into patients, Gd3+
-
based magnetic resonance imaging contrast agents are the
most widely used lanthanide complexes in medicine.4 By
reducing the water proton relaxation times, these paramagnetic
*To whom correspondence should be addressed. E-mail: eva.jakabtoth@
cnrs-orleans.fr.
(1) Parker, D.; Dickins, R. S.; Puschmann, H.; Crossland, C.; Howard, J.
A. K. Chem. Rev. 2002, 102, 1977–2010.
(3) (a) Liu, S. Adv. Drug Delivery Rev. 2008, 60, 1347–1370. (b) Roesch, F.
Radiochim. Acta 2007, 95, 303–311. (c) Volkert, W. A.; Hoffman, T. J. Chem.
Rev. 1999, 99, 2269–2292.
(2) (a) Lanthanide Probes in Life, Chemical and Earth Sciences: Theory
and Practice; Bunzli, J-.C. G., Choppin, G. R., Eds.; Elsevier: New York, 1989.
(b) Caravan, P.; Ellinson, J. J.; McMurry, T. J.; Lauffer, R. B. Chem. Rev. 1999,
99, 2293.
(4) The Chemistry of Contrast Agents in Medical Magnetic Resonance
ꢀ
ꢀ
Imaging; Merbach, A. E.; , E. Toth, Eds.; Wiley: New York, 2001.
r
2009 American Chemical Society
pubs.acs.org/IC
Published on Web 08/05/2009

Article
Inorganic Chemistry, Vol. 48, No. 18, 2009 8879
compounds efficiently enhance image contrast and provide an
invaluable tool to delineate disease-related tissue alterations.
Innovative ligand structures are continuously tested to improve
the relaxation features of the Gd3+ complexes and to better
understand the relations between chemical structure and effi-
cacy and stability of the chelates.5,6 The efficacy of these
contrast agents is governed by their microscopic properties,
the most important ones being the hydration number, the water
exchange rate, and the rotational dynamics of the Gd3+
complex.7 In the past years, significant efforts have been made
to optimize these parameters. The main factors influencing
water exchange have been identified as the steric crowding
around the water binding site8 and the flexibility of the inner
coordination sphere.9 Novel ligands ensuring optimal water
exchange have been synthesized based on a rational design.10
Furthermore, the stability of the chelates to be used in vivo is
also a very important issue.11 The complexes have to be
eliminated intact from the body to avoid any toxicity associated
with the release of free Gd3+. Therefore, high thermodynamic
stability and kinetic inertness are prerequisites for safe medical
use of these chelates.12
Recently, we have reported a series of acyclic chelators
bearing picolinate arms and carboxylate or phosphonate
groups designed for Gd3+ complexation in aqueous solution
(Chart 1).9,13 These ligands form thermodynamically stable
lanthanide complexes.14 Interestingly, the water exchange
rate on the phosphonate derivative Gd3+ complex of L2 has
been found to be extremely high, comparable to that on the
aqua ion itself.9 This fast water exchange has been related to
the flexible nature of the chelate, and indeed, the rigidifica-
tion of the ligand backbone by introducing a cyclohexyl
ring (L3) resulted in a decrease of the water exchange rate.9
In a recent paper we described a novel picolinate-deriva-
tive molecule based on the 1,7-diaza-12-crown-4 platform
Chart 1
(bp12c42-, Chart 1).15 This macrocyclic chelator is expected
to form kinetically more inert complexes with respect to the
previous acyclic picolinate analogues, which will be an
advantage to prevent toxicity. On the other hand, the optimal
water exchange characteristics reported for complexes with
picolinate arms could be preserved. The octadentate ligand
bp12c42- was shown to form lanthanide complexes that
contain one or two inner sphere water molecules resulting
in nine- and ten-coordinate chelates for the heavier and
lighter lanthanides, respectively.15 Complexes of lanthanides
around the middle of the series are present in the form of
hydration equilibrium between mono- and bisaqua species,
as it was proved by luminescence lifetime and UV-vis
absorbance measurements on the Eu3+ analogue (the aver-
age hydration number on the Eu3+ complex is qave = 1.4 at
298 K). While hydration equilibria are relatively common for
Eu3+ and Gd3+ complexes, they usually involve eight- and
nine-coordinate species.16 Indeed, coordination number 10 is
rather unusual in the solution chemistry of lanthanide com-
plexes with amino-carboxylates and related ligands.17 To the
best of our knowledge, no water exchange data are available
on ten-coordinate Gd3+ complexes.
(5) (a) Aime, S.; Calabi, L.; Cavallotti, C.; Gianolio, E.; Giovenzana, G.
B.; Losi, P.; Maiocchi, A.; Palmisano, G.; Sisti, M. Inorg. Chem. 2004, 43,
7588–7590. (b) Baranyai, Z.; Uggeri, F.; Giovenzana, G. B.; Benyei, A.; Brucher,
E.; Aime, S. Chem.;Eur. J. 2009, 15, 1696–1705. (c) Gianolio, E.; Giovenzana,
G. B.; Longo, D.; Longo, I.; Menegotto, I.; Aime, S. Chem.;Eur. J. 2007, 13,
5785–5797. (d) Elemento, E. M.; Parker, D.; Aime, S.; Gianolio, E.; Lattuada, L.
Org. Biomol. Chem. 2009, 7, 1120–1131.
(6) Pellegatti, L.; Zhang, J.; Drahos, B.; Villette, S.; Suzenet, F.;
Guillaumet, G.; Petoud, S.; Toth, E. Chem. Commun. 2008, 6591–6593.
(7) Peters, J. A.; Huskens, J.; Raber, D. J. Prog. NMR Spectrosc. 1996, 28,
283–350.
(8) (a) Ruloff, R.; Toth, E.; Scopelliti, R.; Tripier, R.; Handel, H.;
Merbach, A. E. Chem. Commun. 2002, 2630–2631. (b) Laus, S.; Ruloff, R.;
Toth, E.; Merbach, A. E. Chem.;Eur. J. 2003, 9, 3555–3566.
(9) (a) Mato-Iglesias, M.; Platas-Iglesias, C.; Djanashvili, K.; Peters, J.
A.; Toth, E.; Balogh, E.; Muller, R. N.; Vander Elst, L.; de Blas, A.;
Here we report the protonation constants of bp12c42-
and the thermodynamic stability constants determined for
complexes with a series of lanthanides and some biologi-
cally relevant divalent metal ions. The kinetic inertness
of the Ce3+ and Gd3+ complexes has been assessed in
strongly acidic solutions, as well as at pH 4.5-5.5 by
transmetalation studies. A variable temperature 17O
NMR study performed on [Gd(bp12c4)(H2O)q]+ yielded
parameters characterizing water exchange and rotation.
Finally, the formation of ternary complexes between
Rodrıguez-Blas, T. Chem. Commun. 2005, 4729–4731. (b) Balogh, E.; Mato-
´
Iglesias, M.; Platas-Iglesias, C.; Toth, E.; Djanashvili, K.; Peters, J. A.; de Blas, A.;
Rodríguez-Blas, T. Inorg. Chem. 2006, 45, 8719–8728.
(10) (a) Jaszberenyi, Z.; Sour, A.; Toth, E.; Benmelouka, M.; Merbach, A.
ꢀ
E. Dalton Trans. 2005, 2713–2719. (b) Kotek, J.; Lebduskova, P.; Hermann, P.;
Vander Elst, L.; Muller, R. N.; Geraldes, C. F. G. C.; Maschmeyer, T.; Luke, I.;
Peters, J. A. Chem.;Eur. J. 2003, 9, 5899–5915. (c) Polasek, M.; Sedinova, M.;
Kotek, J.; Vander Elst, L.; Muller, R. N.; Hermann, P.; Lukes, I. Inorg. Chem.
2009, 48, 455–465.
(15) Mato-Iglesias, M.; Roca-Sabio, A.; Palinkas, Z.; Esteban-Gomez,
D.; Platas-Iglesias, C.; Toth, E.; de Blas, A.; Rodriguez-Blas, T. Inorg. Chem.
2008, 47, 7840–7851.
(11) Cheng, S.; Abramova, L.; Saab, G.; Turabelidze, G.; Patel, P.;
ꢀ ꢀ
(16) (a) Graeppi, N.; Powell, D. H.; Laurenczy, G.; Zekany, L.; Merbach,
Arduino, M.; Hess, T.; Kallen, A.; Jhung, M. JAMA 2007, 297, 1542–1544.
::
A. E. Inorg. Chim. Acta 1995, 235, 311–326. (b) Platas-Iglesias, C.; Corsi, D.
(12) Brucher, E.; Sherry, A. D. Stability and Toxicity of Contrast Agents.
::
ꢀ
ꢀ
M.; Vander Elst, L.; Muller, R. N.; Imbert, D.; Bunzli, J.-C. G.; Toth, E.;
In The Chemistry of Contrast Agents in Medical Magnetic Resonante
Maschmeyer, T.; Peters, J. A. J. Chem. Soc., Dalton Trans. 2003, 727–737.
´ ´
(17) (a) Valencia, L.; Martınez, J.; Macıas, A.; Bastida, R.; Carvalho, R.
ꢀ
Imaging; Toth, E., Merbach, A. E., Eds.; Wiley: Chichester, 2001; pp 243-279.
(13) Platas-Iglesias, C.; Mato-Iglesias, M.; Djanashvili, K.; Muller, R. N.;
~
A.; Geraldes, C. F. G. C. Inorg. Chem. 2002, 41, 5300–5312. (b) Nunez, C.;
Bastida, R.; Macías, A.; Mato-Iglesias, M.; Platas-Iglesias, C.; Valencia, L.
Dalton Trans. 2008, 3841–3850. (c) Chatterton, N.; Bretonniere, Y.; Pecaut, J.;
Mazzanti, M. Angew. Chem., Int. Ed. 2005, 44, 7595–7598.
Vander Elst, L.; Peters, J. A.; de Blas, A.; Rodrı
2004, 10, 3579–3590.
(14) Mato-Iglesias, M.; Balogh, E.; Platas-Iglesias, C.; Toth, E.; de Blas,
A.; Rodrıguez-Blas, T. Dalton Trans. 2006, 5404–5415.
´
guez-Blas, T. Chem.;Eur. J.
´

ꢀ
Palinkas et al.
ꢀ
8880 Inorganic Chemistry, Vol. 48, No. 18, 2009
[Gd(bp12c4)(H2O)q]+ and some biologically relevant
small anions has been investigated.
concentration of the gadolinium(III)-complex was 0.0001 M,
the Cu2+ concentration varied between 0.001 and 0.02 M, and
MES was used as buffer. The rate of the exchange reaction was
studied at different pHs (4.6-5.3). The least-squares fit of the
kinetic data was performed by using Micromath Scientist version
2.0 (Salt Lake City, UT, U.S.A.). The reported errors correspond
to one standard deviation obtained by the statistical analysis.
17O NMR Measurements. The transverse and longitudinal
17O relaxation rates (1/T1,2) and the chemical shifts were mea-
sured in a [Gd(bp12c4)(H2O)q]+ aqueous solution in the tem-
perature range 280-345 K, on a Bruker Avance 500 (11.75 T,
67.8 MHz) spectrometer. The temperature was calculated
according to previous calibration with ethylene glycol and
methanol.20 An acidified water solution was used as reference
(HClO4, pH 3.3). Longitudinal 17O relaxation times (T1) were
measured by the inversion-recovery pulse sequence,21 and the
transverse relaxation times (T2) were obtained by the Carr-
Purcell-Meiboom-Gill spin-echo technique.22 The technique
of the 17O NMR measurements on Gd3+ complexes have been
described elsewhere.23 The samples were sealed in glass spheres
fitted into 10 mm NMR tubes, to eliminate susceptibility
corrections to the chemical shifts.24 To improve sensitivity in
17O NMR, 17O-enriched water (10% H217O, CortecNet) was
added to the solutions to yield around 1% 17O enrichment. The
molal concentration of the [Gd(bp12c4)(H2O)q]+ complex in
the sample used for 17O NMR was 0.0263 mol/kg (pH 6.0). The
17O NMR data have been evaluated according to the Solomon-
Bloembergen-Morgan theory of paramagnetic relaxation,25
though using a simplified exponential model for the temperature
dependency of the electron spin relaxation rate (see Supporting
Information). The least-squares fit of the 17O NMR data was
performed by using Micromath Scientist version 2.0 (Salt Lake
City, UT, U.S.A.). The reported errors correspond to one
standard deviation obtained by the statistical analysis.
Experimental Section
The ligand has been synthesized as described in ref 15. The
stock solutions of LnCl3 (Ln: La, Ce, Eu, Gd, Dy, and Yb)
were made by dissolving Ln2O3 in a slight excess of concen-
trated HCl in double distilled water. The excess of aqueous
HCl solution was removed by evaporation. The other metal
chloride solutions were made from chloride salts in double
distilled water. The concentration of the solutions was deter-
mined by complexometric titration with a standardized
Na2H2EDTA solution (H4EDTA = ethylenediaminetetraa-
cetic acid) using xylenol orange as indicator. Ligand stock
solutions were prepared in double distilled water. The exact
ligand concentration was determined from the difference in
the potentiometric titration curves with KOH in the absence
and in the presence of 50-fold excess of Ca2+. The difference
intheinflection pointsofthetwo titrationcurves corresponds
to 2 equiv of the ligand. The [Gd(bp12c4)(H2O)q]+ complex
has been prepared by mixing equimolar quantities of a GdCl3
and ligand solution and adjusting the pH to 6.0.
pH-Potentiometry. Ligand protonaton constants and stability
constants of Ln3+ complexes (Ln: La, Ce, Eu, Gd, Dy and Yb)
were determined by pH-potentiometric titration at 25 °C in 0.1 M
KCl. The samples (3 mL) were stirred while a constant N2 flow was
bubbled through the solutions. The titrations were carried out
adding standardized KOH solution with a Methrom 702 SM
Titrino automatic buret. A Metrohm 692 pH/ion-meter was used
to measure pH. The H+ concentration was obtained from the
measured pH values using the correction method proposed by
Irving et al.18 The protonation and stability constants were
calculated from parallel titrations with the program PSEQUAD.19
The errors given correspond to one standard deviation.
Ternary Complex Formation. The formation of ternary com-
plexes between [Gd(bp12c4)(H2O)q]+ and carbonate, phos-
phate, and citrate ions has been studied by measuring the
longitudinal relaxation rates (1/T1) of water protons on a Bruker
NMR spectrometer at 500 MHz and 25 °C. The concentration
of the Gd3+ complex was 1 mM while the concentration of each
anion varied between 0 and 36 mM by adding 15 μL portions of
the anion stock solution to 0.5 mL of complex solution in the
NMR tube. The exact concentration of the complex and the
anions were calculated at each point; 0.01 M HEPES was used as
buffer to maintain a constant pH (pH = 7.4), while the ionic
strength was 0.15 M NaCl.
Kinetic Studies. The proton assisted dissociation of the Ce3+
complex of bp12c42- has been studied in the presence of a large
excess of HCl, where the complex is unstable. The reaction was
followed at 25 °C by monitoring the decrease of the absorbance
of the complex at 280 nm in the range of proton concentrations
0.001-0.3 M. The concentration of the complex was 0.0001 M.
The exchange reactions between [Gd(bp12c4)(H2O)q]+ and
Zn2+ have been studied by measuring the longitudinal relaxation
rates (1/T1) of water protons on a Bruker Avance 500 (11.75 T)
NMR spectrometer in the pH range 4.6-5.3. The Zn2+ concen-
tration varied between 0.005 and 0.03 M, while the concentration
of the Gd3+ complex was 0.0002 M. 0.02 M N-methyl-piperazine
was used as buffer, and the ionic strength was 0.1 M KCl. The
relaxivities, r1, were calculated from the measured 1/T1obs water
proton relaxation rates according to eq 1, where 1/T1w is the
relaxation rate of water at the given temperature, and [Gd] is the
Gd3+ concentration in mM.
Excitation and emission spectra were recorded on a Perkin-
Elmer LS-50B spectrometer. Luminescence lifetimes were calcu-
lated from the monoexponential fitting of the average decay data,
and they are averages of at least 3-5 independent determinations.
Anion binding studies were performed by using the same instru-
ment on about 10-5 M solutions of the [Eu(bp12c4)(H2O)q]+
complex at 25 °C. Typically, aliquots of a fresh standard 10-5
M
1
1
1
1
solution of the Eu3+ complex containing the anion (up to 0.01 M)
were used, and the emission spectra of the samples were recorded.
The pH of the solutions was adjusted to 7.4 ( 0.05 by using HEPES
buffer (0.01 M), and the ionic strength was 0.15 M NaCl.
¼
þ
¼
þ r1 ꢀ ½Gdꢁ
ð1Þ
T1obs
T1w T1p
T1w
The pseudo-first-order rate constants (kobs) were calculated
by fitting the relaxation rate data to eq 2,
r1t ¼ ðr10 -r1eÞ expð -kobstÞ þ r1e
ð2Þ
(20) Raiford, D. S.; Fisk, C. L.; Becker, E. D. Anal. Chem. 1979, 51, 2050–
2051.
where r1t, r10, and r1e are the relaxivity values at time t, time zero,
and at equilibrium, respectively.
The dissociation rate constants of the metal exchange reac-
tion with Cu2+ were determined by UV-vis measurements
following the increase of the absorbance at 300 nm. The
(21) Vold, R. L.; Waugh, J. S.; Klein, M. P.; Phelps, D. E. J. Chem. Phys.
1968, 48, 3831–3832.
(22) Meiboom, S.; Gill, D. Rev. Sci. Instrum. 1958, 29, 688–691.
::
(23) Micskei, K.; Helm, L.; Brucher, E.; Merbach, A. E. Inorg. Chem.
1993, 18, 3844.
(24) Hugi, A. D.; Helm, L.; Merbach, A. E. Helv. Chim. Acta 1985, 68,
508–521.
ꢀ
ꢀ
(25) Toth, E., Helm, L. Merbach, A. E. Relaxivity of Gadolinium(III)
(18) Irving, H. M.; Miles, M. G.; Pettit, L. Anal. Chim. Acta 1967, 28, 475–
Complexes: Theory and Mechanism. In The Chemistry of Contrast Agents in
488.
ꢀ
ꢀ ꢀ
ꢀ
ꢀ
(19) Zekany L.; Nagypal, I. In ComputationMethods for Determination of
Medical Magnetic Resonance Imaging; Merbach, A. E., Toth, E., Eds.; Wiley:
New York, 2001; Chapter 2, p 45-120.
Formation Constants; Leggett, D. J., Ed.; Plenum: New York, 1985; p 291.

Article
Computational Details. All calculations were performed
employing hybrid density functional theory (DFT) with the
B3LYP exchange-correlation functional,26,27 and the Gaussian
03 package (Revision C.01).28 Full geometry optimizations of
Inorganic Chemistry, Vol. 48, No. 18, 2009 8881
Table 1. Protonation Constants of bp12c42- and Related Ligands and Stability
Constants of Their Ln3+ and M2+ Complexes (25 °C; I = 0.1 M (KCl))
bp12c42-
bp18c6a ba12c4b EDTAc DTPAd
the [Gd(bp12c4)]+ HCO3
,
[Gd(bp12c4)]+ H2PO4
, and
-
-
3
3
logK1
logK2
logK3
logK4
9.16 (0.03)
7.54 (0.04)
3.76 (0.05)
2.79 (0.04)
7.41
6.85
3.32
2.36
9.53
7.46
2.11
10.17
6.11
2.68
10.34
8.59
4.25
2.71
2-
[Gd(bp12c4)]+ HPO4 systems were performed in vacuo by
3
using the effective core potential (ECP) of Dolg et al. and the
related [5s4p3d]-GTO valence basis set for Gd,29 and the
6-31G(d) basis set for C, H, N, O, and P atoms. The stationary
points found on the potential energy surfaces as a result of the
geometry optimizations have been tested to represent energy
minima rather than saddle points via frequency analysis.
logKLaL
logKLaLOH
logKCeL
logKCeLOH
logKEuL
16.81 (0.06) 14.99
10.87 (0.08)
16.94 (0.09) 15.11
9.78 (0.07)
18.62 (0.08) 13.01
9.96 (0.07)
18.82 (0.01) 13.02
15.46
15.94
17.32
17.35
18.28
19.48
18.23
19.09
20.87
20.73
21.11
22.59
Results and Discussion
logKEuLOH
logKGdL
Ligand Protonation Constants and Stability Constants
of the Complexes. The protonation constants of bp12c42-
as well as the stability constants of its metal complexes
formed with Ln3+ and biologically relevant divalent ions
were determined by potentiometric titrations; the con-
stants and standard deviations are given in Table 1, which
also lists those of bp18c62- (Chart 1).32 The ligand
protonation constants are defined as in eq 3, and the
stability and protonation constants of the metal chelates
are expressed in eqs 4 and 5, respectively. The titration
curves of the lanthanide complexes all show a deprotona-
tion step at high pH (>8) indicating the formation of a
monohydroxo complex (Figure 1). The formation of
these monohydroxo complexes has been characterized
by the protonation constant KMLOH, as defined in eq 6.
logKGdLOH 9.49 (0.03)
logKDyL 18.11 (0.08) 11.72
logKDyLOH 10.0(0.1)
logKYbL 18.08 (0.05) 8.89
logKYbLOH 10.15 (0.08)
logKCaL
logKZnL
logKCuL
12.09 (0.06)
18.12 (0.03)
19.56 (0.04)
8.50
12.28
15.95
10.8
16.5
18.8
10.7
18.29
21.5
logKCuHL
6.52 (0.09)
4.8
a Ref 32. b Ref 33. I = 0.1 M Me4NNO3. c Ref 34. d Ref 34b.
½HiLꢁ
Ki ¼
ð3Þ
ð4Þ
ð5Þ
ð6Þ
½Hi -1Lꢁ½Hþꢁ
½MLꢁ
KML
¼
½Mꢁ½Lꢁ
½MHLꢁ
KMHL
¼
½MLꢁ½Hþꢁ
½MLꢁ
KMLOH
¼
½MLðOHÞꢁ½Hþꢁ
Four protonation constants could be determined for
the bp12c42- ligand. The first two logK values correspond
(26) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
(27) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789.
(28) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, Jr., J. A.; Vreven, T.; Kudin, K. N.;
Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa,
J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A.
D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari,
K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.;
Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.;
Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03; Gaussian,
Inc.: Wallingford, CT, 2004.
Figure 1. Top: Potentiometric pH titration curves for the ligand
bp12c42- in the absence and in the presence of some of the metal ions
studied at 1:1 metal to ligand ratio. Bottom: Species distribution of the
[Gd(bp12c4)] system at 1:1 Gd:bp12c4 ratio, [Gd3+] = 1 mM [I = 0.1 M
KCl, 25 °C].
(29) Dolg, M.; Stoll, H.; Savin, A.; Preuss, H. Theor. Chim. Acta 1989, 75,
173–194.

ꢀ
Palinkas et al.
ꢀ
8882 Inorganic Chemistry, Vol. 48, No. 18, 2009
to the protonation of the macrocyclic nitrogens. logK1 is
somewhat lower for bp12c42- than for the bisacetate
derivative of the same macrocycle (ba12c4, Chart 1). It
is in accordance with previous observations where a
diminution of the amine basicity, though more impor-
tant than here, has been observed on replacement of
the acetate arm by 6-methyl-2-pyridinecarboxylate
groups.30 The third and fourth protonation steps of
bp12c42- are likely associated to the carboxylic acid
groups.30,31
The stability constants of the lanthanide complexes of
bp12c42- could be obtained from direct potentiometric
titrations, as the complex formation was fast. The com-
plex stability increases from the early lanthanides to the
middle of the series, then remains relatively constant or
slightly declines for the heavier lanthanides. In this
respect, this chelator is similar to DTPA, in contrast to
most of the poly(amino carboxylate) ligands, such as
EDTA, which form complexes of increasing stability all
across the lanthanide series because of the increase of
charge density on the metal ions. As expected, the
peculiar selectivity for the light lanthanides of the eigh-
teen-membered macrocyclic ligand bp18c62-32 is not
observed for bp12c42-. Given its much smaller size,
this twelve-membered macrocycle provides an optimal
fit for smaller lanthanides. For the first half of the
lanthanide series (Ln = La-Gd), the stability con-
stants fall in between those reported for EDTA and
DTPA complexes. For the heaviest lanthanides, the
stability constants are somewhat lower for bp12c42-
complexes than for EDTA ones. We should note the
high stabilities of Cu2+, Zn2+, and Ca2+ complexes
formed with bp12c42- with respect to those for the
bis(acetate) analogue ba12c4.33
diagram obtained for the representative Gd3+ com-
plex is depicted in Figure 1.
To better compare the stability of Gd(bp12c4)+ to other
Gd3+ complexes, we have calculated the pGd value at
pH 7.4, cL = 1 ꢀ 10-5 M and cGd = 1 ꢀ 10-6 M (pGd =
-log[Gd3+
]free). pGd values reflect the influence of the
ligand basicity and the protonation of the complex on
the stability; the higher the pGd, the more stable is
the complex under the given conditions. For Gd(bp12c4)+
we obtain pGd = 17.6, a slightly lower value than that
calculated for GdDTPA (pGd = 19.1), which clearly
shows the good complexing ability of our ligand toward
lanthanides.
Kinetic Studies on [Ln(bp12c4)](H2O)q]+ Complexes. It
has been recognized early on that the fate of the Gd3+
complexes in vivo cannot be predicted on the basis
of equilibrium considerations.12,36 The kinetic inert-
ness of the complexes plays a crucial role in determin-
ing the Gd3+ release in vivo. The kinetic stability can
be characterized by rate constants of exchange reac-
tions that might take place in the plasma. Among
those, the most important is probably the displacement
of Gd3+ by the endogenously abundant Cu2+ and
Zn2+ ions. Since the dissociation of the complexes
and the metal exchange reactions are relatively
slow processes around physiological pH, the dissocia-
tion of many complexes has been studied in acidic
solutions.12,36 More recently, detailed kinetic studies
have been conducted at higher pHs (3.5-6) to investi-
gate the role of transmetalation with Zn2+ and
Cu2+ 37 In general terms, linear DTPA-type Gd3+
.
chelates are kinetically much less inert than the macro-
cyclic, DOTA-type complexes. At physiological pH,
the dissociation of linear chelates proceeds mainly via
transmetalation, which largely takes over the acid-
assisted pathway. The reaction rates of the metal
exchange between GdDTPA and Cu2+ or Zn2+ are
independent of pH above pH 4.5, indicating that the
exchange predominantly occurs via direct attack of
these metals on the Gd3+ complex.37 In contrast, the
dissociation of macrocyclic chelates is much slower
and independent of the exchanging metal ion concen-
tration, showing that it occurs through proton-assisted
pathways.38-40
We have assessed the kinetic inertness of the
[Ln(bp12c4)]+ complexes by following the acid-assisted
dissociation in strongly acidic solutions for the Ce3+
complex, and the transmetalation with Zn2+ and Cu2+
at pH 4.5-5.5 for the Gd3+ analogue. In the presence of a
large excess of acid (0.001-0.3 M HCl), the complex is
thermodynamically unstable. The reaction was followed
for the Ce3+ complex by monitoring the decrease of the
absorbance of the complex at 280 nm. The observed first
order dissociation rate constants follow a quadratic
dependence on the proton concentration (eq 7) showing
At high pH, all lanthanide complexes undergo
deprotonation, likely occurring on the coordinated
water molecule, to form monohydroxo complexes.35
The KMLOH constants characterizing this deprotona-
tion step decrease from the beginning to the middle
of the series to reach the lowest value for [Gd-
(bp12c4)(OH)], then again slightly increase for the
heavier lanthanides. The trend observed in the first
part of the series can be easily rationalized by the
increasing charge density of the metal ions. The in-
crease observed toward the end of the series could be
related to a diminution in the hydration from two to
one for the small lanthanides. The species distribution
(30) Chatterton, N.; Gateau, C.; Mazzanti, M.; Pecaut, J.; Borel, A.;
Helm, L.; Merbach, A. Dalton Trans. 2005, 1129–1135.
(31) (a) Ferreiros-Martinez, R.; Esteban-Gomez, D.; Platas-Iglesias, C.;
de Blas, A.; Rodriguez-Blas, T. Dalton Trans. 2008, 5754–5765. (b) Pellissier,
A.; Bretonniere, Y.; Chatterton, N.; Pecaut, J.; Delangle, P.; Mazzanti, M. Inorg.
Chem. 2007, 46, 3714–3725.
(32) Roca Sabio, A.; Mato-Iglesias, M.; Esteban-Gomez, D.; Toth, E.; de
Blas, A.; Platas-Iglesias, C.; Rodriguez-Blas, T. J. Am. Chem. Soc. 2009, 131,
3331–3341.
::
(33) Amorim, M. T. S.; Delgado, R.; Frausto da Silva, J. J. R. Polyhedron
1992, 11, 1891–1899.
(36) Brucher E. In Contrast Agents I. Magnetic Resonance Imaging;
Krause, W., Ed.; Springer Verlag: Berlin, 2002; Topics in Current Chemistry,
(34) (a) Lacoste, R. G.; Christoffers, G. V.; Martell, A. E. J. Am. Chem.
Soc. 1965, 87, 2385–2388. (b) Martell, A. E.; Motekaitis, R. J.; Smith, R. M.
NIST Critically selected stability constants of metal complexes database,
Version 8.0 for windows; National Institute of Standards and Technology:
Gaithersburg, MD, 2004; Standard Reference Data Program.
(35) Nonat, A.; Fries, P. H.; Pecaut, J.; Mazzanti, M. Chem.;Eur. J.
2007, 13, 8489–8506.
Vol. 221, p 104.
(37) Sarka, L.; Burai, L.; Brucher, E. Chem.;Eur. J. 2000, 6, 719–724.
::
ꢀ
(38) Toth, E.; Brucher, E.; Lazar, I.; Toth, I. Inorg. Chem. 1994, 33, 4070–
4076.
(39) Balogh, E.; Tripier, R.; Ruloff, R.; Toth, E. Dalton Trans. 2005,
1058–1065.
(40) Brucher, E.; Sherry, A. D. Inorg. Chem. 1990, 29, 1555–1559.

Article
Inorganic Chemistry, Vol. 48, No. 18, 2009 8883
Scheme 1. Possible Reaction Pathways for the Dissociation of
[Gd(bp12c4)]+ in the Presence of a Divalent Metal Ion M2+
respect, this system is intermediate between DTPA-type
and macrocyclic chelates, for which the dissociation is
predominated by metal- or proton-assisted pathways,
respectively. Taking into account previous studies37 and
the pH- and Zn-dependency of the kobs values (Figure 3),
the following possible reaction pathways are assumed
(Scheme 1):
Figure 2. Pseudo-first-order rate constants, kobs, as a function of the
proton concentration in the dissociation of [Ce(bp12c4)]+ (0.0001 M; I =
0.1 M KCl).
According to this reaction scheme, the rate of the
exchange reaction can be given as in eq 8:
d½GdLꢁt
-
¼ kGdL½GdLꢁ þ kGdHL½GdHLꢁ
þ kGdLM½GdLMꢁ
dt
ð8Þ
In eq 8 the first term refers to the spontaneous dissocia-
tion of the complex, the second to the proton-assisted
dissociation, while the last contribution describes the
metal-assisted dissociation. The equilibrium constants
for the protonated and the dinuclear species are given
by eqs 5 and 9, respectively:
Figure 3. Plots of kobs versus hydrogen ion concentration for the reac-
tion between [Gd(bp12c4)]+ and Zn2+. cZn = 5 mM (9), 10 mM (2),
20 mM (() and 30 mM (b).
½GdLMꢁ
KGdLM
¼
ð9Þ
½GdLꢁ½Mꢁ
the role of proton-catalyzed pathways in the dissociation
(Figure 2).
The following expression can be derived for the pseudo-
first-order rate constant of the dissociation:
0
2
kobs ¼ k0 þ k0 ꢀ ½Hþꢁ þ k0 ꢀ ½Hþꢁ
ð7Þ
k0 þ k1½Hþꢁ þ kM½Mꢁ
kobs
¼
ð10Þ
Similar behavior was reported for the dissociation of
various acyclic Ln3+ complexes, including DTPA and
pyridine-aminocarboxylate ligands.,6,41,42 It indicates
that the dissociation might take place by proton-indepen-
dent (characterized by k0) and proton-assisted path-
ways, presumably with the formation and dissociation
of mono- and diprotonated complexes (characterized by
k0 and k00, respectively). The constants fitted are k0 =0.143
( 0.009 M-1 s-1 and k00 = 0.96 ( 0.03 M-2 s-1, while k0
had to be fixed to 0. When fitting k0, we obtained a small
negative value with a large error. This is not surprising
since in such strongly acidic solutions the dissociation
pathways involving protonated complexes are much more
important than the spontaneous dissociation.
1 þ KHGdL½Hþꢁ þ KGdLM½Mꢁ
with k0 =kGdL, k1=kGdHL ꢀ KHGdL, and kM=kGdLM
ꢀ
KGdLM. In the potentiometric study, we could not identify
any protonated complex at pH >1.8, which means that
if there is any, its protonation constant logKHGdL has
to be very low. The pseudo-first-order rate constants of
the exchange reaction between [Gd(bp12c4)]+ and Zn2+
shown in Figure 3 were fitted to eq 10. During the fitting
procedure k0 was fixed at zero, otherwise small negative
values with a large error would be obtained. The rate
constants k1 and kM, as well as the protonation constant
KHGdL and the stability constant of the dinuclear com-
plex, KGdLZn, were calculated; they are listed and com-
pared to those for GdDTPA2- in Table 2.
In the exchange reaction between Gd(bp12c4)+ and
Zn2+, the observed pseudo-first order rate constants
depend on both the proton and the Zn2+ concentration
(pH 4.6-5.3; Figure 3). They increase by increasing
proton and metal ion concentration, indicating that both
proton- and metal-assisted pathways play a role. In this
The observed rate constants for the metal-exchange
reaction between [Gd(bp12c4)]+ and Cu2+ also depend
on both the proton and the Cu2+ concentration; how-
ever, their proton dependency shows an unexpected trend
(Figure 4). In contrast to previously reported complexes
and to the exchange between [Gd(bp12c4)]+ and Zn2+
,
::
(41) Brucher, E.; Laurenczy, G. J. Inorg. Nucl. Chem. 1981, 43, 2089–
here the rate constants decrease with increasing proton
concentration indicating an unusual reaction pathway.
The pH dependency is more important when the Cu2+ is
2096.
::
ꢀ
(42) Jakab, S.; Kovacs, Z.; Burai, L.; Brucher, E. Magy. Kem. Foly. 1993,
99, 391–396.

ꢀ ꢀ
Palinkas et al.
8884 Inorganic Chemistry, Vol. 48, No. 18, 2009
Table 2. Rate Constants Obtained for the Exchange Reaction of [Gd(bp12c4)]+ with Zn2+ and Cu2+ at 25 °C
[Gd(bp12c4)]+
[Gd(DTPA)]2-a
Zn2+ exchange
Cu2+ exchange
Zn2+ exchange
Cu2+ exchange
b
k1/M-1 s-1
kM/M-1 s-1
KGdLM/M-1
logKHGdL
5.0 ( 1.0
0.018 ( 0.006
16 ( 10
0.58c
0.056
7
0.047 ( 0.010
10 ( 7
0.93
13
1.6 ( 0.6
1.6 d
2.0
kOH/ M-2 s-1
(2 ( 0.9) ꢀ 108
a Ref 37. b Set to zero during the fitting procedure (see text). c From Eu3+ exchange. d Fixed in the fit.
include a term kOH ꢀ [Cu2+] ꢀ [OH-] in the numerator
of the expression of kobs to give eq 11:
k0 þ k1½Hþꢁ þ kM½Mꢁ þ kOH½Mꢁ½OH-ꢁ
kobs
¼
ð11Þ
1 þ KHGdL½Hþꢁ þ KGdLM½Mꢁ
The fit of the experimental data to this equation is
shown in Figure 4, and the fitted parameters are reported
in Table 2. We do not have proofs that a hydroxide-
assisted dissociation is responsible for the unprecedented
pH dependence of the Cu2+-exchange kinetics observed
for [Gd(bp12c4)]+, but the data for the Cu2+ exchange
represent a tentative explanation of the experimental
evidence. It is to be mentioned, however, that Margerum
et al. reported a greater kinetic activity for Cu(OH)+ in
Figure 4. Plots of kobs versus hydrogen ion concentration for the reac-
tion between [Gd(bp12c4)]+ and Cu2+. CCu = 1 mM (9), 2 mM (b),
5 mM (2), 10 mM ((), and 20 mM (4).
substitution reactions than for Cu2+ 43
.
In comparison to GdDTPA2-, the proton-assisted dis-
sociation is more rapid, while the zinc- or copper-assisted
transmetalation is less efficient for [Gd(bp12c4)]+. If we
consider that at physiological pH the acid-catalyzed path-
way becomes negligible, the [Gd(bp12c4)]+ complex is
more resistant to dissociation than its DTPA analogue. In
overall, however, the kinetic inertness of [Gd(bp12c4)]+ is
in high concentration, while at lower copper concentra-
tion the kobs values remain almost constant with pH, as
it has been observed for the metal exchange of various
DTPA-type Ln3+ complexes. This phenomenon could
be explained by the formation of a protonated H[Gd-
(bp12c4)]2+ complex that dissociates slower than the
dinuclear Cu-[Gd(bp12c4)]3+ complex transitionally
formed by the direct attack of the exchanging metal
ion. However, a protonated complex could not be
identified in the potentiometric titration, which can be
rationalized by the positive charge of [Gd(bp12c4)]+,
unfavorable for protonation. An attempt to fit the
kobs data to eq 10 resulted in the following values:
k1 = 10 M-1 s-1, kM = 1.2 M-1 s-1, KGdLCu = 11,
and logKHGdL = 5.6. While all the other parameters
have reasonable values, the value of logKHGdL = 5.6 is
inconceivable, since such high protonation constant
should be well detectable on the pH-potentiometric
titration curve, and therefore this model cannot account
for the unprecedented pH dependency of the rate
constants observed for the Cu2+ transmetalation. In
another hypothesis, the transitionally formed dinuclear
Cu-[Gd(bp12c4)]3+ species would dissociate via a hy-
droxide-assisted pathway which is faster than the pro-
ton-assisted dissociation of the [Gd(bp12c4)]+ complex
itself. This hypothesis is based on the three positive
charges of the dinuclear species which makes it more
prone to an eventual hydroxide-catalyzed dissociation.
The lanthanide complexes of bp12c42- all have shown a
tendency to form monohydroxo complexes in the po-
tentiometric study. On the other hand, Cu2+ is known
not considerably higher than that of GdDTPA2-
.
17O NMR Measurements and Water Exchange on
[Gd(bp12c4)(H2O)q]+. To assess parameters characteriz-
ing the water exchange and rotational dynamics on [Gd-
(bp12c4)(H2O)q]+, we have measured variable temperature
transverse and longitudinal 17O NMR relaxation rates and
chemical shifts at 11.7 T. Previous luminescence and high
resolution UV-vis absorbance data obtained on the corre-
sponding Eu3+ analogue proved the existence of a hydra-
tion equilibrium, with an average hydration number of
q
av = 1.4 at 298 K.15 The temperature dependence of the
hydration equilibrium has been also described by the UV-
vis absorbance study which yielded the reaction enthalpy
(ΔH° = 9.5 kJ mol-1) and the reaction entropy (ΔS° =
33 J mol-1 K-1). We assume that the same hydration
equilibrium, characterized by these thermodynamic para-
meters, exists for [Gd(bp12c4)(H2O)q]+ as well. The re-
duced 17O relaxation rates and chemical shifts have been
therefore obtained from the experimental relaxation rates
and chemical shifts by taking into account the temperature
dependency of the hydration equilibrium, which allows
calculating the average q value at each temperature. The
data are presented in Figure 5.
The 17O reduced transverse relaxation rates (1/T2r)
decrease with increasing temperature in the whole tem-
perature range investigated, indicating that they are in the
to form hydroxo-complexes at lower pH than Zn2+
,
which could explain the different behavior of the two
metals. To take into account such a pathway, we can
(43) Margerum, D. W.; Zabin, B. A.; Janes, D. L. Inorg. Chem. 1966, 5,
250.

Article
Inorganic Chemistry, Vol. 48, No. 18, 2009 8885
been adjusted: the water exchange rate, kex298, or the
activation entropy, ΔS‡, the activation enthalpy for water
exchange, ΔH‡, the scalar coupling constant, A/p,
the rotational correlation time (τR298) and its activation
energy, ER, and the electron spin relaxation rate at 298 K,
1/T1e298 and its activation energy, ET. The values obtained
298
for these parameters are presented in Table 3. For 1/T1e
,
we calculated (1.9 ( 0.4) ꢀ 107 s-1, while ET was
fixed to 1 kJ/mol, otherwise small negative values were
obtained.
The water exchange rate is very high on [Gd(bp12c4)-
(H2O)q]+, being within the same order of magnitude as
for the [Gd(H2O)8]3+ aqua ion or for the previously
studied [Gd(L2)]3- complex (Chart 1). To the best of
our knowledge, this is the fastest water exchange on a
macrocyclic Gd3+ complex ever reported. In the fast
exchange regime, the 17O transverse relaxation rates
are influenced by both the water exchange and the
electron spin relaxation. We should note that the electron
spin relaxation, 1/T1e, has a very limited contribution,
representing maximum 10% in the correlation time
τc governing the relaxation rate of the bound water
oxygen, 1/T2m (1/τc =kex + 1/T1e). Therefore, the water
exchange rate can be obtained with a very good certitude.
The activation entropy is strongly negative, indicating
that the water exchange process has an associative char-
acter, that is, the incoming water enters the inner coordi-
nation sphere of the complex before the departure of the
leaving water molecule. So far, few Gd3+ complexes have
been reported to have associative water exchange,48-51
all of them are octa-coordinate. Most Gd3+ complexes
studied with respect to their water exchange, poly-
(aminocarboxylates) in majority, were nine-coordinate
and presented dissociatively activated water exchange.
This could be related to the nature of Gd3+ that prefers
the typical coordination numbers 8 or 9 in solution.
Consequently, the water exchange of octa-coordinate
complexes will proceed via an associatively activated
mechanism, involving a nine-coordinate transition state,
while that of the nine-coordinate complexes will go
through an eight-coordinate transition state in a dissocia-
tively activated process. As a general rule, the water
exchange rate is higher for associatively activated pro-
cesses. In our case, the hydration equilibrium is between a
nine-coordinate, monohydrated, and a ten-coordinate,
bishydrated species, present in comparable quantities.
Given the large negative value obtained for the activation
entropy, one could speculate that the water exchange on
the nine-coordinate species, which can be expected to
proceed via an associative pathway, is mainly responsible
for the observed 17O T2 effect, while the contribution of
the ten-coordinate, bishydrated species, which should
proceed via a dissociative mechanism, would be negligi-
ble. The positive charge of the complex could be another
Figure 5. Reduced longitudinal (2) and transverse (9) 17O NMR
relaxation rates and 17O NMR chemical shifts (() of a [Gd(bp12c4)-
(H2O)q]+ solution at 11.75 T and pH = 6. The lines represent the best fit
of the data as explained in the text.
fast exchange region. Here the reduced transverse relaxa-
tion rate is defined by the transverse relaxation rate of the
bound water oxygen, 1/T2m, which is in turn influenced by
the water exchange rate, kex, the longitudinal electronic
relaxation rate, 1/T1e, and the scalar coupling constant,
A/p. The reduced 17O chemical shifts are determined by
A/p. Transverse 17O relaxation is governed by the scalar
relaxation mechanism and thus contains no information
on the rotational motion of the system. In contrast to
1/T2r, the longitudinal 17O relaxation rates, 1/T1r, are
determined by dipole-dipole and quadrupolar relaxation
mechanisms, both related to rotation. The dipolar term
depends on the Gd3+-water oxygen distance, rGdO, while
the quadrupolar term is influenced by the quadrupolar
coupling constant, χ(1+η2/3)1/2. The experimental data
have been fitted to the Solomon-Bloembergen-Morgan
theory of paramagnetic relaxation using the equations
presented in the Supporting Information. In the fitting
procedure, rGdO has been fixed to 2.50 A, based on
available crystal structures,44,45 and recent ENDOR re-
sults.46 The quadrupolar coupling constant, χ(1+η2/3)1/2
,
has been set to the value for pure water, 7.58 MHz.
We should note that a rGdO value of 2.585 A was obtained
for [Gd(bp12c4)(H2O)]+ from DFT calculations per-
formed in vacuo.15 However, a shorter distance is expected
in aqueous solution, where a stronger water ion interaction
arises from solvent polarization effects that increase
the dipole moment of the free water molecules.47 The
empirical constant describing the outer sphere contribution
to the 17O chemical shift, Cos, was fixed to 0. The electron
spin relaxation rate, 1/T1e, has been fitted to a simple
exponential function. The following parameters have
(48) Powell, H. D.; Ni Dhubhghaill, O, M.; Pubanz, D.; Helm, L.;
Lebedev, Y.; Schlaepfer, W.; Merbach, A. E. J. Am. Chem. Soc. 1996, 118,
9333–9346.
(44) Spirlet, M.-R.; Rebizant, J.; Desreux, J. F.; Loncin, M.-F. Inorg.
Chem. 1984, 23, 359–363.
(45) Stezowski, J. J.; Hoard, J. L. Isr. J. Chem. 1984, 24, 323–334.
(46) Raitsimring, A. M.; Astashkin, A. V.; Baute, D.; Goldfarb, D.;
Caravan, P. J. Phys. Chem. A. 2004, 108, 7318–7323.
(47) Djanashvili, K.; Platas-Iglesias, C.; Peters, J. A. Dalton Trans. 2008,
602–607.
(49) Toth, E.; Helm, L.; Merbach, A. E.; Hedinger, R.; Hegetschweiler,
ꢀ
K.; Janossy, A. Inorg. Chem. 1998, 37, 4104–4113.
(50) Thompson, M. K.; Botta, M.; Nicolle, G. M.; Helm, L.; Aime, S.;
Merbach, A. E.; Raymond, K. N. J. Am. Chem. Soc. 2003, 125, 14274–14275.
(51) Burai, L.; Toth, E.; Bazin, H.; Benmelouka, M.; Jaszberenyi, Z.;
Helm, L.; Merbach, A. E. Dalton Trans. 2006, 629–634.

ꢀ
Palinkas et al.
ꢀ
8886 Inorganic Chemistry, Vol. 48, No. 18, 2009
Table 3. Parameters Obtained from the Fitting of the 17O NMR Relaxation Rates and Chemical Shifts at 11.7 T
[Gd(bp12c4)(H2O)q]+
[Gd(DTPA)(H2O)]2-a
[Gd(H2O)8]3+a
[Gd(L2)]3-b
kex298/ 106 s-1
ΔHq/ kJ mol-1
ΔSq/J mol-1 K-1
A/p/106 rad s-1
τR298/ps
220 ( 15
14.8 ( 3.0
-35 ( 8
-3.4 ( 0.3
105 ( 16
15.0 ( 4.2
3.3
800
15.3
-23
-5.3
41
700
22
-3
-3.5
155
27
51.6
+56
-3.8
103
ER/kJ mol-1
17.3
15
a Ref 48. b Ref 9b.
factor that contributes to the associatively activated pro-
cess. The approach of the second water molecule with
its oxygen bearing a partial negative charge should be
favored by a positive charge on the complex. Conse-
quently, the water exchange on the bishydrated complex
should be considerably slower than that on the monohy-
drated one, in accordance with the general observation
that associative exchanges are faster than dissociative
ones. On the basis of this reasoning, we tried to fit the
1/T2r values by assuming that the only species that con-
tributes to the water exchange is the monohydrated
complex. By taking into account the temperature-depen-
dent variation of its ratio (based on the UV-vis study on
the Eu3+ analogue), we obtained kex298 = (8.0 ( 1.0) ꢀ
108 s-1 for [Gd(bp12c4)(H2O)]+ (ΔH‡ = 18.7 kJ/mol).
However, we do not possess any tool to decide whether
the nine-coordinate species is the only contributor to the
reduced 17O transverse relaxation rates. Therefore, we
prefer to report the effective kex value as shown in Table 3.
In any case, if we consider practical applications of this
complex as an MRI contrast agent, it is the effective kex
value that will determine the efficiency of the chelate in
enhancing water proton relaxation.
The very fast exchange observed for [Gd(bp12c4)-
(H2O)q]+ is likely related to the hydration equilibrium
of the complex, and more importantly, to the very flexible
inner coordination sphere around the metal ion. Other
Gd3+ complexes which also present hydration equilibrium
do not necessarily have such extreme water exchange
rates. For instance, [Gd(DO3A)(H2O)q] and [Gd(DO2A)-
(H2O)q]+ both have differently hydrated species in aqueous
solution (qave = 1.8 and 2.8, respectively), but they show
only a limited increase of the water exchange rate as
compared to the monohydrated [Gd(DOTA)(H2O)]-.52
For these macrocyclic complexes, the most important
factor that limits water exchange is probably the rigidity
of the inner sphere.
Anion Binding Studies. Coordinatively unsaturated
lanthanide complexes containing two inner-sphere water
molecules, such as Ln3+ DO3A-like complexes, are
known to bind biologically relevant anions such as hy-
drogencarbonate, phosphate, or citrate.53 The ability of
these lanthanide complexes to form ternary complexes
with anions is likely related to the presence of the two
inner-sphere water molecules coordinated to the Ln3+
ion in adjacent positions. Indeed, several coordinatively
unsaturated q = 2 lanthanide complexes either do not
bind anions or the observed binding is very weak.54 Most
likely this is due to an unfavorable, non adjacent location
of the inner-sphere water molecules around the metal
ion. In a previous study on the solution structure of
[Ln(bp12c4)(H2O)q]+ complexes, we have shown that in
q = 2 complexes the two inner-sphere water molecules
occupy adjacent positions in the metal ion coordination
sphere.15 Thus, these Ln3+ complexes of bp12c42- are
expected to bind anions in aqueous solution. The forma-
tion of ternary complexes between [Gd(bp12c4)(H2O)q]+
and hydrogencarbonate, phosphate, and citrate ions
has been studied by measuring the longitudinal relaxa-
tion rates (1/T1) of water protons at 500 MHz and 25 °C
(pH = 7.4, 0.01 M HEPES, I = 0.15 M NaCl). The
concentration of the Gd3+-complex was ∼1 mM while
the concentration of each anion varied between 0 and
36 mM. Anion addition caused a decrease in relaxivity, in
agreement with the replacement of inner-sphere water
molecules by the anion (for citrate see Figure 6, curves for
hydrogencarbonate and phosphate are shown in the
Supporting Information). Plots of the relaxivity versus
the [anion]/[complex] ratio show saturation profiles for
all investigated anions. The least-squares fitting of the
titration profiles allowed us to determine the binding
constants listed in Table 4. The fitting of the experimental
data also provides the relaxivity of the ternary complexes
(r1,t, Table 4). [Gd(bp12c4)(H2O)q]+ binds the three
investigated anions rather strongly, as expected for a
positively charged complex, with binding constants ran-
ging between about 250 and 630 M-1, in the following
binding trend: hydrogencarbonate > phosphate ≈ ci-
trate. The relaxivities calculated for the three ternary
complexes are very similar and consistent with q = 0.
In bovine serum, the relaxivity of [Gd(bp12c4)(H2O)q]+ is
slightly diminished (r1=3.19 mM-1 s-1 vs 3.89 mM-1 s-1 in
Concerning the other parameters calculated in the fit, the
rotational correlation time has a reasonable value for a
small molecular weight complex. For the scalar coupling
constant, A/p, we obtained -(3.4 ( 0.3) ꢀ 106 rad s-1
,
which is in good accordance with values reported for Gd3+
complexes.48 This also confirms that the hydration number
used in the calculations is correct. We should also note that
the slope of the curve of the reduced chemical shifts versus
inverse temperature is nicely reproduced by the fit, which
supports that the parameters characterizing the tempera-
ture dependency of the hydration equilibrium are correct
(Figure 5).
(53) (a) Aime, S.; Botta, M.; Bruce, J. I.; Mainero, V.; Parker, D.;
Terreno, E. Chem. Commun. 2001, 115–116. (b) Bruce, J. I.; Dickins, R. S.;
Govenlock, L. J.; Gunnlaugsson, T.; Lopinski, S.; Lowe, M. P.; Parker, D.;
Peacock, R. D.; Perry, J. J. B.; Aime, S.; Botta, M. J. Am. Chem. Soc. 2000, 122,
9674–9684.
(54) Loureiro de Sousa, P.; Livramento, J. B.; Helm, L.; Merbach, A. E.;
Meme, W.; Doan, B.-T.; Beloeil, J.-C.; Prata, M. I. M.; Santos, A. C.;
Geraldes, C. F. G. C.; Toth, E. Contrast Media Mol. Imaging 2008, 3, 78–85.
(52) Toth, E.; Ni Dhubhghaill, O. M.; Besson, G.; Helm, L.; Merbach, A.
E. Magn. Reson. Chem. 1999, 37, 701–708.

Article
Inorganic Chemistry, Vol. 48, No. 18, 2009 8887
Figure 7. Variation of the emission spectrum for [Eu(bp12c4)(H2O)q]+
following addition of hydrogencarbonate [pH = 7.4 (HEPES), [complex] =
ca. 10-5 M, λexc = 272 nm, I = 0.15 M NaCl].
Figure 6. Relaxivity of [Gd(bp12c4)(H2O)q]+ (∼1 mM) in the presence
of increasing citrate concentration, at 500 MHz and 25 °C. The line is the
fitted curve as explained in the text.
Table 4. Binding Constants (Kaff
, ) for the Interaction of [Gd-
M-1
(bp12c4)(H2O)q]+ with Different Anions and Relaxivity (r1,t, mM-1 s-1) of the
Ternary Complex Obtained from the Fit of the Relaxivity Measurements at 25 °Ca
-
citrate
HCO3
phosphate
Kaff
r1,t
280 ( 20
1.90 ( 0.02
630 ( 50
2.15 ( 0.02
250 ( 20
1.90 ( 0.03
a pH = 7.4, 0.01 M HEPES, I = 0.15 M NaCl.
pure water, 25 °C, 500 MHz), reflecting the partial
replacement of the inner sphere water.
The binding of hydrogencarbonate and citrate to the
Eu3+ complex of bp12c42- was also investigated by
observing the changes in the emission intensity and life-
time of the metal centered emission. The emission spec-
trum of a 10-5 M solution of the Eu3+ complex in H2O
[pH 7.4 (0.01 M HEPES), 0.15 M NaCl], obtained under
excitation through the ligand bands at 272 nm displays
the typical 5D0 f 7FJ transitions (Figure 7, J=0-4). The
emission lifetime of the Eu(5D0) excited level determined
under these conditions amounts to 0.50 ms, a value that
is nearly identical to that determined previously in un-
buffered solution at pH 8 and in the absence of NaCl
(0.52 ms).15 Addition of hydrogencarbonate provokes
important changes in the emission spectrum of the com-
plex (Figure 7). In particular, anion addition causes an
important increase of the intensity of the hypersensitive
Figure 8. Calculated minimum energy conformation of the [Gd-
(bp12c4)]+ HCO3 ternary complex as optimized in vacuo at the
-
3
B3LYP/6-31G(d) level. Hydrogen atoms are omitted for simplicity.
magnetic axis of the Yb3+ complex is coincident with
the C2 symmetry axis of the molecule, which is perpendi-
cular to the best plane defined by the donor atoms of the
macrocycle and contains the lanthanide ion. In the case of
the q=1 complexes, the oxygen atom of the inner-sphere
water molecule is on the principal axis. Thus, the increase
5
7
of the relative intensity of the D0 f F2 transition is
attributed to the substitution of hard coordinated water
molecules by more polarizable C-O donors of the anion.
The spectra shown in Figure 7 also evidence substantial
changes in the splitting of the 5D0 f 7F1 manifold upon
anion addition. The splitting of the J = 1 manifold is
related to the magnitude of the second order crystal field
coefficient, which in turns depends on the polarizability
of the donor atoms placed close to the principal axis of the
complex.56 Citrate binding provokes similar spectral
5
7
and electric-dipole allowed D0 f F2 transition, while
f
the intensity of the remaining 5D0 7FJ transitions
increases to a lesser extent. It has been demonstrated that
the nature and polarizability of the group occupying a
position on or close to the principal axis of the complex
affects the relative intensity of the 5D0 f 7F2 transition.55
The more polarizable the axial donor atom is, the greater
the relative intensity of the 5D0 f 7F2 emission band. In a
previous paper, we have demonstrated the principal
(55) (a) Dickins, R. S.; Parker, D.; Bruce, J. I.; Tozer, D. J. Dalton Trans.
2003, 1264–1271. (b) Murray, B. S.; New, E. J.; Pal, R.; Parker, D. Org. Biomol.
Chem. 2008, 6, 2085–2094. (c) Bretonniere, Y.; Cann, M. J.; Parker, D.; Slater, R.
Org. Biomol. Chem. 2004, 2, 1624–1632.
(56) Dickins, R. S.; Aime, S.; Batsanov, A. S.; Beeby, A.; Botta, M.;
Bruce, J. I.; Howard, J. A. K.; Love, C. S.; Parker, D.; Peacock, R. D.;
Puschmann J. Am. Chem. Soc. 2002, 124, 12697–12705.

ꢀ ꢀ
Palinkas et al.
8888 Inorganic Chemistry, Vol. 48, No. 18, 2009
Table 5. Values of the Bond Distances of the Gd3+ Coordination Environment of the Minimum Energy Conformations Calculated for [Gd(bp12c4)(H2O)2]+
,
[Gd(bp12c4)(HCO3)], and [Gd(bp12c4)(H2PO4)] Complexes at the B3LYP/6-31G(d) Levela
[Gd(bp12c4)(H2O)2]+b
[Gd(bp12c4)(HCO3)]
[Gd(bp12c4)(H2PO4)]
Gd-NPY
Gd-OC
2.565
2.646
2.684
2.520
2.351
2.388
2.723
2.806
2.549
2.611
3.038
2.578
2.334
2.422
2.794
2.915
2.513
2.442
2.540
2.666
3.200
2.563
2.313
2.494
2.848
2.986
2.511
2.406
Gd-OCOO
Gd-NAM
Gd-OA
a Distances (A), angles (deg.). NAM = amine nitrogen atoms; NPY = pyridyl nitrogen atoms; OCOO = carboxylate oxygen atoms; OC = crown oxygen
atoms; OA = oxyanion oxygen atom. b Ref 15.
changes in the emission spectrum of the Eu3+ complex
(Figure S2, Supporting Information).
the Λ(δλδλ) one, which is the most stable one in the case of
q=1 complexes.15 The optimized geometry for the complex
with hydrogencarbonate is shown in Figure 8.
The formation of the ternary adduct results in a longer
emission lifetime of the Eu(5D0) excited level. The emis-
sion lifetime determined upon addition of 590 equiv of
citrate amounts to 0.93 ms. This value is substan-
tially longer than that obtained in the absence of anion
(0.50 ms) considering that under these condition the
concentration of the ternary adduct is expected to be
about 62%. These results are in agreement with the
formation of ternary adducts in which inner-sphere water
molecules are replaced by anion binding.
Our calculations predict a slightly asymmetrical binding
of both hydrogencarbonate and dihydrogenphosphate to
the Gd3þ ion, the Gd3þ-Oanion distances amounting to 2.44
and 2.51 A (HCO3-) and 2.41 and 2.51 A (H2PO4-). Anion
coordination also results in pronounced changes in the bond
distances of the Gd3þ coordination environment in com-
parison to [Gd(bp12c4)(H2O)2]þ (Table 5). It provokes an
important lengthening of the distance to one of the oxygen
atoms of the crown moiety (by 0.36 and 0.56 A for
carbonate and phosphate, respectively). The distance be-
tween Gd3þ and one of the amine nitrogen atoms also
increases substantially upon carbonate or phosphate bind-
ing (by 0.10 and 0.18 A, respectively). In overall, the
phosphate coordination induces a more important distor-
tion of the metal coordination environment than carbonate
coordination, in line with the higher binding constant. This
is explained with a more important steric hindrance around
the pendant arms of the ligand caused by the coordination
of a tetrahedral anion such as phosphate compared to the
planar carbonate. The concentration of HPO42- in solution
at pH 7.4 is very similar to that of H2PO4-. Thus, we also
To obtain information about the structure of the ternary
complexes, the [Gd(bp12c4)(HCO3)] and [Gd(bp12c4)-
(H2PO4)] systems were investigated by DFT calculations
(B3LYP model). The effective core potential (ECP) of Dolg
et al. and the related [5s4p3d]-GTO valence basis set was
applied in these calculations.28 This ECP includes 46 + 4f7
electrons in the core, leaving the outermost 11 electrons to
be treated explicitly, and it has been demonstrated to
provide reliable results for several lanthanide complexes
with both macrocyclic57,58 or acyclic59 ligands. We have
previously shown that the complexes of bp12c42- present
two sources of helicity: one associated with the layout of the
picolinate pendant arms (absolute configuration Δ or Λ),
and the other to the four five-membered chelate rings
formed by the binding of the crown moiety (each of them
showing absolute configuration δ or λ).15 Since enantiomers
have the same physicochemical properties in a non-chiral
environment we have only considered four diasteroisomeric
forms of the complexes in our conformational analysis:
Λ(λλλλ), Δ(λλλλ), Λ(δλδλ, and Λ(λδλδ). Previous theore-
tical calculations performed at the DFT (B3LYP) level
evidence that the change in hydration number across the
lanthanide series is accompanied by a change in the con-
formation that the complexes adopt in solution [Δ(λλλλ) for
q = 2 and Λ(δλδλ) for q = 1]. Our calculations on
[Gd(bp12c4)(HCO3)] and [Gd(bp12c4)(H2PO4)] predict
the Δ(λλλλ) conformation to be the most stable for both
ternary complexes. This is attributed to the fact that the
Δ(λλλλ) conformation possesses a more open structure than
2-
investigated the coordination of HPO4 by performing
calculations on the [Gd(bp12c4)(HPO4)]- system. Our
results confirm that the coordination of phosphate induces
a more important distortion of the metal coordination
environment than the coordination of carbonate (see the
Supporting Information).
Conclusions
The stability constants of the [Ln(bp12c4)]þ complexes
have been obtained from direct potentiometric titra-
tions, as the complex formation is fast. The stability
increases from the early lanthanides to the middle of the
series, then remains relatively constant or slightly declines
for the heavier lanthanides, similarly to Ln3þ-DTPA
chelates. In the presence of Zn2þ, the dissociation of
[Gd(bp12c4)]þ proceeds both via proton- and metal-
assisted pathways, and in this respect, this system
is intermediate between DTPA-type and macrocyclic
chelates, for which the dissociation is predominated by
metal- or proton-assisted pathways, respectively. The
Cu2þ exchange shows an unexpected pH dependency,
with the observed rate constants decreasing with increas-
ing proton concentration.
(57) Gonzalez-Lorenzo, M.; Platas-Iglesias, C.; Avecilla, F.; Faulkner, S.;
Pope, S. J. A.; de Blas, A.; Rodriguez-Blas, T. Inorg. Chem. 2005, 44, 4254–
4262.
(58) Cosentino, U.; Villa, A.; Pitea, D.; Moro, G.; Barone, V.; Maiocchi,
A. J. Am. Chem. Soc. 2002, 124, 4901–4909.
(59) Ouali, N.; Bocquet, B.; Rigault, S.; Morgantini, P.-Y.; Weber, J.;
Piguet, C. Inorg. Chem. 2002, 41, 1436–1445.

Article
The [Gd(bp12c4)(H2O)q]þ complex is present in hydration
Inorganic Chemistry, Vol. 48, No. 18, 2009 8889
de Galicia (INCITE08ENA103005ES) for financial sup-
port. Z.P. acknowledges the grant of the RFR program of
the French Ministry of Education and Research. This
research was performed in the framework of the EU
COST Action D38 “Metal-Based Systems for Molecular
Imaging Applications”. The authors are indebted to
equilibrium between nine-coordinate, monohydrated, and
ten-coordinate, bishydrated species. The rate of water ex-
change as assessed by 17O NMR is extremely high, close to
that of the Gd3þ aqua ion itself. This fast exchange can be
accounted for by the flexible nature of the inner coordination
sphere. The activation entropy suggests a strong associative
character for the water exchange which presumably involves
only the nine-coordinate, monohydrated species. Relaxo-
metric and luminescence measurements, together with DFT
calculations, indicate strong anion binding to [Ln(bp12c4)-
(H2O)q]þ complexes, consistent with the complete replace-
ment of the inner sphere water molecules.
ꢀ
Centro de Supercomputacion de Galicia for providing
the computer facilities.
Supporting Information Available: Changes in the relaxivity of
the Gd3þ complex of bp12c42- upon addition of anions (Figure
S1), changes in the emission spectrum of the Eu3þ complex of
bp12c42- upon citrate binding (Figure S2), UV-vis spectra of the
system Gdbp12c4þ-Cu2þ (Figure S3), set of equation used for the
analysis of the 17O NMR data and DFT optimized Cartesian
coordinates (A) for the [Gd(bp12c4)]þ HCO3-, [Gd(bp12c4)]þ
Acknowledgment. A.R.-S., M.M.-I., D.E.-G., C.P.-I.,
ꢀ
A.de B. and T.R.-B. thank the Ministerio de Educacion y
Ciencia and FEDER (CTQ2006-07875/PPQ) and Xunta
2-3
3
H2PO4-, and [Gd(bp12c4)]þ HPO4 systems. This material is
3
available free of charge via the Internet at http://pubs.acs.org.