Synthesis, potentiometric, kinetic, and NMR studies of 1,4,7,10- tetraazacyclododecane-1,7-bis(acetic acid)-4,10-bis(methylenephosphonic acid) (DO2A2P) and its complexes with Ca(II), Cu(II), Zn(II) and lanthanide(III) ions

Article abstract of DOI:10.1021/ic7024704

A cyclen-based ligand containing trans-acetate and trans- methylenephosphonate pendant groups, H₆DO2A2P, was synthesized and its protonation constants (12.6, 11.43, 5.95, 6.15, 2.88, and 2.77) were determined by pH-potentiometry and ¹H NMR spectroscopy. The first two protonations were shown to occur at the two macrocyclic ring N-CH ₂-PO₃^2- nitrogens while the third and fourth protonations occur at the two phosphonate groups. In parallel with protonation of the two -PO₃^2- groups, the protons from the NH ⁺-CH₂-PO₃^2- are transferred to the N-CH₂-COO^- nitrogens. The stability constants of the Ca²⁺, Cu²⁺, and Zn²⁺ (ML, MHL, MH₂L, and M₂L) complexes were determined by direct pH-potentiometry. Lanthanide(III) ions (Ln³⁺) form similar species, but the formation of complexes is slow; so, out-of-cell pH-potentiometry (La ³⁺, Eu³⁺, Gd³⁺, Y³⁺) and competitive spectrophotometry with Cu(II) ion (Lu³⁺) were used to determine the stability constants. By comparing the log K_ML values with those of the corresponding DOTA (H₄DOTA = 1,4,7,10-tetraazacyclododecane-1,4, 7,10-tetraacetic acid) and DOTP (H₈DOTP = 1,4,7,10- tetraazacyclododecane-1,4,7,10-tetramethylenephosphonic acid) complexes, the order DOTA < DO2A2P < DOTP was found for all the metal ion complexes examined here with the exception of the Ca²⁺ complexes, for which the order is reversed. The relaxivity of Gd(DO2A2P) decreases between pH 2 and 7 but remains constant in the pH range of 7 < pH < 12 (r₁ = 3.6 mM^-1 s^-1). The linewiths of the ¹⁷O NMR signals of water in the absence and presence of Gd(DO2A2P) (at pH = 3.45 and 8.5) between 274 and 350 K are practically the same, characteristic of a q = 0 complex. Detailed kinetic studies of the Ce³⁺ and Gd³⁺ complexes with DO2A2P showed that complex formation is slow and involves a high stability diprotonated intermediate Ln(H₂DO2A2P)*. Rearrangement of the diprotonated intermediate into the final complex is an OH^- assisted process but, unlike formation of Ln(DOTA) complexes, rearrangement of Ln(H₂DO2A2P)* also takes place spontaneously likely as a result of transfer of one of the protons from a ring nitrogen to a phosphonate group. The order of the OH^- assisted formation rates of complexes is DOTA > DO2A2P > DOTP while the order of the proton assisted dissociation rates of the Gd³⁺ complexes is reversed, DOTP > DO2A2P > DOTA. ¹H and ¹³C NMR spectra of Eu(DO2A2P) and Lu(DO2A2P) were assigned using two-dimensional correlation spectroscopy (2D COSY), heteronuclear multiple quantum coherence (HMQC), heteronuclear chemical shift correlation (HETCOR), and exchange spectroscopy (EXSY) NMR methods. Two sets of ¹H NMR signals were observed for Eu(DO2A2P) characteristic of the presence of two coordination isomers in solution, a twisted square antiprism (TSAP) and a square antiprism (SAP), in the ratio of 93% and 7%, respectively. Line shape analysis of the ¹H NMR spectra of Lu(DO2A2P) gave lower activation parameters compared to La(DOTP) for interconversion between coordination isomers. This indicates that the Ln(DO2A2P) complexes are less rigid probably due to the different size and spatial requirements of the carboxylate and phosphonate groups.

Full text of DOI:10.1021/ic7024704

Inorg. Chem. 2008, 47, 3851-3862
Synthesis, Potentiometric, Kinetic, and NMR Studies of
1,4,7,10-Tetraazacyclododecane-1,7-bis(acetic
acid)-4,10-bis(methylenephosphonic acid) (DO2A2P) and its Complexes
with Ca(II), Cu(II), Zn(II) and Lanthanide(III) Ions
Ferenc K. Kálmán,† Zsolt Baranyai,†,‡ Imre Tóth,† István Bányai,§ Róbert Király,† Ernö Brücher,†
Silvio Aime,‡ Xiankai Sun,| A. Dean Sherry,⊥,# and Zoltán Kovács*,⊥
Department of Inorganic and Analytical Chemistry and Department of Colloid and EnVironmental
Chemistry, UniVersity of Debrecen, P.O. Box 21, Egyetem tér 1, Debrecen H-4010, Hungary,
Dipartimento di Chimica I. F. M., UniVersità di Torino, Via P. Giuria 7, I-10125 Torino, Italy,
Department of Radiology and AdVanced Imaging Research Center, UniVersity of Texas
Southwestern Medical Center, 5323 Harry Hines BlVd, Dallas, Texas 75390-8568, and Department of
Chemistry, UniVersity of Texas at Dallas, P.O. Box 830688, Richardson, Texas 75083-0688
Received December 20, 2007
A cyclen-based ligand containing trans-acetate and trans-methylenephosphonate pendant groups, H6DO2A2P, was synthesized
1
and its protonation constants (12.6, 11.43, 5.95, 6.15, 2.88, and 2.77) were determined by pH-potentiometry and H NMR
2–
spectroscopy. The first two protonations were shown to occur at the two macrocyclic ring N–CH2–PO3 nitrogens while the
third and fourth protonations occur at the two phosphonate groups. In parallel with protonation of the two -PO32– groups, the
protons from the NH+-CH2-PO3 are transferred to the N-CH2-COO– nitrogens. The stability constants of the Ca2+
,
2–
Cu2+, and Zn2+ (ML, MHL, MH2L, and M2L) complexes were determined by direct pH-potentiometry. Lanthanide(III) ions
(Ln3+) form similar species, but the formation of complexes is slow; so, “out-of-cell” pH-potentiometry (La3+, Eu3+, Gd3+, Y3+
and competitive spectrophotometry with Cu(II) ion (Lu3+) were used to determine the stability constants. By comparing the log
ML values with those of the corresponding DOTA (H4DOTA ) 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid) and
)
K
DOTP (H8DOTP ) 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetramethylenephosphonic acid) complexes, the order DOTA <
DO2A2P < DOTP was found for all the metal ion complexes examined here with the exception of the Ca2+ complexes, for
which the order is reversed. The relaxivity of Gd(DO2A2P) decreases between pH 2 and 7 but remains constant in the pH
range of 7 < pH < 12 (r1 ) 3.6 mM-1 s-1). The linewiths of the 17O NMR signals of water in the absence and presence of
Gd(DO2A2P) (at pH ) 3.45 and 8.5) between 274 and 350 K are practically the same, characteristic of a q ) 0 complex.
Detailed kinetic studies of the Ce3+ and Gd3+ complexes with DO2A2P showed that complex formation is slow and involves
a high stability diprotonated intermediate Ln(H2DO2A2P)*. Rearrangement of the diprotonated intermediate into the final complex
is an OH– assisted process but, unlike formation of Ln(DOTA) complexes, rearrangement of Ln(H2DO2A2P)* also takes place
spontaneously likely as a result of transfer of one of the protons from a ring nitrogen to a phosphonate group. The order of
the OH– assisted formation rates of complexes is DOTA > DO2A2P > DOTP while the order of the proton assisted dissociation
rates of the Gd3+ complexes is reversed, DOTP > DO2A2P > DOTA. 1H and 13C NMR spectra of Eu(DO2A2P) and Lu(DO2A2P)
were assigned using two-dimensional correlation spectroscopy (2D COSY), heteronuclear multiple quantum coherence (HMQC),
heteronuclear chemical shift correlation (HETCOR), and exchange spectroscopy (EXSY) NMR methods. Two sets of 1H NMR
signals were observed for Eu(DO2A2P) characteristic of the presence of two coordination isomers in solution, a twisted square
antiprism (TSAP) and a square antiprism (SAP), in the ratio of 93% and 7%, respectively. Line shape analysis of the 1H NMR
spectra of Lu(DO2A2P) gave lower activation parameters compared to La(DOTP) for interconversion between coordination
isomers. This indicates that the Ln(DO2A2P) complexes are less rigid probably due to the different size and spatial requirements
of the carboxylate and phosphonate groups.
Introduction
1,4,7,10-tetraacetic acid) are widely used in different fields
of medical diagnosis and therapy due to their high thermo-
Complexes of lanthanides formed with the cyclen deriva-
§
tive DOTA (H4DOTA ) 1,4,7,10-tetraazacyclododecane-
Department of Colloid and Environmental Chemistry, University of
Debrecen.
* To whom correspondence should be addressed. E-mail: zoltan.kovacs@
utsouthwestern.edu.
| Department of Radiology, University of Texas Southwestern Medical
Center.
†
Department of Inorganic and Analytical Chemistry, University of
⊥ Advanced Imaging Research Center, University of Texas Southwestern
Medical Center.
Debrecen.
‡
#
Università di Torino.
University of Texas at Dallas.
10.1021/ic7024704 CCC: $40.75  2008 American Chemical Society
Inorganic Chemistry, Vol. 47, No. 9, 2008 3851
Published on Web 04/02/2008

Kálmán et al.
dynamic and kinetic stabilities.1–4 These favorable complex-
ation properties are related to the rigid structure of the
lanthanide(III) (Ln3+) DOTA complexes in which the Ln3+
ion occupies a position in the coordination cage formed by
the four ring nitrogens and four carboxylate oxygens. The
many advantageous properties of the Ln(DOTA)- complexes
have attracted widespread interest in the synthesis of new
DOTA derivatives that may have even more favorable
complexation behavior. The macrocyclic 1,4,7,10-tetraaza-
cyclododecane backbone plays an important role in the
rigidity of these complexes so it is widely used in the design
of new ligands where one or more acetate groups of the
DOTA are replaced with some other charged or uncharged
pendant arms.1,2,5,6
useful as a Li, 23Na, or 39K NMR shift reagent in isolated
cells, perfused tissues, and intact animals.19–21 Gd(HDOTP)4-
has a larger effect on the relaxation rates of water protons
than expected for a q ) 0 complex. This has been interpreted
as due to the formation of strong H-bonds in the second
sphere between the phosphonate oxygens and the surrounding
water molecules.18 A few alkyl/aryl phosphinate and phos-
phonate monoester derivatives of DOTP have also been
reported. Compared to DOTP, the stability constants of the
Ln3+ complexes of these ligands are much lower but the
kinetic stabilities are somewhat higher.14,22 The presence of
the alkyl or aryl phosponate groups render these ligands and
their complexes significantly more hydrophobic than Ln-
(HDOTP)4-.23
7
The tetrakis(methylene phosphonate) derivative of cyclen,
DOTP (H8DOTP ) 1,4,7,10-tetraazacyclododecane-1,4,7,10-
tetramethylenephosphonic acid) was reported soon after
DOTA.7,8 The protonation and complexation behavior of
DOTP and the relaxation properties of its Gd3+ complex have
been studied in detail.9–17 While the stability constants of
the Ln(DOTP)5- complexes are higher than those of the
corresponding Ln(DOTA)- complexes, potential biomedical
applications of DOTP are more limited than DOTA partly
because of the high negative charge on the Ln(HnDOTP)(5-n)-
(n ) 1, 2, 3, and 4) complexes over the pH range 3–9.13
Gd(HDOTP)4- exists largely as a monoprotonated species
near pH 7, so the high negative charge of this complex plus
the absence of an inner-sphere water molecule16–18 gives
Gd(HDOTP)4- no particular advantage over Gd(DOTA)- as
a contrast agent for magnetic resonance imaging. However,
Tm(HDOTP)4- has been shown to induce large paramagnetic
shifts in ion-paired, NMR active cations so it has proven
More recently, some new DOTA derivatives have been
prepared in which one, two, and three carboxylate groups
were replaced with phosphonates. Taborsky et al. synthesized
the triacetate-monophosphonate derivative (DO3AP) and
reported its complexation properties.24 We synthesized the
bis(acetate)-bis(methylene phosphonate) (DO2A2P) and
mono(acetate)-tris(methylene phosphonate) (DO3PA) deriva-
tives, and while that work was in progress, Lacerda et al.
independently reported the synthesis of DO2A2P and
examined some potential nuclear medicine applications of
this ligand with 153Sm(III) and 166Ho(III).25 The availability
of DO3AP, DO2A2P, and DO3PA offers the possibility of
a systematic study of the complexation properties of cyclen
based ligands with mixed pendant arms. In this paper, we
discuss the synthesis and complexation properties of the
ligand, DO2A2P, the relaxation properties of Gd(DO2A2P),
and the solution structures of the Eu3+ and Lu3+ complexes
as determined by multinuclear NMR spectroscopy.
(1) The Chemistry of Contrast Agents in Medical Magnetic Resonance
Imaging; Merbach, A. E., Tóth, É., Eds.; Wiley: Chichester 2001.
(2) Caravan, P.; Ellison, J. J.; McMurry, T. J.; Lauffer, R. B. Chem. ReV.
1999, 99, 2293.
Experimental Details
Synthesis of DO2A2P. Paraformaldehyde (0.47 g, Aldrich) was
added to a mixture of 1,7-DO2A-tert-butyl ester26 (3.00 g) and
triethyl phosphite (2.86 g, Aldrich), and the mixture was stirred
for 4 days at room temperature. The volatile impurities were
removed by keeping the mixture in vacuum at 60 °C for 1 day to
yield a clear oil (5.15 g) that consisted of the ester intermediate
and a small amount of diethyl hydroxymethyl phosphonate. This
product was used without further purification in the acid hydrolysis
of the ethyl phosphonate and tert-butyl carboxylate ester groups.
The oil was dissolved in hydrochloric acid (20%, 75 mL), and the
(3) Volkert, W. A.; Hoffman, T. J. Chem. ReV. 1999, 99, 2269.
(4) Rösch, F.; Aronsson, E. F. In Metal Ions in Biological Systems; Siegel,
A., Siegel, H., Eds.; Marcel and Dekker, Inc.: New York 2004; Vol.
42, p 77.
(5) Alexander, V. Chem. ReV. 1995, 95, 273.
(6) Jacques, V.; Desreux, J.-F. In Chemistry of Contrast Agents in Medical
Magnetic Resonance Imaging; Merbach, A. E., Tóth, É., Eds.; Wiley:
Chichester, 2001; p 157.
(7) Kabachnik, I. M.; Medved, T. Ya.; Bel’skii, F. I.; Pisareva, S. A. IzV.
Akad. Nauk SSSR, Ser. Khim. 1984, 844.
(8) Sherry, A. D.; Malloy, C. R.; Jeffrey, F. M. H.; Cacheris, W. P.;
Geraldes, C. F. G. C. J. Magn. Reson. 1988, 76, 528.
(9) Geraldes, C. F. G. C.; Sherry, A. D.; Cacheris, W. P. Inorg. Chem.
1989, 28, 3336.
(19) Ramaprasad, S.; Lindquist, D. M.; Wall, P. T. J. EnViron. Sci. Heal.
A 1999, A34, 1839–1848.
(10) Lázár, I.; Hrncir, D. C.; Kim, W. D.; Kiefer, G. E.; Sherry, A. D.
Inorg. Chem., 1992, 31, 4422.
(20) Winter, P. M.; Seshan, V.; Makos, J. D.; Sherry, A. D.; Malloy, C. R.;
Bansal, N. J. Appl. Physiol. 1998, 85, 1806–1812.
(11) Delgado, R.; Siegfried, L. C.; Kaden, T. A. HelV. Chim. Acta 1990,
73, 140.
(21) Radford, N. B.; Babcock, E. E.; Richman, A.; Szczepaniak, L.; Malloy,
C. R.; Sherry, A. D. Magn. Reson. Med. 1998, 40, 544–550.
(22) Kim, W. D.; Kiefer, G. E.; Huskens, J.; Sherry, A. D. Inorg. Chem.
1997, 36, 4128.
(12) Aime, S.; Botta, M.; Terreno, E.; Anelli, P. L.; Uggeri, F. Magn. Reson.
Med. 1993, 30, 583.
(13) Sherry, A. D.; Ren, J.; Huskens, J.; Brücher, E.; Tóth, É.; Geraldes,
C. F. C. G.; Castro, M. M. C. A.; Cacheris, W. P. Inorg. Chem. 1996,
35, 4604.
(23) Aime, S.; Batsanov, A. S.; Botta, M.; Dickins, R. S.; Faulkner, S.;
Foster, C. E.; Harrison, A.; Howard, J. A. K.; Moloney, J. M.; Norman,
T. J.; Parker, D.; Royle, L.; Williams, J. A. G. Chem. Soc. Dalton
Trans. 1997, 3623–3636.
(14) Burai, L.; Király, R.; Lázár, I.; Brücher, E. Eur. J. Inorg. Chem., 2001,
813.
(15) Geraldes, C. F. G. C.; Brown, R. D. III.; Cacheris, W. P.; Koenig,
S. H.; Sherry, A. D.; Spiller, M. Magn. Reson. Med. 1989, 9, 94.
(16) Geraldes, C. F. G. C.; Sherry, A. D.; Kiefer, G. E. J. Magn. Reson.
1992, 97, 290–304.
(24) Taborsky, P.; Lubal, P.; Havel, J.; Kotek, J.; Hermann, P.; Lukes, I.
Collect. Czech. Chem. Commun. 2005, 70, 1909.
(25) Lacerda, S.; Campello, M. P.; Marques, F.; Gano, L.; Santos, I. In
Metal Ions in Biology and Medicine; Alpoim, M. C., Morais, P. V.,
Santos, M. A., Cristovao, A. J., Centeno, J. A., Collery, P., Eds.; John
Libbey Eurotext: Paris , 2006; Vol. 9, p 46.
(17) Avecilla, F.; Peters, J. A.; Geraldes, C. F. G. C. Eur. J. Inorg. Chem.
2003, 4179.
(18) Botta, M. Eur. J. Inorg. Chem. 2000, 399.
(26) Kovács, Z.; Sherry, A. D. Synthesis 1997, 759–763.
3852 Inorganic Chemistry, Vol. 47, No. 9, 2008

DO2A2P and its Complexes with Ca2+, Cu2+, Zn2+, and Ln(III)
solution was refluxed for 2 days. The hydrochloric acid was
removed by rotary evaporation, and the residue was redissolved in
water (75 mL), treated with activated carbon (1 g), filtered, and
evaporated again by rotary evaporation. The residue was redissolved
in a small amount of water (5 mL), and ethanol (50 mL) was added
in small portions. A white amorphous solid precipitated. The
solution was decanted, and the residue was redissolved in water (5
mL) and stirred for 24 h, over which period the product separated
as a white crystalline solid. The solid was filtered, washed with
water (2 × 3 mL), and dried in air to give 2.38 g (66.6%) of product.
Anal. calcd for C14H30N4O10P2: C, 35.30; H, 6.35; N, 11.76. Found:
values were used to calculate the H+ concentrations from the
experimentally measured pH values as described by Irving et al.27
The ion product of water was determined from the same HCl-KOH
titration experiments over the pH range of about 11–12 (pKw
)
13.83).
The protonation and stability constants were evaluated from the
titration data using the program PSEQUAD.28 For the calculation
of the stability constant of Lu(DO2A2P), in addition to the
concentrations and pH values, the measured absorbances and the
molar absorptivity values of Cu(DO2A2P) were also taken into
account.
Relaxation Rates. The longitudinal water proton relaxation times
(T1) were measured using a Stelar Spinmaster (Mede, Pavia, Italy)
spectrometer operating at 0.47 T and a standard inversion–recovery
sequence. A typical 90° pulse width was 3.5 µs, and the reproduc-
ibility of the T1 data was (0.5%. The temperature was controlled
with a Stelar VTC-91 air-flow heater equipped with a copper-con-
stantan thermocouple (uncertainty (0.1 °C).
Kinetic Studies. The rate of formation of Ce(DO2A2P) was
studied in 1.0 M KCl by spectrophotometry at 312 nm in a
thermostatted cell holder (25 °C) using a Cary 1E spectrophotom-
eter. The concentration of the Ce3+ was 4 × 10-4 M, and DO2A2P
was used in 5-20-fold excess. MES buffer (0.05 M) was used to
keep the pH constant. The absorbance values measured at different
t time intervals were fitted to eq 1 for the calculation of the first
order rate constants (kobs or kp):
1
C, 35.27; H, 6.46; N, 11.58. H NMR (400 MHz, D2O, ND3, pD
∼ 9–10) δ ) 3.31 (8H, s br, NCH2 ring), 3.16 (4H, s, CH2COOH),
2.94–2.92 (8H, m br, NCH2 ring), 2.85 (4H, d, JH-P ) 11.2 Hz
CH2PO3H2). 13C NMR (100 MHz, D2O, ND3, pD ∼ 9–10) δ )
179.13 (COOH), 57.44 (NCH2COOH), 51.57 (NCH2 ring), 50.49
(NCH2PO3H2, JC-P ) 126.41 Hz), 49.66 (NCH2 ring). 31P NMR
(161 MHz, D2O, ND3, pD ∼ 9–10) δ ) 6.24. MALDI-TOF: m/z,
477.6 (100% [M + H]+).
Equilibrium Measurements. The chemicals used in the experi-
ments were of the highest analytical grade. For the preparation of
the LnCl3 solutions, Ln2O3 (Fluka, 99.9%) was dissolved in 6.0 M
HCl and the excess acid was evaporated. An aqueous solution of
CeCl3 was prepared from CeCl3 ·7H2O (Aldrich, 99.9%). The
concentration of the LnCl3, CaCl2, CuCl2, and ZnCl2 solutions were
determined by complexometric titration with standardized
Na2H2EDTA and xylenol orange (LnCl3), murexide (CuCl2, ZnCl2),
and Patton & Reeder (CaCl2) as indicator. The concentration of
DO2A2P was determined by pH-potentiometric titration in the
presence and absence of a large (40-fold) excess of CaCl2.
The protonation constants of DO2A2P and the stability constants
of the complexes formed with Ca2+, Cu2+, and Zn2+ have been
determined by pH-potentiometric titration. The metal/ligand con-
centration ratios were 1:1 and 2:1 (the concentration of the ligand
was generally 2.0 mM). Due to slow complex formation, the
stability constants of the Ln(DO2A2P) complexes were determined
by use of an “out-of-cell” technique as follows. Nine Ln3+(2
mM)-DO2A2P (2 mM) samples (1.5 mL) were prepared with pH
values in the pH range about 1.8–3.8 (depending on the Ln3+) at
equilibrium. The samples contained 1.0 M KCl to maintain constant
ionic strength. The samples were sealed from air maintained at 50
°C for 3 weeks followed by an additional 1 week at 25 °C to reach
equilibrium. In the case of Lu(DO2A2P), a competition reaction
was used to determine the stability constant because of solubility
problems. Nine samples were prepared containing Cu2+ (0.1 mM),
DO2A2P (0.1 mM), and Lu3+ of different concentrations, and after
reaching equilibrium, the pH and the absorbance of the samples
were measured. The absorbance values were determined at nine
different wavelengths between 310 and 350 nm. The molar
absorptivity of Cu(DO2A2P) was determined in a separate experi-
ment. In addition to the protonation constants of the ligand, the
protonation constants of the complexes determined in separate
titration experiments were also used in the calculation of the stability
constants of Ln(DO2A2P) complexes.
(Ao - Ae)
(At - Ae)
1
kp ) ln
t
(1)
where A0, At, and Ae are the absorbance values measured at the
start of the reaction, at time t, and at equilibrium, respectively.
The formation and dissociation rates of Gd(DO2A2P) were
studied by relaxometry at 25 °C with an MS-4 NMR spectrometer
(Institute Jozef Stefan, Ljubljana) at 9 MHz. The temperature of
the sample holder was controlled with thermostatted air. The
longitudinal relaxation times were measured by the “inversion–re-
covery” method (180°-τ-90°) using 6–8 different τ values, and
the sample relaxivity was determined from these data by dividing
by the solution concentration in millimolar. Complex formation was
studied for a mixture of 1.0 mM Gd3+ and 20 mM DO2A2P
solutions at different pH values (5.3 < pH < 6.5), a condition that
insures complete formation of the intermediate species, Gd-
(DO2A2P)*.
The acid assisted dissociation rate of Gd(DO2A2P) was deter-
mined in 0.1–1.0 M HCl at constant ionic strengh ([HCl] + [KCl]
) 1.0 M) at 25 °C. The first order rate constants (kobs or kp) were
calculated using eq 1 except that instead of the absorbance values
A0, At, and Ae, relaxivities measured at the start of the reaction (r10),
at time t (r1t), and at equilibrium (r1e) were used in the fitting. The
relaxivities were obtained from the relaxation times, r1 ) 1/T1m
-
1/T1w, where T1m and T1w are the water proton longitudinal
relaxation times measured in the presence and absence of Gd3+
respectively.
,
17O NMR. Variable temperature 17O NMR measurements were
performed on nonspinning samples at 14.1 T using a Bruker Avance
600 spectrometer. A capillary filled with D2O was used as an
external lock. Experimental settings were as follows: spectral width
10 000 Hz, 90° pulse (7 µs), acquisition time 10 ms, 1000 scans.
The pH measurements and titrations were performed in the pH
range 1.7–12.5 with a Radiometer PHM93 pH meter, an ABU 80
autoburette, and a Metrohm 6.0234.100 combined electrode. The
stirred samples (15 mL) were thermostatted at 25 °C, and the
titrations were performed under an inert gas atmosphere by bubbling
Ar across the solutions. KH-phthalate (pH ) 4.005) and borax (pH
) 9.180) buffers were used to calibrate the pH meter. A 0.01 M
HCl solution (1.0 M for KCl) was titrated with standardized KOH
solution. The differences between the measured and calculated pH
(27) Irving, H. M.; Miles, M. G.; Pettit, L. D. Anal. Chim. Acta 1967, 38,
475.
(28) Zékány, L.; Nagypál, I. In Computional Methods for Determination
of Formation Constants; Legett, D. J., Ed.; Plenum: New York, 1985;
p 291.
Inorganic Chemistry, Vol. 47, No. 9, 2008 3853

Kálmán et al.
Table 1. Protonation Constants of the Ligands DO2A2P, DOTA, and
DOTP (25 °C)
DO2A2P
DOTPa
DOTAb
1.0 M KCl
12.6 (0.01)
11.43 (0.01)
5.95 (0.01)
6.15 (0.01)
2.88 (0.01)
2.77 (0.01)
0.1 M Me4N(NO3)
0.1 M Me4NCl
12.6
9.70
4.50
4.14
2.32
log KH1
log KH2
log KH3
log KH4
log KH5
log KH6
13.7c
12.2c
9.28
8.09
6.12
5.22
a Reference 13. b Reference 32. c Determined by 1H NMR titration
without control of the ionic strength.
Figure 1. Structures of the ligands and the numbering of the ring carbon
DOTA and DOTP (the charges of the ligands and complexes
will be indicated only when necessary). Since the first two
protonation constants in DOTA and DOTP are quite high,
one would anticipate that DO2A2P would also have two high
protonation constants characteristic of two macrocyclic ring
nitrogen protonations. The protonation constants, log KHi (KHi
) [HiL]/[Hi-1L][H+]) were calculated from the pH-poten-
tiometric titration data (Table 1). The log KH1 value is high,
but it could be calculated from the pH-potentiometric data
because the titration was performed up to pH ) 12.5, where
the formation of the fully deprotonated ligand L6- was close
to 50%. At such high pH values, the ligand DO2A2P6-
probably forms complexes with K+ ions and so the log KH1
would have been even higher if Me4NCl had been used to
maintain constant ionic strength. Analogous to DOTA and
DOTP,9,30 protonation of trans-ring nitrogens is expected
to occur first followed by the protonation of the phosphonate
groups by the third and fourth protons and the carboxylates
by the fifth and sixth protons. The log KH3 and log K4H values
are associated with the protonation of the phosphonate groups
while the log KH5 and log K6H characterize the protonation of
the two carboxylates. These protonation constants are so-
called “macroscopic” constants, which can be used in the
description of protonation and complexation equilibria.31 To
obtain more information on the sites of protonation, 1H and
31P NMR spectra of the ligand were also recorded as a
function of pH. The ring methylene protons of DO2A2P give
rise to two well resolved triplet signals in the entire pH range
investigated. The signals of the methylenephosphonate and
acetate methylene protons were doublet and singlet, respec-
tively. The 31P signal of the phosphonate group was broad
atoms.
The samples contained water enriched to 2.6% 17O (Yeda, Israel).
The half-height line width of the 17O NMR signal (∆ν1/2) was
determined in the presence and absence of Gd(DO2A2P) (5.42 ×
10-2 M, 1.0 M KCl).
1
NMR Experiments. The H and 31P NMR chemical shifts of
DO2A2P were determined as a function of pH to evaluate some of
the protonation constants. A 0.01 M solution of the ligand in H2O
with 10% D2O was prepared. The pH was adjusted by stepwise
addition of a solution of KOH and HCl (both prepared in H2O).
The protonation constants were determined by fitting of the
chemical shift versus pH data using Micromath Scientist, version
2.0 (Salt Lake City, UT).
NMR spectra of some metal ion complexes of DO2A2P were
1
measured in 0.07 M samples prepared in D2O. H, 13C, and 31P
NMR spectra were collected using either a JEOL EX-400 (9.4 T)
or Bruker DRX 360 NMR spectrometer. Variable temperarture 13
C
NMR measurements were carried out using a Bruker DRX 360
NMR spectrometer equipped with Bruker VT-1000 thermocontrol-
ler. The 31P NMR spectra of the Eu(DO2A2P) complex were
1
recorded using frequency-selective H decoupling. The chemical
1
shifts are reported in ppm, relative to TMS for H and 13C and
H3PO4 for 31P as the external standard. The H-H correlation
spectroscopy (COSY) and exchange spectroscopy (EXSY) spectras
were collected by using gradient pulses in the z direction with the
standard JEOL pluse program (in the case of EXSY spectra, the
mixing time (τM) was 10 ms.) The H-C correlation spectrum was
recorded on JEOL EX-400 spectrometer in inverse mode by using
gradient pulses in the z direction with the usual JEOL heteronuclear
multiple quantum coherence (HMQC) pulse sequence.
Results and Discussion
1
in the entire pH range. The chemical shifts of the H and
Synthesis of DO2A2P. DO2A2P (Figure 1) was synthe-
sized in good yields by reacting the tert-butyl ester of
1,4,7,10-tetraazacyclododecane-1,7-bis(acetic acid)26 with
paraformaldehyde and triethyl phosphite in a Michaelis-
Arbuzov type reaction. This method appears to work well
for the synthesis of methylenephosphonate derivatives of
cyclen.29 The final product crystallized from an aqueous
solution as a zwitterion as shown by elemental analysis. The
31P signals as a function of pH are presented in Figure 2.
In the pH-range 13.5–10, addition of two equivalents of
acid to DO2A2P6- brought about a significant downfield shift
of both the 31P and the proton signal of the CH2-
2-
N-CH2-PO3 ring methylene groups indicating that the
2-
first two protons are attached to the N-CH2-PO3 ring
nitrogens. The protonation of these nitrogens results in the
formation of strong H-bonds between the NH+ and the PO3
2-
1
molecular structure and purity were also confirmed by H,
groups, which explains the high values of log KH1 and log
KH2 . The proton signals of the acetate methylene protons were
also shifted, which is interpreted by the dissociation of the
13C, and 31P NMR and MALDI-TOF mass spectrometry.
pH-Potentiometric Studies. The ligand H6DO2A2P (H6L)
has two acetate and two methylenephosphonate pendant
arms, and so, its properties are expected to be similar to
(30) Desreux, J. F.; Merciny, E.; Loncin, M. F. Inorg. Chem. 1981, 20,
987.
(29) Burai, L.; Ren, J.; Kovács, Z.; Brücher, E.; Sherry, A. D. Inorg. Chem.
1998, 37, 69–75.
(31) Beck, M. T.; Nagypál, I. Chemistry of Complex Equilibria; Akadémiai
Kiadó: Budapest, 1990.
3854 Inorganic Chemistry, Vol. 47, No. 9, 2008

DO2A2P and its Complexes with Ca2+, Cu2+, Zn2+, and Ln(III)
Table 2. Stability and Protonation Constants of the Complexes of
DO2A2P (25 °C, 1.0 M KCl), DOTA, and DOTP
ML
MHL
MH2L
M2L
MLb MLc
Ca2+ 15.1 (0.07) 7.33 (0.01) 6.40 (0.04) 3.05 (0.07) 17.2d 10.3g
Cu2+ 24.9 (0.04)a 6.79 (0.05) 6.30 (0.03) 4.72 (0.04) 22.2e 25.4h
Zn2+ 22.5 (0.03) 6.92 (0.03) 6.34 (0.01) 3.47 (0.08) 21.1e 24.8h
La3+ 23.3 (0.08) 5.40 (0.08) 4.49 (0.08) 2.77 (0.05) 22.9f 27.6i
Eu3+ 25.6 (0.04) 5.34 (0.05) 4.64 (0.05) 3.78 (0.07) 24.7f 28.1i
Gd3+ 25.7 (0.09) 5.32 (0.08) 4.55 (0.08) 3.75 (0.07) 24.7f 28.8i
Y3+ 26.6 (0.09) 5.19 (0.09) 4.27 (0.09) 3.10 (0.05)
Lu3+ 26.4 (0.09)a 5.26 (0.02) 4.42 (0.02) 4.33 (0.04) 25.4f 29.6i
a Obtained by spectrophotometry. b DOTA. c DOTP. d Reference 33.
e Reference 34. f Reference 35. g Reference 7. h Reference 36. i Reference 13.
well with the potentiometric data but, surprisingly, both the
NMR and potentiometric log KH4 is larger than log KH3 , which
is the opposite of the usual decreasing order of protonation
constants. This unexpected finding is probably due to the
peculiar structure of the triprotonated ligand H3DO2A2P3-.
1
Interpretation of the H NMR chemical shifts suggests that
in the triprotonated species one phosphonate and two
neighboring (cis) nitrogens are protonated. Due to the
electrostatic repulsion between these two protonated nitro-
gens the stability (and consequently the concentration) of
this intermediate is low. The attachment of the fourth proton
is favorable since it leads to the formation of the symmetric,
more stable tetraprotonated species H4DO2A2P2-. This
cooperativity in the protonation steps reverses the expected
order of protonation constants.
Stability Constants of Complexes. Complex formation
equilibria of DO2A2P were studied with Ca2+, Cu2+, Zn2+,
and Ln3+ ions. The stability (KML, KM2L) and protonation
constants (KMHiL) of the complexes have been defined by
eqs 2-4 (the charges of the species are omitted for
simplicity).
Figure 2. 1H- (A) and 31P NMR chemical shifts (B) of DO2A2P as a
function of pH. (A) N-CH2-COO- (a); N-CH2-PO32- (b, center of the
2-
doublet); CH2-N-CH2-COO- (c); CH2-N-CH2-PO3 (d).
complex K+-DO2A2P6- formed at higher pH values.
1
Similar H and 31P NMR shifts have been observed during
the titration of H8DOTP with Me4NOH and NaOH when
Na+-DOTP8- complexes were formed.9
In the pH interval about 10–7, there is practically no
change in the chemical shift values (Figure 2). Further
lowering of the pH resulted in the strong downfield shift of
the 31P signal between pH 7 and 5, confirming that the log
KH3 and log KH4 values are related to the protonation of the
phosphonate groups. However, in this same pH range the
signals of the N-CH2-PO3 and CH2-N-CH2-PO3 meth-
ylene protons shifted to lower frequencies (upfield). These
findings can be explained by assuming that parallel with the
[ML]
[M][L]
KML
KMH L
)
(2)
(3)
[MHiL]
)
2-
[MHi-1L][H+]
[M2L]
i
protonation of the PO3 groups the protons from the
NH+-CH2-PO32- nitrogens are transferred to the N-CH2-
COO- nitrogens, which also result in a strong downfield shift
of the signal of the acetate methylene protons. (Figure 2A).
At pH values below 4, the carboxylate groups of the ligand
start to be protonated. The protonation constants log KH3 and
log KH4 are much lower than those for the DOTP (Table 1)
probably because for the protonation of the phosphonate
groups to occur, the first two protons must be transferred
from the NH+-CH2-PO32- nitrogens to the N-CH2-COO-
nitrogens. This proton transfer is not favorable as the H-bonds
KM2L
)
(4)
[ML][M]
The complexation reactions between DO2A2P and Ca2+,
Cu2+, or Zn2+ were fast enough to determine the equilibrium
constants of eqs 2-4 in direct titration experiments. On the
other hand, the formation reactions of the Ln(DO2A2P)
complexes were slow and the stability constants (log KML
)
were determined by the out-of-cell method. The log KMHiL
and log KM2L values have been determined by direct titration
of the acidified complex solutions prepared before the
2-
between the protonated nitrogens and PO3 groups are
strong. Calculation of the first four protonation constants
from the H NMR chemical shift data in the pH range
measurements. The protonation constants KMHL and KMH2L
,
1
determined in the separate experiments, were taken into
account in the calculation of the log KML stability constants
of the complexes Ln(DO2A2P). The obtained equilibrium
constants of eqs 2-4 are presented in Table 2.
Generally, the stabilities of DOTP complexes is higher
than those of DOTA complexes, with the exception of
Ca-DOTP, whose stability is much lower than that of
Ca-DOTA. As expected, the stability constants of the
3.7–13.5 gave log KH1 ) 12.8, log KH2 ) 11.7, log KH3 ) 5.93,
and log KH4 ) 6.25. The first two protonation constants are
somewhat higher than those obtained by pH-potentiometry
(Table 1), probably because the solution samples used for
1
the H NMR titration experiments did not contain KCl to
keep the ionic strength constant. The value obtained from
1
the H NMR study for third and fourth protonation agrees
Inorganic Chemistry, Vol. 47, No. 9, 2008 3855

Kálmán et al.
DO2AP complexes ML were found to be between the log
KML values of the complexes of DOTA and DOTP. This is
also true for the complexes of Ca2+, but in this case the order
of log KML values is Ca(DOTA) > Ca(DO2A2P) > Ca-
(DOTP). Thus, with the exeption of the Ca2+, which prefers
the carboxylate groups over the phosphonates, the replace-
ment of two carboxylates of DOTA with two phosphonate
groups leads to an increase in the stability of the complexes.
Similar results were found for the Ln3+ complexes of the
monophosphonate derivative of DOTA (DO3AP) by Tabor-
sky et al.24 (The stability constants of the Ln(DO3AP)
complexes determined by Taborsky in 0.1 M Me4NCl are
somewhat higher than those of the Ln(DO2A2P) complexes
measured by us in 1.0 M KCl. This probably reflects the
effect of K(DO2A2P) formation as under identical conditions
the opposite order of stabilities is expected.) The stabiliy
constants of the Ln(DO2A2P) complexes, similarly to the
Ln-complexes of DOTA, DO3AP, and DOTP, slightly
increase with decreasing ionic size of lanthanides (Table 2).24
The protonation of the M(DO2A2P) complexes probably
occurs at the phosphonate or carboxylate groups in the same
way as the protonation of Ln(DOTP)5- or Ln(DOTA)-
complexes.13,32 The protonation of the ring nitrogens is not
likely, particularly for the Cu2+ and Zn2+ complex. Surpris-
ingly, the first and second protonation constants log KMHL
and log KMH2L of the Ca2+-, Cu2+-, and Zn2+-DO2A2P
complexes are higher than the log KH3 and log KH4 protonation
constants of the phosphonate groups of the free ligand. This
interesting observation can be interpreted by assuming the
presence of uncoordinated pendant arms in these complexes
due to the lower coordination number of Cu2+ and Zn2+.
Interestingly, the stability constant of the Zn(DO2A2P) is
very similar to the that of the Zn2+ complex of 1,4,7,10-
tetraazacyclododecane-1,7-bis(methylenephosphonic acid)
(DO2P) (log KZn(DO2P) ) 21.2)29], a ligand similar to
DO2A2P except lacking the two acetate pendant arms. The
nearly identical stability of the Zn2+-complexes suggest that
the carboxylates do not play a significant role in the
formation of Zn(DO2A2P). Unlike the free diprotonated
ligands, H-bond formation between the ring nitrogens and
phosphonate groups is not possible in the Cu(DO2A2P) and
Zn(DO2A2P) complexes, so protonation of these complexes
at the phosphonates may take place at higher pH values. For
the complexes of lanthanides, where all donor atoms in
DO2A2P act as ligands for the metal ion, the log KMHL and
log KMH2L values are lower than the log KH3 and log K4H
protonation constants of the ligand. This indicates the
coordinated phosphonate groups can be protonated in these
complexes. The third protonation constants of the Cu2+ and
Zn2+ complexes are low (2.46 and 1.96, respectively), so
Figure 3. Relaxivity values of Gd(DO2A2P) at 25 °C.
these must reflect protonation of the carboxylate groups. The
stability constants of the dinuclear complexes formed with
Zn2+, Cu2+, and Ln3+ ions are fairly similar (Table 2)
probably because electrostatic interactions predominate in
the formation of these complexes.
Relaxation Properties of Gd(DO2A2P). The relaxivity
of Gd(DO2A2P) as a function of pH at 20 MHz is shown in
Figure 3. Protonated complexes are present in the pH range
2–7 so these contribute to the relaxation rates but at pH
values above 7 where the complex is fully deprotonated, the
relaxivity is relatively constant at 3.6 ( 0.05 mM-1 s-1. This
value is most consistent with a q ) 0 complex with a small
outer-sphere contribution to relaxivity, much like that noted
for Gd(DOTP) previously. The higher relaxivity of the
protonated complexes can be explained by faster proton
exchange due to the protonation of the noncoordinated
2-
oxygen atoms of the PO3 group.
It is known that Gd(DOTA) has one inner-sphere coor-
dinated water molecule while Gd(DOTP) has none.16–18 The
presence of one (or more) water molecule(s) in the inner
sphere of the paramagnetic Gd3+ results in a broadening in
the 17O NMR signal of water via the exchange of the
coordinated H2O with the bulk water.37 The half-widths
(∆ν1/2) of the 17O NMR signal have been determined in 17
O
enriched (2.6%) water in the presence and absence of
Gd(DO2A2P) at pH ) 3.45 and at pH ) 8.5 in the
temperature range 273–350 K. The ∆ν1/2 values (Figure S1,
Supporting Information) are virtually identical in the presence
and absence of Gd(DO2A2P), showing that the presence of
the paramagnetic complex has no effect on the 17O NMR
water linewidths. This finding indicates that neither the
protonated nor deprotonated Gd(DO2A2P) complexes have
an inner-sphere water molecule.
Kinetics of Formation of Ln(DO2A2P) Complexes. The
formation of Ce(DO2A2P) and Gd(DO2A2P) complexes
below pH 7 were slow enough so that the rates of reactions
could be studied by conventional techniques. The rates
of formation of Ce(DO2A2P) in the pH range 5.3–6.5 were
measured in the presence of both metal and ligand excess
to afford pseudo-first-order rate constants (kobs). The plot
of kobs values against the DO2A2P or Ce3+ concentration
(32) Burai, L.; Fábián, I.; Király, R.; Szilágyi, E.; Brücher, E. J. Chem.
Soc., Dalton Trans. 1998, 243.
(33) Delgado, R.; Da Silva, J. J. R. F. Talanta 1982, 72, 567.
(34) Chaves, S.; Delgado, R.; Da Silva, J. J. R. F. Talanta 1992, 39, 249.
(35) Cacheris, W. P.; Nickle, S. K.; Sherry, A. D. Inorg. Chem. 1987, 26,
958.
(36) Martell, A. E.; Smith, R. M. Critical Stability Constants, Vols. 1–6;
Plenum Press: New York, 1974–1989.
(37) Tóth, É.; Helm, L.; Merbach, A. E. Top. Curr. Chem. 2002, 221, 61.
3856 Inorganic Chemistry, Vol. 47, No. 9, 2008

DO2A2P and its Complexes with Ca2+, Cu2+, Zn2+, and Ln(III)
Figure 5. First-order rate constants for the formation reaction of Ce-
(DO2A2P) (() and Gd(DO2A2P) () ([Ce3+] ) 4 × 10-4 M, [DO2A2P]
) 8 × 10-3 M; [Gd3+] ) 1 × 10-3 M, [DO2A2P] ) 1 × 10-2 M, 1.0 M
KCl, 25 °C).
Figure 4. Absorption spectra of the Ce3+ + DO2A2P system at the start
of the reaction (t0) and after 100 min, at equilibrium (te). ([Ce3+] )
[DO2A2P] ) 4 × 10-4 M, 25 °C, pH ) 5.33).
gave a saturation curve indicating the formation of a
intermediate.38 This was also confirmed by spectropho-
tometry (Figure 4).
DO2A2P)* with an absorption maximum at 304 nm and the
product, Ce(DO2A2P) showing an absorption maximum at
309 nm.
During formation of Ce(DOTA), a diprotonated intermedi-
ate, Ce(H2DOTA)* was detected by spectrophotometry and
was shown to have two protonated macrocyclic ring nitrogens
(in trans postions) while Ce3+ is coordinated only to the
carboxylate oxygens.32,38 We assumed that fast formation
of a similar diprotonated intermediate, Ce(H2DO2A2P)*,
might make it possible to measure the stability constant of
this species by a direct pH-potentiometric titration. Indeed,
titration of a Ce3+(2 × 10-3 M)-DO2A2P(2 × 10-3 M)
solution in the pH range 1.8–4.0 gave a stability constant
log KCeH2L* ) 10.4 (0.07) and protonation constant log
KHCeH2L* ) 3.9 (0.03) for this intermediate. During the
calculation of the stability constant of Ce(H2DO2A2P)*
lower standard deviation values were obtained by assuming
the formation of a triprotonated intermediate as well. It is
very likely that the species Ce(H2DO2A2P)* has a similar
structure to Ce(H2DOTA)* with the Ce3+ ion coordinated
to two carboxylate oxygens and two phosphonate oxygens.
The stability constant of the intermediate Ce(H2DO2A2P)*
is significantly higher than that of Ce(H2DOTA)* (log KHCeH2L*
) 4.538]), probably due to the participation of the phospho-
nates in coordination to the metal ion. The stability constant
of La(H2DO2P), where again the protons are located on two
macrocyclic nitrogens, is also significantly lower (log K(LaH2L)
) 7.6)29] than that of the Ce(H2DO2A2P)*. These data
strongly suggest that both the phosphonate and carboxylate
groups are coordinated in the Ce(H2DO2A2P)* intermediate.
For the formation of Ce(DO2A2P), the Ce3+ must displace
the two protons from the macrocyclic ring nitrogens before
entering the coordination cage of the macrocycle. Since
removal of these protons from the very basic ring nitrogens
is not favored, formation of the final complex is slow. In
the pH range 5.3–6.5, where the formation rates of the
complexes were studied, the concentration of the triproto-
nated intermediate is very low, so its role in the complex
formation is probably negligible. This assumption was
confirmed by spectrophotometric studies. The presence of
isosbestic points in the absorption spectra (Figure 4) indicates
the formation of two species, the intermediate Ce(H2-
The rates of formation of Ce(DO2A2P) and Gd(DO2A2P)
were studied in the presence of excess ligand ([L]t/[Ln3+] )
20) because the saturation values in the plots of kobs versus
[Ce3+] showed a decreasing trend with increasing Ce3+
concentration. We assumed that the decrease in kobs observed
at high Ce3+ concentrations reflected formation of a dinuclear
diprotonated intermediate. Ce2(H2DO2A2P)*. The stability
constants of the dinuclear intermediate [Ce(H2DO2A2P)- +
Ce3+] was determined by pH-potentiometry in the pH range
about 1.8–4.0, exploiting the competition between the H+
and the excess Ce3+ ions for the intermediate,
Ce(H2DO2A2P)*. The stability constant of the dinuclear
intermediate, Ce2(H2DO2A2P)*, was found to be log KCe2H2L
) 3.2 (0.07). To avoid any complications that might arise
from the formation of the dinuclear intermediate, all further
kinetic studies were performed in the presence of ligand
excess. Under pseudofirst-order conditions the formation rate
of the Ln(DO2A2P) complexes can be described by
d[LnL]t
) kobs[Ln3+]
(5)
dt
where [LnL]t is the total concentration of the complex,
Ln(DO2A2P). The kobs values obtained at different pH values
show a direct proportionality with the increase of the OH-
concentration (Figure 5). Unlike the results obtained for the
formation of Ln(DOTA) complexes, the straight line obtained
by plotting the kobs values against the [OH-] has a positive
intercept. Similarly, when values of kobs for Gd(DO2A2P)
were plotted against the OH- concentration, that graph also
had a positive intercept (Figure 5). Thus, the rate constants,
kobs, obtained for the formation reactions of Ce(DO2A2P)
and Gd(DO2A2P) can be expressed as:
kobs)kL + kOH[OH-]
(6)
where kOH is the rate constant characterizing the transforma-
tion of the intermediate Ln(H2DO2A2P)* to product cata-
lyzed by OH- while the rate constant, kL, reflects “sponta-
neous” transformation of the intermediate to product,
Ln(DO2A2P).
The formation of Ln(DO2A2P) complexes by the OH-
assisted pathway presumably occurs via mechanism similar
to that proposed for the formation of Ln(DOTA)- com-
(38) Tóth, É.; Brücher, E.; Lázár, I.; Tóth, I. Inorg. Chem. 1994, 33, 4070.
(39) Kumar, K.; Chang, C. A.; Tweedle, M. F. Inorg. Chem. 1993, 32,
587.
Inorganic Chemistry, Vol. 47, No. 9, 2008 3857

Kálmán et al.
Table 3. Rate Constants, Characterizing the Formation of the Ce3+ and Gd3+ Complexes of DOTA, DO3AP, DO2A2P, and DOTP (25 °C)
DOTA
DO3APc
DO2A2P
DOTPd
kOH (M-1 s-1
)
kOH (M-1 s-1
)
kL (s-1
)
kOH (M-1 s-1
)
kOH (M-1 s-1
)
a
Ce3+
Gd3+
3.5 × 106
9.6 × 105
(9.1 ( 1) × 10-5
(1.7 ( 0.04) × 105
b
5.9 × 106
9.0 × 104
(3.0 ( 0.3) × 10-5
(6.6 ( 0.14) × 104
7.2 × 103
a Reference 32. b Reference 39. c Reference 24. d Reference 14.
plexes.32,38 It is likely that the diprotonated intermediate is
in equilibrium with a monoprotonated intermediate:
experimental data, we assume that the extent of preorga-
nization of the ligands (or more accurately, the intermedi-
ary metal complexes) decreases with increasing number
of the phosphonate groups, since the electrostatic repulsion
Ln(H2DO2A2P)* h Ln(HDO2A2P)* + H+
(7)
between the PO3 groups or between PO3 and COO-
groups is larger than that between multiple COO- groups.
The formation rates are directly proportional to the [OH-]
for all the complexes indicating a similar formation
mechanism, which is the deprotonation of a ring nitrogen
of the monoprotonated intermediate. However, the increase
of the number of the phosphonate groups of the ligand
results in an increase in the negative charge of the
intermediate, which may slow down the rate of the proton
loss. In addition, high negative charge of the pendant arm
functional groups results in the formation of a more stable
intermediate (the Ln3+ is coordinated by the donor atoms
of the pendant groups) and the rearrangement rate of the
intermediary complex to the final product is slower. Such
effects have been implicated in the slow formation of the
Ln(DOTP)5- complexes.
2-
2-
This equilibrium strongly favors Ln(H2DO2A2P)* because
the protonation constant of Ln(HDO2A2P)* is large,
presumably only somewhat lower than the protonation
constant of the monoprotonated ligand HDO2A2P5-, log
KH2 ) 11.43 (Table 1). (In the case of the formation of
DOTA complexes, log K2H ) 9.70 and the protonation
constant of the intermediate, Ce(HDOTA)*, is log KCHeHL*
) 8.64.32) The formation of the product Ln(DO2A2P) can
occur via the formation of the monoprotonated intermedi-
ate with the rate determining step being the deprotonation
of the species Ln(HDO2A2P)*. Thus, the formation rate
of the complex Ln(DO2A2P) is directly proportional to
the concentration of the intermediate Ln(HDO2A2P)*,
which is directly proportional to the [H+]-1 or [OH-], as
predicted by equilibrium (7).32 The deprotonation of the
intermediate may take place with the participation of a
general base such as a water molecule. This is the only
plausible pathway for the formation of the Ln(DOTA)-
complexes in the absence of other general bases and in
this case the rate expression (6) does not contain the term
kL. However, the intermediates Ln(H2DO2A2P)* contain
two phosphonate groups, which are more basic than the
carboxylates of the Ln(H2DOTA)* and one of these
phosphonate groups may accept a proton transferred from
a ring nitrogen of the diprotonated intermediate. This
process can lead to a diprotonated intermediate which is
protonated at one ring nitrogen and one phosphonate
group. Transfer of the proton from the ring nitrogen to a
water molecule in the rate determining step yields
Ln(HDO2A2P)2-, in which a phosphonate group is
protonated, and the formation of this complex is inde-
pendent from the H+ (or OH-) concentration. Over the
pH intervals investigated here, the complexes exist as
nonprotonated Ln(DO2A2P), monoprotonated Ln-
(HDO2A2P) and diprotonated Ln(H2DO2A2P) species as
expected on the basis of the protonation constants obtained
by pH-potentiometric titration (Table 2). The mono and
diprotonated complexes are protonated at the phosphonate
oxygens, while the other oxygen of the phosphonate
groups is coordinated to the Ln ion located in the
coordination cage.
Kinetics of Dissociation of Gd(DO2A2P). High kinetic
inertness of Ln3+ chelates is vital for biomedical applications.
The kinetic inertness is often characterized by comparisons
of acid-catalyzed dissociation rates of complexes. Therefore,
the rates of dissociation of Gd(DO2A2P) to free Gd3+ in
0.1–1.0 M HCl and constant ionic strength ([HCl] + [KCl]
) 1.0 M) were measured by water relaxation. Release of
free Gd3+ from the complex in acid solution resulted in an
increase of the relaxivity (r1) since r1(Gd3+) ) 17.0 mM-1
s-1 while r1(Gd(HxL)) ≈ 7 mM-1 s-1 at 9 MHz. Equlibrium
calculations using the known protonation constants, KGdHL
and KGdH2L (Table 2), indicate that the complex is largely in
the diprotonated form, Gd(H2DO2A2P) (GdH2L), under these
experimental conditions. This species likely has two protons
associated with the two appended and coordinated phospho-
nate groups. The proton from a phosphonate group can be
transferred to a ring nitrogen, which eventually leads to the
dissociation of the complex. The transfer of the proton is
very slow because of the electrostatic repulsion between the
coordinated Ln3+ ion and the proton.
At a given [H+] the rate of dissociation is directly
proportional to the total concentration of the complex, [GdL]t
[GdL]
-
t ) kd[GdL]t
(8)
dt
where kd is a pseudofirst-order rate constant, since [H+] >>
[GdL]t. The kd values were calculated from the relaxivity
measurements and a plot of kd as a function of [H+] showed
weak saturation at high acid concentrations (Figure 6).
Dissociation of GdH2L was assumed to take place both
spontaneously and by a proton assisted mechanism through
A comparison of the kOH values characterizing the
formation of the DO2A2P, DOTA, DO3AP, and DOTP
complexes of Ce3+ and Gd3+ (Table 3) clearly indicates
that the gradual replacement of the carboxylate groups of
DOTA with phosphonate groups results in a decreased
rate of complex formation. For the interpretation of these
3858 Inorganic Chemistry, Vol. 47, No. 9, 2008

DO2A2P and its Complexes with Ca2+, Cu2+, Zn2+, and Ln(III)
by several groups.16,17,51–54 These were found to be quite
similar both in solid state and in solution, probably because
of the rigid structure of the coordinated DOTA or DOTP
ligands. The four ring nitrogens and four carboxylate or
phosphonate oxygen donor atoms form two parallel planes
and the Ln3+ is located between the planes, somewhat closer
to the oxygens. The four N and four O donor atoms
approximately form a square antiprism. In the case of the
Ln(DOTA) complexes, a water molecule is coordinated to
the Ln3+ ion in a capping position at the center of plane of
the four oxygens. The Ln(DOTP) complexes do not contain
water molecule in the inner sphere of the Ln3+ ion.17,18
Because of the rigid structure of the macrocyclic ring, the
four ethylene groups adopt identical staggered conformations,
which lead to clockwise (λλλλ) or counterclockwise (δδδδ)
helicity of the ring. The acetate groups are accommodated
similarly and thus two enantiomers, Λ(λλλλ) and ∆(δδδδ),
are formed which give rise to identical NMR spectra.
However, the 1H and 13C NMR spectra of several Ln(DOTA)
complexes show two sets of signals which have been shown
to correspond to two different coordination isomers where
the conformation of the macrocyclic ring remains the same
but the layout of the acetate sidearms differs.47–50 For
the enantiomer pair Λ(λλλλ) and ∆(δδδδ), the twist angle
between the 4N and 4O planes is about 40° corresponding
to a square antiprismatic geometry (SAP). Twisting the
acetate arms in the opposite direction gives rise to the
enantiomer pair Λ(δδδδ) and ∆(λλλλ) having a twist angle
of about -30° corresponding to a twisted square antipris-
matic geometry (TSAP).47–50 Ln(DOTP) complexes exist as
the TSAP isomer both in solid state and in solution.17 Both
enantiomerization and isomerization of DOTA complexes
have been observed by NMR in the temperature range 0 –
100 °C while only enantiomerization was detected in case
of the more rigid DOTP complexes.
Figure 6. First-order rate constants characterizing the dissociation of
Gd(DO2A2P) ([Gd(DO2A2P)] ) 1 × 10-3 M, [HCl] + [KCl] ) 1.0 M,
25 °C).
formation of a triprotonated species, Gd(H3L), characterized
by the rate constants k0 and k1, respectively. The rate of
dissociation can be expressed by eq 9:
d[GdL]t
-
) k0[GdH2L] + k1[GdH2L][H+]
(9)
dt
Considering that [GdL]t ) [GdH2L] + [GdH3L] and the
protonation constant KGdH3L ) [GdH3L]/[GdH2L][H+], kd can
be expressed from eqs 8 and 9 as follows:
k0 + k1[H+]
1 + KGdH3L[H+]
kd )
(10)
The rate constants k0 and k1 were obtained by fitting the kd
values to eq 10. These are presented in Table 4 together with
the reported values for the Gd(DOTA), Gd(DO3AP), and
Gd(DOTP).
From a comparison of these data, it can be concluded that
replacement of acetate sidearms of DOTA with methylene-
phosphonates leads to an increase in the proton-assisted
dissociation rate of the Gd3+ complexes. This decrease in
kinetic inertness likely reflects the mechanism of dissociation.
The first step is protonation of an oxygen atom of the
coordinated ligand followed by the protonation of a ring
nitrogen via an intramolecular proton transfer from a
protonated carboxylate or phosphonate. Since it is much
easier to protonate a coordinated phosphonate (protonation
occurs at pH 5–7) than a carboxylate, the Gd3+ complexes
containing more phosphonate groups form more stable
protonated complexes and therefore transfer of a proton to
a ring nitrogen occurs with greater frequency.
1H and 13C NMR Studies of the Ln(DO2A2P) Com-
plexes. The structure of the Ln(DO2A2P) complexes, like
many other properties, is expected to be similar to those of
the corresponding Ln(DOTA) and Ln(DOTP) complexes.
The structure of the Ln(DOTA) complexes has been studied
in detail both in the solid state by X-ray diffraction method
and in solution by 1H and 13C NMR spectroscopy.41–50 The
structure of different Ln(DOTP) complexes was also studied
Ln(DOTA) and Ln(DOTP) complexes have C4 symmetry
1
and so the H and 13C NMR spectra consist of 6 and 4
signals, respectively.47 The Ln(DO2A2P) complexes how-
(41) Spirlet, M. R.; Rebizant, J.; Loncin, M. F.; Desreux, J. F. Inorg. Chem.
1984, 23, 4278–83.
(42) Dubost, J. P.; Leger, J. M.; Langlois, M. H.; Meyer, D.; Schaefer, M.
C. R. Acad Sci. (Paris) 1991, 312, 349–54.
(43) Chang, C. A.; Francesconi, L. C.; Malley, M. F.; Kumar, K.;
Gougoutas, J. Z.; Tweedle, M. F.; Lee, D. W.; Wilson, L. J. Inorg.
Chem. 1993, 32, 3501–8.
(44) Parker, D.; Pulukkody, K.; Smith, F. C.; Batsanov, A.; Howard, J. A. K.
J. Chem. Soc., Dalton Trans. 1994, 689–93.
(45) Aime, S.; Barge, A.; Botta, M.; Fasano, M.; Ayala, J. D.; Bombieri,
G. Inorg. Chim. Acta 1996, 246, 423–429.
(46) Desreux, J. F. Inorg. Chem. 1980, 19, 1319–24.
(47) Aime, S.; Botta, M.; Ermondi, G. Inorg. Chem. 1992, 31, 4291–9.
(48) Hoeft, S.; Roth, K. Chem. Ber. 1993, 126, 869–73.
(49) Jacques, V.; Desreux, F. Inorg. Chem. 1994, 33, 4048.
(50) Aime, S.; Botta, M.; Fasano, M.; Marques, M. P. M.; Geraldes,
C. F. G. C.; Pubanz, D.; Merbach, A. E. Inorg. Chem. 1997, 36, 2059–
2068.
(51) Geraldes, C. F. G. C.; Brown, R. D., III; Cacheris, W. P.; Koenig, S. H.;
Sherry, A. D.; Spiller, M. Magn. Reson. Med. 1989, 9, 94–104.
(52) Aime, S.; Botta, M.; Parker, D.; Williams, J. A. G. J. Chem. Soc.,
Dalton Trans. 1996, 17–23.
(53) Sherry, A. D.; Geraldes, C. F. G. C.; Cacheris, W. P. Inorg. Chim.
Acta 1987, 139, 137–9.
(54) Ren, J.; Sherry, A. D. J. Magn. Reson. 1996, B111, 178–82.
(55) Aime, S.; Botta, M.; Garino, E.; Crich, S. G.; Giovenzana, G.; Pagliarin,
R.; Palmisano, G.; Sisti, M. Chem.sEur. J. 2000, 6, 2609–2617.
(40) Wang, X.; Jin, T.; Comblin, V.; Lopez-Mut, A.; Merciny, E.; Desreux,
J. F. Inorg. Chem. 1992, 31, 1095.
Inorganic Chemistry, Vol. 47, No. 9, 2008 3859

Kálmán et al.
Table 4. Rate Constants, Characterizing the Dissociation of the Gd3+ Complexes of DOTA, DO3AP, DO2A2P, and DOTP (25 °C)
Gd(DOTA)a
Gd(DO3AP)b
Gd(DO2A2P)
Gd(DOTP)c
k0 (s-1
k1 (M-1 s-1
)
<5 × 10-8
(4.3 ( 0.5) × 10-6
)
8.4 × 10-6
2.78 × 10-3
(1.95 ( 0.04) × 10-4
5.4 × 10-4
a Reference 40. b Reference 24. c Reference 14.
1
ever have only C2 symmetry so twelve H resonances and
chemical exchange between the two isomers was also
confirmed by saturation transfer experiments (Figure S3,
Supporting Information). The assignment of the NMR spectra
of Eu(DO2A2P) was aided by 2D NMR methods (COSY,
EXSY, and HMQC). The HMQC spectrum is presented in
the Supporting Information (Figure S4).
seven 13C NMR resonances were expected and observed
experimentally (Figure 7). In the H NMR spectrum, the
signals of the two -CH2-PO3 protons overlap at -10.4
ppm and the coupling patterns could not be observed due to
the paramagnetism of Eu3+. In the 13C NMR spectrum,
however, the signal of the -CH2-PO32- carbon is split into
a doublet due to coupling to the 31P nucleus (JCP ) 148.8
Hz). The axial (a) and equatorial (e) protons of the ethylene
groups are labeled as Ha, He, Ha, He, Ha, He, Ha, and
1
2-
The solution structure of Lu(DO2A2P) was examined by
1D (1H and 13C) and various 2D NMR methods (COSY,
EXSY, heteronuclear chemical shift correlation (HETCOR),
1
Figures S5–S7, Supporting Information). The H and 13C
1
1
2
2
3
3
4
4He according to the numbering of carbon atoms (Figure 1).
The 1H NMR spectrum of Eu(DO2A2P) recorded at 274 K
showed two sets of signals indicating the presence of two
isomers at lower temperatures. The chemical shifts of the
major component are similar to those of the Eu(DOTP) and
the m isomer (TSAP) of Eu(DOTA) while the minor
component appears to have chemical shifts similar to those
of the M isomer (SAP) of Eu(DOTA). From the comparison
of the integrals of the two sets of signals, it was estimated
that under these conditions Eu(DO2A2P) exists 93% as the
NMR spectra (Figures 8 and 9) recorded at different
temperatures revealed the presence of only the TSAP isomer.
It is known from the H NMR studies of Lu(DOTA) and
1
Lu(DOTP) complexes that the ring axial protons are coupled
with the geminal and vicinal axial protons, while the
equatorial protons couple with the geminal protons only. Due
to these couplings the axial and equatorial protons give rise
to a triplet and doublet, respectively,47 and similar couplings
and multiplicities can be expected for Lu(DO2A2P). The
1Ha-4He, He-4Ha, Ha-3He, and He-3Ha proton pairs of
1
2
2
1
TSAP isomer and 7% as the SAP isomer. The H NMR
spectra were also recorded at 6 different temperatures from
274 to 303 K. The signals of the two isomers broaden and
move closer to one other and finally the two sets of signals
coalesced at 298 K (Figure S2, Supporting Information). The
1
Figure 8. H NMR spectra of Lu(DO2A2P) at 274 K (A) and 298 K (B)
([LuL] ) 0.08 M).
Figure 7. The 1H and 13C NMR spectra of the Eu(DO2A2P) complex.
(1H NMR spectra at 298 K (A) and 274 K (B), 13C NMR spectra at 298
K(C), [EuL] ) 0.04 M).
Figure 9. 13C NMR spectra of Lu(DO2A2P) at 278 K (A), 288 K (B), and
298 K (C) ([LuL] ) 0.08 M).
3860 Inorganic Chemistry, Vol. 47, No. 9, 2008

DO2A2P and its Complexes with Ca2+, Cu2+, Zn2+, and Ln(III)
the complex exchange through fluxional motion as confirmed
by the EXSY cross peaks (Figure S6, Supporting Information).
13C NMR studies of Lu(DOTP) have shown that the ring
carbon nuclei positioned above the plane defined by the four
N atoms (e.g., 1C and 3C) are strongly coupled to the 31P of
the phosphonate group while coupling between the carbons
located under the plane (e.g., 2C and 4C) and the 31P atoms
are quite weak.16 In the 13C NMR spectrum of Lu(DO2A2P),
the doublet at 51.6 ppm (at 278 K) is assigned to the 3C
atoms, which are coupled to the 31P nuclei (3JCP ) 15.47
Hz). At higher temperature the rate of the interconversion
of the macrocyclic ring and the chemical exchange between
the 3C and 2C (and between the 1C and 4C) carbons is faster.
At 298 K, a broad signal appears (the signal of the 2C
carbons is at 51.96 ppm). The chemical exchange process
between the 1C (48.77 ppm) and 4C carbons (53.66 ppm)
could directly be detected by magnetization transfer experi-
ments (Figure S8, Supporting Information). The lower
intensity signals at 53.66 and 54.66 ppm are assigned to the
1
Figure 10. Experimental (A) and the simulated (B) H NMR spectra of
the Lu(DO2A2P).
Table 5. Rate Constants ki (s-1) Characterizing the Interconversion of
the Ring of Lu(DO2A2P)
T (K)
ki (s-1
276
3.9
282
9.0
288
13
294
19
300
36
306
63
)
-CH2-PO3 carbons, which are coupled to the 31P atoms
2-
(∆Hq ) 101 kJ mol-1) obtained for La(DOTP).16,46 These
data indicate that the extremely high rigidity of the macro-
cyclic ring in the Ln(DOTP) complexes significantly de-
creases if two of the phosphonates of DOTP are replaced
with carboxylate groups.
(1JCP ) 138.21 Hz). The two doublet signals at about 3.7
and 2.7 ppm in the 1H NMR spectrum belong to the acetate
methylene protons (2JHH ) 15.95 Hz). The protons of the
-CH2-PO32- groups appear as two double-doublets at about
2.93 and 2.17 ppm (2JHP ) 13.85 and 16.7 Hz, respectively).
3
The coupling constants JHH, characterizing the coupling
Conclusion
between the vicinal pairs of protons Ha–2Ha and Ha–4Ha
1
3
were found to be about 13–14 Hz. The dihedral angle φ
between these vicinal protons was estimated at ∼180° from
the Karplus equation (3JHH ) 7 - cos φ + 5 cos2 φ).16 The
1H and 13C NMR signals of Lu(DO2A2P) broaden at higher
temperatures as a result of an increased rate of enantiomeriza-
tion. The rates of enantiomerization of the Lu(DOTA) and
La(DOTP) complexes have been studied by NMR spectroscopy
and higher activation parameters were found for the La(DOTP),
which was interpreted with its more rigid structure.16,46
In the 1H NMR spectra of Lu(DO2A2P), the 4Ha and 4He
proton signals are well separated, so it was possible to carry
out line-shape analysis by simulating the spectra gathered
at different temperatures. Since the linewiths practically do
not change below 272 K, the transverse relaxation time for
the protons was calculated from the spectra obtained at 272
K (T2 ) 0.07 s). The experimental spectra were simulated
using coupling constants, 2JHH ) 13.45 Hz and 3JHH ) 13.6
Hz. Examples of typical experimental and simulated spectra
are shown in Figure 10. The rate constants characterizing
the interconversion (ki ) 1/τ) of the macrocyclic ring were
calculated from the average conformation life-times (τ)
(Table 5) and the activation parameters characterizing the
enantiomerization process were estimated using the Eyring
equation as ∆Hq ) 60.2 ( 3.6 kJ mol-1 and ∆Sq ) -14 (
12 kJ mol-1 K-1. The interconversion rate constant (ki) at
298 K was also calculated from the selective 13C magnetiza-
tion transfer experiments and the ki ) 34 ( 3 s-1 value agrees
well with the data obtained from the line shape analysis
(Table 5). The activation enthalpy and enthropy are similar
to those reported for the La(DOTA) (∆Hq ) 59.4 kJ mol-1
and ∆Sq ) -4.6 kJ mol-1 K-1) but lower than the values
The cyclen based ligand, DO2A2P, was synthesized in
reasonable yields. This ligand has two acetate and two
methylenephosphonate pendant arms and, as expected, its
protonation and chelation behavior was similar to those
of DOTA and DOTP. The protonation of DO2A2P,
DOTA, and DOTP is characterized by two high constants
due to the protonation of two ring nitrogens. The stability
constants of the DO2A2P complexes fall between those
of the corresponding DOTA and DOTP complexes. The
relaxivity of Gd(DO2A2P) decreases from about 4.6 to
about 3.6 mM-1 s-1 between pH 2 and 7, then remains
constant over the pH range 7–12. Similar to Gd(DOTP),
17O NMR measurements indicate the absence of an inner-
sphere water molecule. Kinetic studies on the Ce3+ and
Gd3+ complexes revealed that the rate determining step
in the formation of these complexes corresponds to
rearrangement of
a
diprotonated intermediate,
Ln(H2DO2A2P)*, which occurs both spontaneously and
by a OH– assisted process. The proton assisted dissociation
of Gd(DO2A2P) occurs through the diprotonated species
GdH2L both spontaneously and by a proton assisted
mechanism through formation of a triprotonated species,
Gd(H3L). The order of proton assisted dissociation rates
of the Gd3+ complexes is DOTP > DO2A2P > DOTA,
which reflects the relative stability of the corresponding
protonated complexes. NMR data suggest that Eu(DO2A2P)
exists as 93% twisted square antiprismatic (TSAP) and
7% square antiprismatic (SAP) isomer. Activation enthalpy
and entropy values for the enantiomerization process by
ring interconversion suggest that Ln(DO2A2P) complexes
are less rigid than the corresponding DOTP complexes
Inorganic Chemistry, Vol. 47, No. 9, 2008 3861

Kálmán et al.
Supporting Information Available: Additional figures (17O
NMR line width data, variable temperature H NMR studies of
EuDO2A2P, 1H NMR saturation transfer experiments with
EuDO2A2P, HMQC spectrum of EuDO2A2P, COSY, EXSY, and
HETCOR spectrum of LuDO2A2P, and magnetization transfer on
the 13C NMR spectra of LuDO2A2P). This material is available
free of charge via the Internet at http://pubs.acs.org.
and that the rigidity of the complexes can be fine-tuned
by varying the number of methylenephosponate and
acetate side arms on the ligand.
1
Acknowledgment. This research was supported in part
by grants from the National Institutes of Health (CA-115531
and RR-02584, A.D.S.), the Robert A. Welch Foundation
(AT-584, A.D.S.), OTKA K 69098 (E.B.), COST Action
D-38 program (E.B.), and the EMIL Project funded by the
EC FP6 Framework program (E.B.).
IC7024704
3862 Inorganic Chemistry, Vol. 47, No. 9, 2008