Lanthanide(III) complexes of phosphorus acid analogues of H4DOTA as model compounds for the evaluation of the second-sphere hydration

FULL PAPER

DOI: 10.1002/ejic.200800859

Lanthanide(III) Complexes of Phosphorus Acid Analogues of H₄DOTA as

Model Compounds for the Evaluation of the Second-Sphere Hydration

Zuzana Kotková,^[a]Giovannia A. Pereira,^[b]Kristina Djanashvili,^[c]Jan Kotek,^[a]

Jakub Rudovský,^[a]Petr Hermann,^[a]Luce Vander Elst,^[d]Robert N. Muller,^[d]

[b]

[a]

[c]

ˇ

Carlos F. G. C. Geraldes, Ivan Lukes,* and Joop A. Peters*

Keywords: Contrast agents / Lanthanides / NMR spectroscopy / Phosphinate complexes / Phosphonate complexes

Five DOTA-like ligands lacking a water molecule in the first

coordination sphere of their Gd^IIIcomplexes, namely the

phosphinates H₄DOTP^H, H₄DOTP^hmand H₄DOTP^Et, and the

phosphonate monoesters H₄DOTP^OEtand H₄DOTP^OBu, were

synthesized with the aim of exploring the influence of the

second hydration sphere on the relaxivity of Gd^IIIcomplexes.

The H₄DOTP^H, H₄DOTP^hmand H₄DOTP^OEtligands and their

Ln^IIIcomplexes were characterized by potentiometry and

time-resolved luminescence; the Gd^IIIcomplexes are thermo-

dynamically much less stable than that of H₄DOTA, and no

water is coordinated in the inner sphere. The crystal struc-

plexes show that these compounds have a clear preference

for a specific arrangement of phosphorus atoms which gives

rise to the symmetrical RRRR (or SSSS) isomer. However,

many diastereoisomers were observed for all other com-

plexes. Ln^III-induced ¹⁷O NMR shift data reveal that the spa-

tial location of the second-sphere water molecules for the two

groups of complexes differs. The parameters governing the

effect of the second hydration sphere on the relaxivity of the

Gd^IIIcomplexes of all ligands were evaluated by EPR, vari-

able-temperature ¹⁷O NMR spectroscopy and ¹H NMRD re-

laxometry. The presence of second-sphere water molecules is

tures of the free ligand H₄DOTP^OEtand of the Gd^IIIcom- clearly confirmed, depending on the character of the pendant

plexes of H₄DOTP^Hand H₄DOTP^OEtwere determined by X-

arms. As the relaxivity does not depend significantly on the

ray diffraction. The complexes have the expected octadent- nature of the phosphorus substituents and/or on the isomer-

ate coordination mode with an N₄O₄arrangement; no water

molecule is bound to the Gd^IIIion. Information on the struc-

tures of the Ln^IIIcomplexes of all five ligands in aqueous

solution were obtained from ¹H and ³¹P NMR spectra. The

NMR spectra of the [Ln(DOTP^hm)]^–and [Ln(DOTP^Et)]^–com-

ism present in solution, the second-sphere water molecules

should be located close to the phosphorus–oxygen atoms.

Germany, 2009)

more than a decade.^[1–5]The research undertaken has led to

a much better understanding of the parameters governing

the effectiveness of these CAs, which is usually expressed

by the relaxivity (the enhancement of the water proton re-

laxation rate in a 1 m solution of CA).^[1–6]The structure

of the ligand can be tuned to optimize some of the param-

eters for optimal contrast efficiency, including the rotational

correlation time (τ_R), the number of water molecules di-

rectly bound to the central metal ion (q) and their residence

time in the first coordination sphere (τ_M). These parameters

mainly influence the so-called inner-sphere contribution to

the overall relaxivity. Other parameters, such as the elec-

tronic relaxation rates (T_1e,2e) of the Gd^IIIion^[1,2,7,8]or the

contribution of water molecule(s) in the second hydration

sphere,^[9]are less well understood.

Introduction

Magnetic resonance imaging (MRI) is one of the most

widely used medical diagnostic techniques. Gadolini-

um(III)-based contrast agents (CAs) have significantly ex-

tended the applicability of this imaging method, and opti-

mizing the efficacy of CAs has been an important goal for

[a] Department of Inorganic Chemistry, Univerzita Karlova

(Charles University),

Hlavova 2030, 12840 Prague, Czech Republic

Fax: +420-22195-1253

E-mail: lukes@natur.cuni.cz

[b] Departamento de Bioquímica, Faculdade de Ciências e Tecnol-

ogica, e Centro de Neurociências, Universidade de Coimbra,

3049 Coimbra, Portugal

[c] Laboratory for Biocatalysis and Organic Chemistry, Depart-

ment of Biotechnology, Delft University of Technology,

Julianalaan 136, 2628 BL Delft, The Netherlands

Fax: +31-152-784-289

The second-sphere contribution to the relaxivity origi-

nates from water molecules held in the close proximity to

the Gd^IIIion by hydrogen bonds to the organic ligand of

the complex.^[9]This contribution can enhance the overall

relaxivity by 5–15% depending on the structure of the com-

plex. Some years ago, it was demonstrated that complexes

E-mail: j.a.peters@tudelft.nl

[d] NMR Laboratory, Department of Organic Chemistry, Univer-

sity of Mons-Hainaut,

7000 Mons, Belgium

Supporting information for this article is available on the

WWW under http://ww w .eurjic.org or from the author.

Eur. J. Inorg. Chem. 2009, 119–136

119

I. Lukesˇ, J. A. Peters et al.

FULL PAPER

of phosphorus-containing ligands exhibit a higher hy- tain more bulky phosphorus atoms.^{[10,15,16,18,19]}The pres-

dration (i.e. more extended second hydration sphere) than ence of just one phosphorus-based pendant arm causes an

complexes of carboxylate ligands. This is particularly true increased population of the TSA isomer.^[20–24]As stated

for lanthanide(III) complexes of H₈DOTP [H₈DOTP = above, the Gd^IIIcomplexes of all the tetraphosphorus li-

1,4,7,10-tetraazacyclododecane-1,4,7,10-tetrakis(methyl- gands under study have no coordinated water molecules. In

phosphonic acid), Scheme 1], whose Gd^IIIcomplex does fact, direct coordination of a water molecule was observed,

not contain any directly coordinated water molecule^[10]but in the solid state, only for the La^IIIand Ce^IIIcomplex-

has a relaxivity (r₁) comparable to that of the clinically used es.^[15,16b]

[Gd(H₂O)(DOTA)]^–complex (H₄DOTA = 1,4,7,10-tetra-

A rich, second-sphere hydration has also been observed

azacyclododecane-1,4,7,10-tetraacetic acid, Scheme 1), in Gd^IIIcomplexes of pyridine-containing macrocycles con-

which has one water molecule in the first coordination taining phosphonic acid pendant arms.^[25]It has been

sphere.^[11]The water relaxation enhancement by the DOTP shown recently that the second-sphere contribution to the

complex results exclusively from the second-sphere contri- overall relaxivity is significant even in complexes of

bution after conjugation to a macromolecule.^[12]A previous H₅DTPA (H₅DTPA = 1,4,7-triazaheptane-1,1,4,7,7-penta-

relaxometric study concerning Gd^IIIcomplexes of the re- acetic acid, Scheme 1)^[26]or H₄DOTA derivatives contain-

lated ethylphosphinate ligand (H₄DOTP^Et), and of the eth- ing only one phosphonic/phosphinic acid group.^[20–22]This

yl- and butylphosphonate monoesters (H₄DOTP^OEtand contribution was found to be more pronounced when com-

H₄DOTP^OBu, respectively), indicated only poor hydration plexes of these monophosphorus acid ligands were bound

(coordination number lower than one).^[13]The absence of to macromolecules.^[27,28]Theoretical approaches to treat the

directly coordinated water molecules and the presence of second-sphere hydration have confirmed that phosphonic

second-sphere water molecule(s) has been demonstrated for acid groups enhance the hydration of complexes of DOTA-

Gd^IIIcomplexes of phosphinic acid analogues of H₄DOTA like ligands^[29]as well as of open-chain ligands.^[30]In this

having methyl (H₄DOTP^Me),^[14]phenyl (H₄DOTP^{Ph [15]}

or paper, we evaluate the second-sphere contribution to the

)

benzyl (H₄DOTP^{Bn [16]}

atom (Scheme 1).

)

substituents on the phosphorus overall relaxivity in a series of Ln^IIIcomplexes of tetraphos-

phorus acid analogues of H₄DOTA with phosphinic

(H₄DOTP^H), hydroxymethylphosphinic (H₄DOTP^hm), eth-

ylphosphinic (H₄DOTP^Et) and phosphonate ethyl mono-

ester (H₄DOTP^OEt) and butyl monoester (H₄DOTP^OBu

)

pendant arms (Scheme 1) by means of a combination of

relaxometric and other spectroscopic methods. In addition,

the influence of the substituent on the phosphorus atom to

this contribution is investigated.

Results and Discussion

Syntheses

H₄DOTP^Hwas prepared by a Mannich reaction between

cyclen, paraformaldehyde and hypophosphorous acid ac-

cording to a slightly modified literature procedure;^[31]

milder conditions were applied, for this reaction and during

the workup, to reduce the amount of by-products, including

Scheme 1. Formulas of ligands mentioned in the text.

The structures of complexes of these tetraphosphorus derivatives with various numbers of pendant arms with hy-

acid derivatives are analogous to those of complexes of the droxymethyl and hydrogen substituents (e.g. intermediates

parent ligand H₄DOTA, where the lanthanide(III) ion is in H₄DOTP^hmformation) and partially N-methylated deriv-

sandwiched between N₄- and O₄-planes. The H₄DOTA atives of cyclen having a smaller number of phosphinate

complexes are present in solution in two diastereoisomeric pendants. Despite the optimized reaction conditions, the

forms that differ in the mutual rotation of N₄- and O₄-pla- amount of by-products was still high, and the yield of iso-

nes to give a square-antiprismatic (SA) arrangement [tor- lated product was only 20%. H₄DOTP^hmwas obtained as

sion angle Ͼ35°, with opposite signs of rotation of the pen- a by-product during the synthesis of H₄DOTP^H, but could

dant arms (∆/Λ) and the conformation of the ethylene brid- also be prepared more conveniently from pure H₄DOTP^H

ges in the macrocycle ring (δ/λ), represented as the ∆-λλλλ/ upon treatment with an excess of paraformaldehyde. In this

Λ-δδδδ isomeric pair, traditionally termed “M”] and a way, less by-products were present in the final reaction mix-

twisted-square-antiprismatic (TSA) arrangement (torsion ture and the isolated yield of H₄DOTP^hmwas 50%.

angle Ͻ30°, with the same sign of rotation leading to Λ- H₄DOTP^OEtwas prepared by treating cyclen with para-

λλλλ/∆-δδδδ antipodes, traditionally termed “m”).^[17]Com- formaldehyde and triethyl phosphite (without solvent) at

plexes of H₄DOTP^Rligands, on the other hand, occur ex- 40 °C.^[32]The octaethyl ester Et₈DOTP obtained was hy-

clusively in a TSA arrangement as their pendant arms con- drolyzed in aqueous NaOH to give the desired product in

120

www.eurjic.org

Eur. J. Inorg. Chem. 2009, 119–136

Lanthanide(III) Complexes of Phosphorus Acid Analogues of H₄DOTA

an overall yield of 90%. H₄DOTP^Etand H₄DOTP^OBuwere trogen atoms of the macrocycle and four oxygen atoms of

synthesized according to literature procedures.^[13,33]All li- the pendant arms to form two parallel N₄- and O₄-planes.

gands were purified, as mentioned in the Experimental Sec- As the complex possesses the symmetry of a crystallo-

tion, to a purity of at least 99%. The lanthanide(III) com- graphic twofold axis, only half of the molecule forms the

plexes were formed at a pH around 7 and at room tempera- independent unit. The sign of rotation of the pendant arms

ture in 1 h (confirmed by ³¹P NMR spectroscopy), which (Λ/∆) is the same as that of the conformation of the ethyl-

demonstrates that the complexation rates are rapid.

ene bridges in the macrocycle ring (δ/λ). This leads to a

TSA (Λ-λλλλ/∆-δδδδ) environment and confirms the re-

sults found by NMR spectroscopy (see below). The torsion

angle between the N₄- and O₄-planes is about 30°, which is

typical for a TSA coordination sphere [an SA isomer (∆-

λλλλ/Λ-δδδδ) would typically have a torsion angle of more

than 35°].^[23,24]The O22 oxygen atom of one pendant arm

was best fitted as disordered in two positions with relative

occupancies 60(A):40(B), with the phosphinate hydrogen

atom placed in the theoretical position. This leads to two

diastereoisomers (each of which is present in both enantio-

meric forms as the space group is centrosymmetric), namely

SSSS-∆-δδδδ (+RRRR-Λ-λλλλ) and SRSR-∆-δδδδ

(+SRSR-Λ-λλλλ). The opening angles O–Gd–O are

around 125°, which is too small for the coordination of a

water molecule.^[37]

Crystal Structures

Crystal Structure of H₄DOTP^OEt·H₂O

The structurally independent unit in the crystal structure

of H₄DOTP^OEt·H₂O contains two ligand molecules, with

similar conformations, and two solvate water molecules (see

Figure 1 and Figure S1 in the Supporting Information).

The free ligand H₄DOTP^OEtcrystallizes as a zwitterion,

with two (mutually “trans”) macrocycle nitrogen atoms pro-

tonated. The two remaining protons are bound to the pen-

dant phosphonate groups attached to the unprotonated

amino groups in opposite positions. The macrocyclic part

adopts a conformation typical for doubly protonated sub-

stituted cyclens, with intramolecular hydrogen bonds be-

tween protonated and unprotonated amino groups and all

substituents pointing in the same direction with respect to

the macrocyclic plane.^[34]Both protonated pendant arms

are turned above the macrocycle (and participate in hydro-

gen bonding with protonated amino groups), and the re-

maining two pendants are turned outwards from the macro-

cycle cavity. The quality of the crystal data was rather

poor – some ethyl ester groups and the solvate water mole-

cules were found to be disordered. A similar orientation of

the pendant arms is also found in the structures of the par-

ent H₈DOTP^[35]and H₄DOTA.^[36]

Figure 2. Molecular structure of the [Gd(DOTP^H)]^–species found

in the crystal structure of Li[Gd(DOTP^H)]·6H₂O. Hydrogen atoms

attached to carbon atoms have been omitted for clarity. Both posi-

tions of the disordered phosphinate pendant arm, with SSSS-∆-

δδδδ (60%, O22A) and SRSR-∆-δδδδ (40%, O22B) configurations,

are shown.

Figure 1. Molecular structure of one independent ligand molecule

found in the crystal structure of H₄DOTP^OEt·H₂O, showing the

intramolecular hydrogen bonds (dashed). Hydrogen atoms attached

to carbon atoms have been omitted for clarity. Only one position

of the disordered parts of molecule (phosphonate ester groups) is

shown.

The molecules in this complex are packed in a polymeric

chain due to coordination through the lithium counterion.

This ion was also refined as being disordered over two posi-

tions as it is coordinated by the disordered oxygen atom

O22. The coordination sphere of the lithium ion is com-

pleted by two water molecules, which are located far from

Crystal Structure of Li[Gd(DOTP^H)]·6H₂O

The central Gd^IIIion in the crystal structure of the metal centre and the pseudo-C₄axis (the closest

Li[Gd(DOTP^H)]·6H₂O (Figure 2) is coordinated to four ni- Gd···Ow distance is 5.67 Å and the QN–QO–Ow angle is

Eur. J. Inorg. Chem. 2009, 119–136

www.eurjic.org

121

I. Lukesˇ, J. A. Peters et al.

FULL PAPER

Table 1. Selected geometrical parameters found in the crystal structures of Li[Gd(DOTP^H)]·6H₂O and Li[Gd(DOTP^OEt)]·

0.5Me₂CO·8H₂O.

Li[Gd(DOTP^H)]·6H₂O

Li[Gd(DOTP^OEt)]·0.5Me₂CO·8H₂O

Molecule 1

Molecule 2

Distances [Å]

Gd1–N1

Gd1–N4

Gd1–N7

Gd1–N10

Gd1–O11

Gd1–O21

Gd1–O31

Gd1–O41

2.636(2)

2.635(2)

–

2.663(4)

2.667(4)

2.662(5)

2.665(4)

2.315(4)

2.332(4)

2.323(4)

2.307(4)

1.6390(2)

1.0495(2)

2.693(5)

2.670(5)

2.649(5)

2.670(5)

2.320(3)

2.308(3)

2.315(3)

2.307(4)

1.6520(2)

1.0491(2)

–

2.311(2)

2.319(2)

–

[a]

Gd1–QN

Gd1–QO

1.6053(2)

1.0503(2)

[a]

Angles [°]

O11–Gd1–O31

O21–Gd1–O41

124.6(1)^[b]

124.7(1)

127.6(1)

26.7(2)

27.4(2)

27.0(2)

28.6(2)

1.09(2)

127.2(1)

124.9(1)

27.7(2)

26.1(2)

28.2(2)

26.3(2)

0.86(3)

127.4(1)^[b]

N1–QN–QO–O11^[a]

N4–QN–QO–O11^[a]

N7–QN–QO–O11^[a]

N10–QN–QO–O11^[a]

N₄-plane–O₄-plane

29.21(8)

30.50(8)

–

0

[a] QN is the centroid (centre of gravity) of the N₄-plane and QO the centroid of the O₄-plane. [b] O31/O11^#and O41/O21^#are related

by a twofold symmetry axis; ^#: –x, y, –z + 1/2.

around 98°, where QN is the centroid of the N₄-plane and phosphonate monoesters is alternating, leading to SRSR-

QO the centroid of the O₄-plane). Selected geometrical pa- ∆-δδδδ and SRSR-Λ-λλλλ species, as was observed in the

rameters are listed in Table 1.

case of H₄DOTP^Phcomplexes.^[15]One of the lithium coun-

terions is coordinated by O23X and by three water mole-

cules in an approximately tetrahedral arrangement. The

other lithium ion is tetrahedrally coordinated by four water

molecules. Similarly to the previous structure, the water

molecules are located away from the metal centre and the

pseudo-C₄axis of the complexes (the distances of the clos-

est water molecules from metal centre are in the range 5.37–

5.60 Å, with corresponding QN–QO–Ow angles of 97–

102°). Selected geometrical parameters are listed in Table 1.

Crystal Structure of Li[Gd(DOTP^OEt)]·0.5Me₂CO·8H₂O

The independent part of the crystal structure of

Li[Gd(DOTP^OEt)]·0.5Me₂CO·8H₂O (Figure 3) consists of

two complex molecules. Similarly to the previous case, the

central Gd^IIIions in both molecules are coordinated by four

nitrogen atoms of the macrocycle and four oxygen atoms of

the pendant arms to form two parallel N₄- and O₄-planes.

The stereochemistry of the complex molecules is also TSA

(Λ-λλλλ/∆-δδδδ), with torsion angles between the N₄- and

O₄-planes of about 28°. The absolute configuration of the

Potentiometry

Protonation Constants of the Ligands

The protonation constants of the free ligands are given

in Table 2. These ligands adopt the general protonation

scheme of DOTA-like compounds. The first two proton-

ations occur on macrocyclic nitrogen atoms (with pK_aval-

ues of about 10–11 and 7–8, respectively). The overall basic-

ity of the macrocyclic nitrogen atoms [pK(HL) + pK(H₂L)

= logβ(H₂L)] is rather low and corresponds well to the gen-

erally observed order phosphinates ≈ phosphonate mono-

esters Ͻ carboxylates Ͻ phosphonates (Table 2).^[37]The

proton dissociation constants of the H₂L^2–species of the

ligands studied, as well as those of H₄DOTP^Ph(ref.^[38]) and

H₄DOTP^Otfe(ref.^[39]), depend on the electronic properties

of the substituent on the phosphorus atoms, in agreement

with previous observations on simple aminomethylphos-

phinic acids – more electronegative substituents result in

lower pK_avalues.^[40]The protonations in the acidic region

Figure 3. Molecular structure of one (the independent molecule la-

belled X) of the [Gd(DOTP^OEt)]^–species found in the crystal struc-

ture of Li[Gd(DOTP^OEt)]·0.5Me₂CO·8H₂O. Hydrogen atoms have

been omitted for the sake of clarity. The picture shows the species

with SRSR-∆-δδδδ configuration.

122

www.eurjic.org

Eur. J. Inorg. Chem. 2009, 119–136

Lanthanide(III) Complexes of Phosphorus Acid Analogues of H₄DOTA

Table 2. Protonation constants^[a]and corresponding pK_avalues^[b]of the studied ligands.

h

Species^[c]

H₄DOTP^H[d]

H₄DOTP^hm[d]

H₄DOTP^Et[e]H₄DOTP^OEt[f]H₄DOTP^OBu[g]H₈DOTP^[h]H₄DOTA^[i]

logβ_h

pK_a

logβ_h

pK_a

1

2

3

4

5

6

7

HL

H₂L

H₃L

H₄L

H₅L

H₆L

H₇L

10.58(2)

17.51(3)

19.41(3)

10.58

10.60(2)

18.46(2)

20.21(3)

21.13(5)

10.60

7.86

1.75

0.92

–

10.94

11.57

7.94

1.80

–

2.3

–

10.34

Ͼ13

12.45

9.18

7.95

6.08

5.20

1.85

11.9

9.72

4.60

4.13

2.36

–

6.93

1.90

–

8.24

3.71

–

7.72

2.42

–

[j]

–

[a] β_h= [H_hL^(h–4)]/([H⁺]^hϫ[L^4–]). [b] pK_a= β_h– β_h–1. [c] Charges have been omitted for clarity. [d] This work. [e] Ref.^[41][f] Ref.^[32][g]

Ref.^[33][h] Ref.^[42][i] Ref.^[43][j] Simultaneous deprotonations over two steps.

(pK_aϽ 2–3) can be assigned to those of the phosphinate/ [Gd(HDOTP^R,OR)] species (formally the [hlm] = [111] spe-

phosphonate monoester moieties and/or the remaining cies) are relatively high at around 4. This suggests that the

macrocyclic amino groups.

proton in these species is probably bound to a nitrogen

atom in the macrocycle under the equilibrium conditions,

in other words the macrocycle is not fully coordinated, and

the Gd^IIIion is bound mainly by the oxygen atoms of the

pendant arms. Proton removal from the protonated species

leads to full coordination of the ligand in the normal N₄O₄

mode. The potentials do not change after prolonged stand-

ing of the samples – the same values were observed after

six weeks as after three weeks, which confirms that thermo-

dynamic equilibrium was reached in the solutions used. Al-

though these Ln^IIIcomplexes have a relatively low thermo-

dynamic stability, they have a high kinetic stability both in

vitro and in vivo.^[32]

Stability of the Gadolinium(III) Complexes

The stability constants of complexes with phosphinate

and phosphonate monoester ligands (Table 3) are much

lower than those of the parent ligands H₈DOTP and

H₄DOTA, which can be ascribed to the lower overall basic-

ity of the ligands. Consequently, the complexes are formed

at higher pH – the Gd^IIIion, for example, is only fully com-

plexed above a pH of about 5. The distribution diagram for

Gd^III-containing species in the Gd^III-H₄DOTP^OEtsystem

under equilibrium conditions is shown in Figure 4 as an

example. The proton dissociation constants of the

Protonation of the Gadolinium(III) Complexes

The relaxivity of [Gd(DOTP^hm)]^–appears to be depend-

ent on the pH (see below), whereas the relaxivity of the

other complexes is pH-independent. To gain more insight

into the observed behaviour, we performed a potentiometric

study of pre-formed Gd^IIIcomplexes of H₄DOTP^hmand,

for comparison, of H₄DOTP^OEt. During these measure-

ments, we explored the kinetically controlled pseudo-equi-

librium associated with the complex protonation as the

complexes are kinetically stable under these conditions.

Solutions of the Gd^IIIcomplexes of H₄DOTP^hmand

H₄DOTP^OEtwere prepared in sealed ampoules (see Experi-

mental Section) and used for the determination of their pro-

tonation/dissociation constants. The obtained constants are

given in Table 4, and the corresponding distribution dia-

gram is shown in Figure 4. The [Gd(HDOTP^R,OR)] com-

plexes behave as monovalent strong acids with a pK_aof

about 1.4. Such dissociation constants correspond well with

those of the pendant arms in the free ligands, although the

pK_avalues are somewhat lower in the complexes due to the

Figure 4. Distribution of Gd^III-containing species in the Gd^III

H₄DOTP^OEt(H₄DOTP^OEt= H₄L) system under equilibrium condi-

/

tions [c(Gd^III) = c(H₄L) = 0.004 ; full lines]. Distribution of Gd^III

-

containing species during titration of the pre-formed [Gd(L)]^–com-

plex (c{[Gd(L)]^–} = 0.004 ; dashed lines, formulae in italics). The

protonated [Gd(HL)] species probably bind the proton at a nitro-

gen atom under equilibrium conditions and an oxygen atom in the

pre-formed complex.

Table 3. Stability constants^[a]of gadolinium(III) complexes of the studied ligands.

h

l

m

H₄DOTP^H[b]

H₄DOTP^hm[b]

H₄DOTP^Et[c]H₄DOTP^OEt[b]H₄DOTP^OBu[d]

H₈DOTP^[e]

logβ_hlm

H₄DOTA^[f]

logβ_hlm

0

1

14.0(1)

18.11(5)^[g]

4.1

16.09(4)

19.25(5)^[g]

3.2

16.50

–

14.4(2)

19.0(1)^[g]

4.6

12.19

–

28.8

36.4^[g]

7.6

24.67

–

[a] β_hlm= [H_hL_lM_m^h–4l+3m]/([H⁺]^hϫ[L^4–]^lϫ[M³⁺]^m). [b] This work. [c] Ref.^[41][d] Ref.^[33][e] Ref.^[44]; the other pK_avalues are: 6.3, 5.4 and

4.0. [f] Ref.^[45][g] The corresponding pK_avalues are given in italics.

Eur. J. Inorg. Chem. 2009, 119–136

www.eurjic.org

123

I. Lukesˇ, J. A. Peters et al.

FULL PAPER

coordination of the pendant arms. Thus, the low pK_avalue complexes with one coordinated water molecule {e.g. [Eu-

points to the kinetic inertness of the complex species, at (H₂O)(DOTA)]^–, where the lifetimes observed in H₂O are

least on the time scale of the potentiometric measurements, about 650 µs}, but still significantly shorter than the life-

and that the complexes are protonated at an oxygen atom times measured in D₂O solution. For [Eu(DOTP^H)]^–and [Eu-

of the pendant arms. As no other protonations occur in (DOTP^hm)]^–complexes, the lifetimes observed in H₂O are

the pH range 3–10, the relaxivity change observed for the even closer to the values of [Eu(H₂O)(DOTA)]^–itself. How-

[Gd(DOTP^hm)]^–complex in this region (see below) cannot ever, using the relevant equations published in the litera-

be associated with the presence of any protonation/depro- ture,^[47,48]the hydration number, q, was estimated to be 0.3–

tonation event (e.g. dissociation of a proton from the 0.7 (Table S1 in the Supporting Information). This fact is

PCH₂OH groups); therefore, the relaxivity change must be usually attributed to a partial hydration of the complexes

associated with the H⁺/OH^–catalysis of the prototropic ex- (i.e. equilibrium between complex species with q = 0 and 1,

change.^[36,46]

respectively). Alternatively, a high second-sphere hydration

also gives such an effect.^[48]As the published equations used

to calculate q were derived mostly for amide derivatives,

which have a lower second-sphere hydration, the correction

for second-sphere contribution can be underestimated. In

our case, the richer hydration sphere of the phosphorus acid

complexes contributes more to the decay rates. All in all,

however, the luminescence data do not allow an unambigu-

ous conclusion regarding the value of q for the Eu^IIIcom-

plexes of the tetraphosphorus acid derivatives; therefore, we

recorded absorption spectra for the ⁵D₀ǟ⁷F₀transition.

Only a single symmetrical absorption peak was observed in

the spectra (Figure S2 in the Supporting Information).

Table 4. Equilibrium constants (pK_a) of the pre-formed Gd^IIIcom-

plexes with H₄DOTP^hmand H₄DOTP^OEt

.

Equilibrium

[Gd(DOTP^hm)]^–[Gd(DOTP^OEt)]^–

[HML] h H⁺+ [ML]^–

1.39(3)

11.29(1)

1.43(3)

12.01(2)

[ML]^–h H⁺+ [H_–1ML]^2–

Interestingly, the fits of the potentiometric data improved

dramatically after inclusion of deprotonated species (pK_aof

around 12) in the chemical model. Since deprotonated spe-

cies were observed for the complexes of both H₄DOTP^hm

and H₄DOTP^OEt, this improvement cannot be explained by

dissociation of a hydroxymethyl moiety of the former com-

plex. As no water is coordinated in the inner sphere of the

complexes (see below), the deprotonation can only be ex-

plained by coordination of a hydroxide anion after decoor-

dination of one of the pendant arms. Such behaviour points

to a relatively weak coordination of the phosphinic acid or

phosphonic monoester functions, probably due to the low

nucleophilicity of their oxygen atoms^[37]and the high affin-

ity of lanthanide(III) ions for the hydroxide anion.

5

Since the D₀ǟ⁷F₀transition is extremely sensitive to the

local coordination environment of the central Eu^IIIion, a

change in hydration [i.e. change in coordination number

(CN) between 8 and 9] would have resulted in a substantial

difference in the corresponding two spectral bands with a

separation of the peaks typically larger than 0.5 nm.^[30,49]

We can therefore conclude that only non-hydrated complex

species (i.e. having CN = 8) are present in the solutions; the

luminescence lifetimes given above (i.e. slight shortening of

From the distribution diagram (Figure 4) and published

data on other lanthanide(III) complexes,^[32]it is clear that

the pre-formed [Gd(DOTP^R,OR)]^–complexes are kinetically

relatively inert on the potentiometry time scale (ca. 20 min).

Above a pH of about 4, the concentration of the

[Gd(HDOTP^R,OR)] species is negligible. Since the most

likely mechanism of dissociation of the complex involves

protonation of the pendant arm followed by proton transfer

to the nitrogen atoms and subsequent dissociation of the

complex, it may be concluded that the complexes are stable

τ_{H O}in comparison with τ_{D O}) should be significantly

2

shorter for any hydrated comp²lex. The non-zero hydration

numbers (q = 0.4–0.6) calculated from the luminescence

lifetimes are therefore due to the effect of the second-sphere

water molecules and/or of neighbouring OH oscillators in

the case of the [Eu(DOTP^hm)]^–complex.^[48]

NMR Studies of the Complexes

1

under the experimental conditions applied for the investiga- ³¹P and H NMR Spectra

1

tion of the pH dependence of the H NMRD profiles (see

¹H and ³¹P NMR spectra of the complexes with Nd^III

,

Eu^IIIand Yb^IIIwere recorded in order to obtain infor-

mation about the solution structure of the complexes. It is

well known that all lanthanide(III) complexes of tetraphos-

phonate/phosphinate analogues of H₄DOTA selectively

adopt TSA-type (Λ-λλλλ/∆-δδδδ) structures.^{[10,15,16,18,19]}

below).

Luminescence and UV/Vis Measurements of Eu^III

Complexes

To confirm the absence of a directly coordinated water However, since the presently studied ligands have a substit-

molecule in the Gd^IIIcomplexes studied, we measured lumi- uent on the phosphorus atoms, four additional chirality

nescence lifetimes of the excited states of selected Eu^IIIcom- centres are created upon coordination of the prochiral

plexes in H₂O and D₂O (see Table S1 in the Supporting phosphorus atoms, which leads to R/S isomerism centred

Information).

on the phosphorus atoms. Six diastereoisomers (RRRR,

The luminescence lifetimes obtained for the [Eu- RRRS, RRSS, RSRS, SSSR and SSSS) with a relative sta-

(DOTP^OEt)]^–and [Eu(H_1.5DOTP)]^3.5–(pH ≈ 7) complexes tistical abundance of 1:4:4:2:4:1 are therefore possible for

in H₂O solution are relatively long compared to those for the TSA arrangement. This, in principle, gives rise to a

124

www.eurjic.org

Eur. J. Inorg. Chem. 2009, 119–136

Lanthanide(III) Complexes of Phosphorus Acid Analogues of H₄DOTA

quite complicated isomeric mixture. The isomerism of the SSSS), and very weak signals (four for each of the eight

lanthanide(III) complexes with analogous ligands such as types of protons), which correspond to a very minor asym-

1

DOTP^Otfe(ref.^[39]) and DOTP^Ph(ref.^[15]) has been discussed metric isomer (RRRS or SSSR, Figure S10, H NMR). A

in detail previously.

similar situation occurs for the [Eu(DOTP^Et)]^–complex.

In summary, the ³¹P and ¹H NMR spectra show the pres-

The ³¹P NMR spectra of the [Eu(DOTP^H)]^–, [Eu-

(DOTP^OEt)]^–and [Eu(DOTP^OBu)]^–complexes show the full ence of complex diastereomeric mixtures for [Ln-

set of possible isomers (sixteen ³¹P NMR signals for the six (DOTP^OEt)]^–, [Ln(DOTP^H)]^–and [Ln(DOTP^OBu)]^–com-

isomers) with no predominant isomer, as observed pre- plexes, whereas the [Ln(DOTP^hm)]^–and [Ln(DOTP^Et)]^–

viously for other complexes.^[15,39]The ³¹P NMR spectra of complexes have a preference for a single configuration.

the Yb^IIIand Nd^IIIcomplexes of H₄DOTP^H, H₄DOTP^OEt

,

The solution structures of the H₄DOTP^OEt, H₄DOTP^H

and DOTP^OBushow a similar pattern (Figures S3, S6 and and DOTP^OBucomplexes can be explained in a similar

S7 in the Supporting Information). The ³¹P NMR spectra manner to the complexes of the trifluoroethyl phosphonate

of the [Eu(DOTP^hm)]^–and [Eu(DOTP^Et)]^–complexes, on monoester derivative.^[39]There, the ¹⁹F NMR spectra show

the other hand, are quite simple, with one major signal (in- the presence of all six possible isomers, with the abundances

tegral intensity ca. 70%) and four minor ones (integral ratio of the RRRR/SSSS pair slightly higher than, and that of

1:1:1:1, total integral intensity ca. 25%) being present in the the RSRS isomers slightly lower than, the statistical values

spectra of the [Eu(DOTP^hm)]^–complex (Figure S4 in the by a factor of 1.4–1.5. These deviations were interpreted by

Supporting Information). The major resonance can be as- using a neighbouring interaction model whereby the pop-

signed to the totally symmetric RRRR (or SSSS) and the ulation differences between the various isomers depend only

four minor resonances to the RRRS (or SSSR) isomers, on the energy difference of the interactions of the substitu-

respectively. Signals for other possible diastereomers are ent groups on neighbouring phosphorus atoms in the RR

negligible (total integral intensity Ͻ5%). The Yb^IIIand (or SS) (E_RR≈E_SS) and RS (or SR) (E_RS≈E_SR) orienta-

Nd^IIIcomplexes of H₄DOTP^hmand H₄DOTP^Etalso have tions; E_RSis smaller than E_RRin these compounds by

³¹P NMR spectra very similar to those of the corresponding 0.47 kJmol^–1. This small difference should result from a

Eu^IIIcomplexes (Figures S4 and S5 in the Supporting In- higher steric hindrance between the ester groups and elec-

formation). This points to a strong preference for a single tronic repulsions between the partially charged oxygen

arrangement around all the phosphorus atoms in complexes atoms in the RS orientation. Such a small energy difference

of these two phosphinate ligands, in other words the forma- could also be present in the complexes of the phosphonate

tion of only one of the possible diastereomeric pairs – monoester derivatives. The absence of the ester groups

RRRR-Λ-λλλλ

(+SSSS-∆-δδδδ)

or

SSSS-Λ-λλλλ should decrease the energy difference between those two in-

(+RRRR-∆-δδδδ). A similar preference for these isomers teractions, with isomer abundances closer to the statistical

was observed previously for complexes of H₄DOTP^Bndistribution.

(ref.^[16a]) as well as for complexes of tris(phosphinate)

The preference for a single configuration observed for

monoacetamide cyclen derivatives.^[50]Unfortunately, it was H₄DOTP^hmcould be rationalized by hydrogen-bonding in-

not possible to unambiguously determine the absolute con- teractions between the hydroxy moieties on one phosphorus

figuration of these complexes. The chemical exchange be- acid side-chain and the oxygen atoms of the neighbour

tween the major (RRRR/SSSS) and minor isomers (RRRS/ phosphinate groups. However, such hydrogen bonds are not

SSSR) could be demonstrated by the presence of exchange possible for [Ln(DOTP^Et)]^–complexes, which also show a

cross-peaks in the two-dimensional exchange (EXSY) strong preference for a single configuration. We can specu-

NMR spectrum of the [Yb(DOTP^H)]^–derivative (Figure S8 late that this isomerism may be a consequence of the elec-

in the Supporting Information). In addition, the dynamic tronic properties rather than the steric demands of the sub-

behaviour of the isomer mixture (by variable-temperature stituents on the phosphorus atoms. Thus, ligands in com-

³¹P NMR spectra) was also observed in the cases of other plexes of H₄DOTP^OEt, H₄DOTP^Hand H₄DOTP^OBu(and

ligands (Figure S9 in the Supporting Information).

also H₄DOTP^Phand H₄DOTP^Otfe),^[23,39]which form a mix-

The H NMR spectra of the Eu^IIIcomplexes of all li- ture of isomers, contain electron-withdrawing substituents

gands (see example in Figure S10 in the Supporting Infor- and those of H₄DOTP^Etand H₄DOTP^hm(and also

mation) have resonances for the axial protons at δ = 22– H₄DOTP^Bn),^[16a]which show a preferential isomerism, have

32 ppm, which confirms the exclusive presence of the TSA electron-donating alkyl substituents on the phosphorus

isomer as no signal was observed at the characteristic chem- atoms. The electronic properties could alter the electron dis-

1

–

ical shifts for axial protons of the SA isomer (above δ = tribution inside the PO₂moiety and therefore the different

35 ppm)^{[17,20,21a,22]}The [Eu(DOTP^OEt)]^–, [Eu(DOTP^H)]^–preferences for the isomers.

and [Eu(DOTP^OBu)]^–complexes have very complex ¹H

NMR spectra, thus indicating the presence of many dia-

¹⁷O and ³¹P Lanthanide-Induced Shifts

stereoisomers (a set of resonances was observed for each

proton). In contrast, the ¹H NMR spectrum of the

The ¹⁷O lanthanide-induced shifts (LISs) were measured

[Eu(DOTP^hm)]^–complex contains a set of eight dominant for selected complexes, namely the [Ln(DOTP^hm)]^–com-

peaks, which corresponds to the non-equivalent protons of plexes as representatives of the group of compounds

a dominant isomeric species with C₄symmetry (RRRR or with a dominant diastereoisomer (RRRR/SSSS) and the

Eur. J. Inorg. Chem. 2009, 119–136

www.eurjic.org

125

I. Lukesˇ, J. A. Peters et al.

FULL PAPER

[Ln(DOTP^OEt)]^–and [Ln(DOTP^H)]^–complexes as represen-

The plots for the lanthanide complexes of the other li-

tatives of the other group which shows no preference for gands deviate from linearity towards more negative values

either of the possible isomeric forms.

for one or more of the lighter lanthanides (Ce^IIIfor

The LIS (∆, the lanthanide-induced shift for a concentra- H₄DOTP^Hand H₄DOTP^OEt; Ce^III, Pr^III, and Nd^IIIfor

tion of 1  of Ln^IIIion)^[51]consists of diamagnetic (∆_d), H₈DOTP).^[51]The line through the data according to Equa-

contact (∆_c) and pseudocontact (∆_p) contributions [Equa- tion (3) for Ce^III, Pr^IIIand Nd^IIIcomplexes of H₈DOTP has

tion (1)].

a slope (qF) of –80, which indicates that these light lantha-

nides form complexes of H₈DOTP with one water molecule

in the first coordination sphere of the lanthanide(III) ions.

The large negative values of ∆Ј/C^Dfor the Ce^IIIcomplexes

of H₄DOTP^Hand H₄DOTP^OEtsuggest that these com-

plexes contain the water molecule in the first coordination

sphere as well. The plots obtained from Equation (4)

(Figure 5), where the points corresponding to the Ce^III

complexes of H₄DOTP^Hand H₄DOTP^OEtand those

for the Ce^III, Pr^IIIand Nd^IIIcomplexes of H₈DOTP lie

significantly far from the linear fits, support this conclu-

sion.

∆ = ∆_d+ ∆_c+ ∆_p

(1)

The paramagnetic contribution to the LIS, ∆Ј, obtained

after subtraction of the diamagnetic term, ∆_d, can be ex-

pressed as Equation (2), which contains terms that are

characteristic of each lanthanide(III) ion but independent

of the ligand (ϽS_zϾ and C^Dfor ∆_cand ∆_p, respectively)

and characteristic of the nucleus under study but indepen-

dent of the lanthanide(III) cation (F and G for ∆_cand ∆_p,

respectively).

The negative sign of the slope of the plots obtained from

Equation (4) for the lanthanide(III) complexes of

H₄DOTP^hmis rather surprising when compared with those

observed for the other ligands (Figure 5). As such slopes

are proportional to the geometric term G of the ¹⁷O nuclei

of the water molecules, a change in the sign of the slope

could be a consequence of a different spatial orientation of

the second-sphere hydration shell around the lantha-

nide(III) complexes of H₄DOTP^hmin comparison with the

complexes of the other ligands. As all the ligands are sym-

metrical, the main magnetic axis of their lanthanide(III)

complexes should have the same direction as their pseudo-

C₄axis. Thus, the sign of the ¹⁷O G value is given by the

∆Ј = q(ϽS_zϾϫF + C^DϫG)

(2)

Equation (2) can be linearized to obtain either qF in the

contact contribution from the slope of the plot obtained

from Equation (3) or the geometric term qG in the pseu-

docontact contribution from the slope of the plot corre-

sponding to Equation (4).

∆Ј/C^D= (ϽS_zϾ/C^D)ϫqF + qG

(3)

∆Ј/ϽS_zϾ = qF + (C^D/ϽS_zϾ)ϫG

(4)

Equation (3) is best used to study nuclei with a dominant spatial location of the second-sphere water molecules rela-

contact contribution, such as the ¹⁷O nucleus of the water tive to the dipolar cone defined by the “magic” angle θ =

oxygen atom directly bound to the central lanthanide(III) 54.7° which cancels out the geometric term G = (3ϫcos²θ –

ion in the complexes with inner-sphere water, whereas 1)/r³, where θ is the angle between the main magnetic axis

Equation (4) is preferably used to study nuclei where the of the complex and the line connecting the ion with the ¹⁷

O

pseudocontact contribution dominates, usually those nuclei nucleus at distance r. When these water molecules pass from

lacking any direct bonding to the central lanthanide(III) the inside to the outside part of the cone, the sign of G and

ion, which in the present case is the phosphorus atoms. of the corresponding pseudocontact shift is reversed. The

When the series of lanthanide(III) complexes with a par- value of G is positive inside the cone, whereas it is negative

ticular ligand are isostructural, the plots obtained from outside. The second-sphere water molecules are therefore

Equations (3) and (4) are straight lines.

located outside of the dipolar cone in the H₄DOTP^hmcom-

Analysis of the ¹⁷O LIS data for the Ln^IIIcomplexes of plexes, whereas for H₄DOTP^H, H₄DOTP^OEtand H₈DOTP

H₄DOTP^H, H₄DOTP^hmand H₄DOTP^OEtaccording to they are inside this cone (Figure S11 in the Supporting In-

Equations (3) and (4) is shown in Figure 5. Data for the formation).

complexes of H₈DOTP are also given for comparison. The

These findings suggest that all alkyl groups in the com-

data for the complexes of H₄DOTP^hmshow a linear trend plexes of H₄DOTP^hmand H₄DOTP^Etare oriented above

across the lanthanide series, which indicates that these com- the O₄-plane, thus avoiding an approach of the water mole-

plexes are isostructural. The slope (qF) of the plot accord- cule inside the cone. This is supported by the crystal struc-

ing to Equation (3) is –30, and since the value of F is tures of several lanthanide(III) complexes with H₄DOTP^Bn

,

around –90 under the conditions applied,^[51]this gives q in which all benzylic groups are directed above the O₄-plane

value of around 0.3. A comparable value was obtained by to form a hydrophobic cavity. This arrangement leads exclu-

luminescence lifetime measurements on Eu^IIIcomplexes. sively to RRRR-Λ-λλλλ + SSSS-∆-δδδδ enantiomeric pairs

However, combining these results with the UV/Vis study, in the solid state, and these complexes are also present in

which clearly showed the presence of a single coordination the solution as the single isomer.^[16]It is also supported by

structure, suggests that q is actually 0. The relatively small the high stability of the symmetric arrangement, as evi-

negative slope is probably due to the presence of water in denced by the relatively high interconversion energy be-

the second coordination sphere.

tween isomers of [Yb(DOTP^hm)]^–compared to more labile

126

www.eurjic.org

Eur. J. Inorg. Chem. 2009, 119–136

Lanthanide(III) Complexes of Phosphorus Acid Analogues of H₄DOTA

Figure 5. Linearization of the ¹⁷O LIS data for the lanthanide(III) complexes of H₄DOTP^H(top), H₄DOTP^hm(center top), H₄DOTP^OEt

(center bottom) and H₈DOTP (bottom) according to Equations (3) (slope = contact shift, left) and (4) (slope = pseudocontact shift,

right); T = 301 K, pH 7.

[Yb(DOTP^OEt)]^–(see the variable-temperature ³¹P NMR [Ln(DOTP^Et)]^–were determined for the major RRRR

spectra in Figure S9 of the Supporting Information).

(SSSS) isomer as its signal can easily be assigned in the

Similar equations can be derived for the ³¹P NMR LIS spectra (Figure 6). The plots obtained from Equations (3)

values [q in Equations (3) and (4) should be replaced by 1]. and (4) clearly show a break between the first and second

In the case of [Ln(DOTP^hm)]^–and [Ln(DOTP^Et)]^–com- half of the Ln series, thus indicating a non-isostructurality

plexes, the ³¹P LIS values for [Ln(DOTP^hm)]^–and of these complexes across the lanthanide series. The com-

Eur. J. Inorg. Chem. 2009, 119–136

www.eurjic.org

127

I. Lukesˇ, J. A. Peters et al.

FULL PAPER

plex isomeric mixtures present for the [Ln(DOTP^H)]^–, {the data for [Ln(DOTP^H)]^–are included in Figure 6}.

[Ln(DOTP^OEt)]^–and [Ln(DOTP^OBu)]^–complexes do not al- Analogous ³¹P LIS phenomena have also been observed

low a reliable assignment of the resonances, therefore the previously for Ln^IIIcomplexes of H₄DOTP^Phand

weighted average of the chemical shifts for all isomers (i.e. H₄DOTP^{Bn [15,16]}

.

the chemical shift corresponding to half of the total inten-

The positions of the breaks observed in the plots of the

sity of all signals in the ³¹P NMR spectra) was used. The ³¹P NMR spectroscopic data do not fully correspond with

corresponding plots are very similar to the previous case those observed in the ¹⁷O NMR spectroscopic data, poss-

Figure 6. Separation of the contact [Equation (3), left] and pseudocontact [Equation (4), right] contributions to the ³¹P LIS for the

weighted chemical shift for the [Ln(DOTP^H)]^–complexes (top), the RRRR (SSSS) isomers of the [Ln(DOTP^hm)]^–(center top) and

[Ln(DOTP^Et)]^–(center bottom) complexes, and [Ln(H_1.5DOTP)]^3.5–(bottom); T = 301 K, pH 7.

128

www.eurjic.org

Eur. J. Inorg. Chem. 2009, 119–136

Lanthanide(III) Complexes of Phosphorus Acid Analogues of H₄DOTA

ibly due to a change in the water coordination of the lan- of geometry of the donor atom environment between large

thanide ions. Since the breaks in Figure 6 indicate a change (La^III–Eu^III) and small (Eu^III–Lu^III) ions, which is appar-

of both F and G for the ³¹P nuclei of these complexes across ently reflected in changes of both F and G for the ³¹P nuclei.

the lanthanide series, it is highly likely that some geometric

change in the TSA coordination sphere is occurring. The

ions along the lanthanide(III) series gradually move inside

the ligand cavity (closer to the N₄-plane) in the TSA iso-

Relaxometric and EPR Measurements

mers – the Ln–QN distance (QN is the centre of gravity of

Variable-temperature ¹⁷O NMR spectroscopic data [T₁

the N₄-plane) drops steeply from around 1.85 Å for La^IIIto and T₂relaxation times and ¹⁷O angular frequencies (ω_O)]

around 1.65 Å for Eu^III, and remains in the range 1.65– were collected for the Gd^IIIcomplexes of H₄DOTP^hm

,

1.55 Å for the rest of the series.^[23,37]This leads to a change H₄DOTP^Hand H₄DOTP^OEt. For comparative purposes,

Figure 7. The ¹H NMRD profiles for the Gd^IIIcomplexes of H₄DOTP^H, H₄DOTP^hm, H₄DOTP^Et, H₄DOTP^OEt, H₄DOTP^OBuand

H₄DOTP. Profiles were measured at pH 7 (except for H₈DOTP) and 10 (H₈DOTP). The full lines represent the results of the best

1

simultaneous fits of H NMRD and EPR data.

Eur. J. Inorg. Chem. 2009, 119–136

www.eurjic.org

129

I. Lukesˇ, J. A. Peters et al.

FULL PAPER

the data for the [Gd(DOTP)]^5–complex were also measured method from their molecular weights, and the correspond-

at a pH of around 10 (where all pendant arms are fully ing activation energy was fixed at 18.2 kJmol^–1.^[53]The

deprotonated) and a pH of about 7 {where approximately shortest distance of a water proton to the Gd^IIInucleus (the

1.5 protons are bound to the ligand, i.e. [Gd(H_1.5

-

closest approach, a_GdH) was fixed at 3.50 Å, the same value

DOTP)]^3.5–[44]}. In this case, no significant influence of the as in most previous studies. The water molecules in the sec-

pH on the data obtained was observed. The ¹⁷O NMR ond coordination sphere are assumed to be hydrogen-

spectroscopic data are shown in Figure S12 (see the Sup- bonded to the Gd^III-coordinated O-atoms of the ligands.

porting Information). Since q = 0 for these systems, the Accordingly, the effective distance between Gd^IIIand these

paramagnetic effects on the ¹⁷O NMR parameters are so protons (R_GdH) was fixed at 3.6 Å. The residence time of

small that a fitting to theoretical models is not feasible.

The ¹H NMRD profiles of the Gd^IIIcomplexes of 56 ps, the value calculated by Borel et al. for [Gd-

H₄DOTP^H, H₄DOTP^hmH₄DOTP^Et, H₄DOTP^OEt(DOTP)]^5–by molecular dynamics, and the corresponding

the second-sphere water molecules (τ_M,ss) was taken as

,

H₄DOTP^OBuand H₈DOTP at 5, 25 and 37 °C are shown activation energy was fixed at 35 kJmol^–1.^[29]The rotational

in Figure 7. The Gd^IIIcomplexes of H₄DOTP^Et, H₄DOTP^OEtcorrelation times (τ_R) were estimated from the previously

and H₈DOTP were re-measured, and the published ¹H evaluated correlation time for [Gd(H₂O)(DOTA)]^–after ap-

NMRD profile for the H₄DOTP^OBucomplex^[13]was used. plication of a correction for the molecular volume by means

These profiles were re-analysed as the second-sphere hy- of the Stokes–Debye–Einstein equation, and the activation

dration was not considered in previous relaxometric stud- energy for this correlation time was fixed at 25.4 kJmol^–1

ies.^[13]The values of the r₁relaxivities of phosphinates and for all complexes. Only q_ss, the mean-square zero-field split-

phosphonate monoesters are about half those measured for ting energy (∆²) and the correlation time for the modulation

[Gd(H₂O)(DOTA)]^–; for example, the relaxivity at 20 MHz of the zero-field splitting interaction (τ_v) remained as ad-

and 310 K is about 2.1 s^–1mm^–1for all these complexes justable parameters in the fitting procedure. The activation

(Table 6). The ¹H NMRD profiles of the [Gd(DOTP^H)]^–energy of the latter parameter was fixed to a reasonable

and [Gd(DOTP^hm)]^–complexes are similar to each other, in value of 1 kJmol^–1because fittings with this as a variable

contrast to those of the [Gd(DOTP^Et)]^–, [Gd(DOTP^OEt)]^–parameter led to negative values. Inclusion of the ¹H

and [Gd(DOTP^OBu)]^–complexes, which have significantly NMRD profiles measured at 5 °C resulted in very unrea-

lower relaxivities at low fields. As all these complexes have sonable values of ∆²and τ_v, therefore these profiles were

approximately the same molecular size, these differences given zero weight in the fittings.

The best values of the adjustable parameters for the Gd^III

complexes of these ligands are given in Table 6. The values

of T_2e,exp(from EPR data) and T_2e,calc, calculated by using

the best-fit parameters τ_vand ∆², are in good agreement

(Table 5). The calculated ¹H NMRD profiles are repre-

sented by the curves in Figure 7.

can be attributed to a slower electronic spin relaxation and/

or to a somewhat higher polarity of the H₄DOTP^Hand

H₄DOTP^hmcomplexes in comparison to the other com-

plexes, which can lead to a richer second hydration sphere

in the former complexes.

To obtain more information about electronic parameters,

transverse electronic spin relaxation times (T_2e,exp) were cal-

culated from the peak-to-peak EPR line widths (∆H_pp) ac-

cording to Equation (5), where g_Lis the Landé g factor

The values of the electronic parameters τ_vand ∆²could

1

not be determined accurately from the H NMRD profiles,

although the profiles appeared to be very sensitive to the

value of q_ss. The best-fit parameters obtained (Table 6) also

support the preliminary conclusion made above, namely

that the [Gd(DOTP^H)]^–and [Gd(DOTP^hm)]^–complexes

have a slower electronic spin relaxation and a somewhat

richer second hydration sphere than the [Gd(DOTP^Et)]^–,

[Gd(DOTP^OEt)]^–and [Gd(DOTP^OBu)]^–complexes (see also

the luminescence measurements). As expected, the highest

1

(Table 5).^[52]The H NMRD data were then fitted simulta-

neously with the X-band EPR data by using a set of equa-

tions suitable for complexes with an extended second hy-

dration sphere.^[22,27b]

(5)

Since the number of parameters in these equations is second-sphere hydration was found in the case of

rather large, and some of them correlate strongly, we fixed [Gd(DOTP)]^5–. The values for ∆²are an order of magnitude

many of them. The diffusion coefficients of the various larger than that for [Gd(H₂O)(DOTA)]^–, which may attrib-

complexes were fixed at values obtained by a semi-empirical uted to the lower rigidity of the complexes of H₄DOTP^R

1

Table 5. Experimental (1/T_2e,exp, calculated from EPR ∆H_pp) and calculated (1/T_2e,fit, simultaneous fitting of H NMRD and EPR data)

1/T_2evalues for the Gd^IIIcomplexes.

Parameter

[Gd(DOTP^H)]^–[Gd(DOTP^hm)]^–[Gd(DOTP^Et)]^–[Gd(DOTP^OEt)] [Gd(DOTP^OBu)] [Gd(DOTP)]^5–[Gd(H₂O)(DOTA)]^–

–

∆H_pp[Gauss]

275Ϯ5

4.19

280Ϯ5

4.27

415Ϯ2

6.33

425Ϯ5

6.40

362Ϯ16

5.49

605Ϯ10

9.22

9.25

91^[a]

1.40

–

1/T_2e,exp[10⁹s^{–1 [b]}

]

1/T_2e,fit[10⁹s^–1

]

[a] Ref.^[3][b] Values calculated according to Equation (5).

130

www.eurjic.org

Eur. J. Inorg. Chem. 2009, 119–136

Lanthanide(III) Complexes of Phosphorus Acid Analogues of H₄DOTA

1

Table 6. Results of multi-parameter simultaneous fitting of H NMRD and EPR data for Gd^IIIcomplexes of the H₄DOTP^Rderivatives

(at pH 7) and H₈DOTP (at pH 10). The values in italics were fixed during the fitting.

Parameter

[Gd(DOTP^H)]^–[Gd(DOTP^hm)]^–[Gd(DOTP^OEt)] [Gd(DOTP^OBu)] [Gd(DOTP^Et)]^–[Gd(DOTP)]^5–[Gd(H₂O)(DOTA)]^–

–

[s^–1mm^–1

]

2.2^[a]

2.6^[c]

2.2^[a]

2.6^[c]

2.1^[a]

2.4^[c]

2.3^[a]

2.5^[c]

2.8^[a]

2.5^[c]

3.5^[a]

3.5^[c]

3.8^[b]

4.4^[d]

2.4^[f]

0^[h]

310

r

1,exp

r

[s^–1mm^–1

]

1,calc

310

r

[s^–1mm^–1

]

2.1^[e]

2.2^[e]

2.1^[e]

1os,calc

310

r

[s^–1mm^–1

]

0.5^[g]

0.3^[g]

0.4^[g]

0.3^[g]

1.4^[g]

1ss,calc

q_ss

1.32Ϯ0.05

1.22Ϯ0.06

106.0^[i]

2.6Ϯ2.0

1.5Ϯ1.0

214

0.68Ϯ0.04

127.1^[i]

3.8Ϯ1.3

1.6Ϯ0.5

137

0.80Ϯ0.04

162.0^[i]

4.8Ϯ1.1

1.1Ϯ0.2

158

0.79Ϯ0.06

117.3^[i]

5.0Ϯ4.6

1.0Ϯ0.8

157

3.3Ϯ0.1

90.7^[i]

12.7Ϯ0.2

0.9Ϯ0.1

73

0

²⁹⁸τ_R[ps]

²⁹⁸τ_v[ps]

86.6^[i]

77^[j]

11^[j]

0.16^[j]

473

–

2.7Ϯ1.7

1.4Ϯ0.8

220

∆²[10²⁰s^–2

]

298

τ

[ps]^[k]

S0

298

D

[10^–9m²s^{–1 [l]}

]

2.48

2.40

2.37

2.33

2.38

2.43

GdH

[a] Experimental relaxivity values as measured at 310 K and 20 MHz. [b] Ref.^[54], 312 K. [c] Relaxivity at 310 K and 20 MHz as calculated

with the best-fit parameters reported in this paper. [d] Relaxivity at 310 K and 20 MHz as calculated with the best-fit parameters reported

in ref.^[54][e] Outer-sphere contribution to relaxivity at 310 K and 20 MHz as calculated with the best-fit parameters reported in this

paper. [f] Outer-sphere contribution to relaxivity at 310 K and 20 MHz as calculated with the best-fit parameters reported in ref.^[54][g]

Second-sphere contribution to relaxivity at 310 K and 20 MHz as calculated with the best-fit parameters reported in this paper. [h]

Second-sphere contribution to relaxivity at 310 K and 20 MHz as calculated with the best-fit parameters reported in ref.^[54][i] Calculated

using a molecular volume as evaluated with HyperChem^®and the Stokes–Debye–Einstein equation. [j] Ref.^[54][k] Low field limiting value

of the electronic relaxation correlation time as calculated with τ_S0= (12∆²τ_v)^–1. [l] Calculated from the molecular weight using a previously

reported equation.^[52]

and H₈DOTP due to the lack of an inner-sphere water parent system H₄DOTA and, consequently, the Gd^IIIcom-

molecule, which leads to larger deformation of the com- plexes are thermodynamically much less stable. Titrations

plexes upon collision with solvent molecules. As a result, have shown that there is no (de)protonation of the pre-

the electronic relaxation rates of the presently studied com- formed complexes in the pH range 3–10. Luminescence

plexes are lower than that of [Gd(H₂O)(DOTA)]^–(ref.^[54]), data do not allow an unambiguous conclusion regarding

which means that the outer-sphere contribution to the re- the value of q for the Eu^IIIcomplexes of the tetraphospho-

laxivity of these complexes is somewhat lower. The second- rus acid derivatives as the measured lifetimes suggest q

sphere water molecules {except for [Gd(DOTP)]^5–complex} valus of around 0.5. However, UV/Vis measurements on the

have a contribution of 10–20% to the total relaxivity Eu^IIIcomplexes confirm that heavier lanthanide(III) com-

(Table 6).

plexes lack an inner-sphere water molecule.

The pH-dependence of the relaxivity of the Gd^IIIcom-

This finding was confirmed by the X-ray structures of

plexes in solution was also examined. No protonation of the Gd^IIIcomplexes. Exclusive formation of the TSA iso-

the pendant arms is expected above a pH of around 3 for mer was observed by both multinuclear NMR spectroscopy

phosphinic acid or phosphonic monoester groups as they and by X-ray diffraction. The presence of inner-sphere co-

are highly acidic (see above),^[37]therefore no pH depen- ordinated water was found only for the [Ce(H₂O)-

dence was expected. Surprisingly, the relaxivity of the (DOTP^H)]^–,

[Ce(H₂O)(DOTP^OEt)]^–

and

[Ln(H₂O)-

[Gd(DOTP^hm)]^–complex was found to be highly dependent (DOTP)]^5–(Ln = Ce, Pr, Nd) complexes through the mea-

on the pH, in sharp contrast to the relaxivities of other surements of ¹⁷O lanthanide-induced shifts. The ³¹P NMR

complexes, which were found to be independent of the pH study revealed striking differences in isomerism between the

(tested at pH 4,

7

and 10). The relaxivity of the [Ln(DOTP^hm)]^–and [Ln(DOTP^Et)]^–complexes on the one

[Gd(DOTP^hm)]^–complex increases by more than 50% upon hand and the [Ln(DOTP^H)]^–, [Ln(DOTP^OEt)]^–and

going from neutral to both acidic and alkaline pH regions [Ln(DOTP^OBu)]^–complexes on the other. Thus, whereas

(Figure S13 in the Supporting Information). This observa- rich isomeric mixtures were found in the latter cases, the

tion can be explained by a prototropic exchange similar to [Ln(DOTP^hm)]^–and [Ln(DOTP^Et)]^–complexes show a clear

that observed for the Gd^IIIcomplexes of H₄DOTA preference for the RRRR/SSSS arrangement on the phos-

amides.^[46]In the strongly acid region (pH Ͻ 2), a partial phorus atoms, although they probably form RRRR-Λ-λλλλ

decomplexation probably takes place leading to a further + SSSS-∆-δδδδ enantiomeric pairs similar to those found

relaxivity increase.

previously in the case of [Ln(DOTP^Bn)]^–complexes.^[16]This

is probably caused by a strong interaction of the side-chains

in the preferential arrangement.

Conclusions

To explore the second-hydration sphere in more detail,

a relaxometric study on all the Gd^IIIcomplexes was also

performed. The ¹H NMRD and EPR data were fitted

simultaneously, and the best-fit parameters suggest that

about three second-sphere water molecules are present in

the case of [Gd(DOTP)]^5–and about one second-sphere

We have synthesized and studied a series of tetraphos-

phorus acid ligands and, in particular, their complexes with

Ln^IIIions. These ligands and their complexes were charac-

terized by several methods in order to gain insight into their

stabilities and structures. The ligands are less basic than the

Eur. J. Inorg. Chem. 2009, 119–136

www.eurjic.org

131

I. Lukesˇ, J. A. Peters et al.

FULL PAPER

hypophosphorous acid (7.2 mL, 66 mmol) were mixed in water

(20 mL). The mixture was heated to 40 °C whilst being stirred, and

solid paraformaldehyde (0.78 g, 26 mmol) was added in small por-

tions over 60 min. The mixture was further heated at this tempera-

ture for 1 d. Excess paraformaldehyde was then filtered off. After

concentration to a small volume under reduced pressure, the reac-

tion mixture was purified on a cation exchange resin (Dowex 50,

H⁺-form, 300 mL, elution with water). Phosphinic acid and other

acidic impurities were eluted first followed by the product. Frac-

tions containing the pure ligand (checked by ³¹P NMR) were com-

water molecule is present in the other complexes. The elec-

tronic spin relaxation parameters τ_vand ∆²obtained sug-

gest that the more hydrophilic [Gd(DOTP^H)]^–and

[Gd(DOTP^hm)]^–complexes have a slower electronic spin re-

laxation and a somewhat richer second hydration sphere

than the [Gd(DOTP^Et)]^–, [Gd(DOTP^OEt)]^–and [Gd-

(DOTP^OBu)]^–complexes. The pH-dependence of the re-

laxivity observed for the [Gd(DOTP^hm)]^–complex reflects

the proton/hydroxide catalysis of the prototropic exchange

mechanism, similar to that observed for the Gd^IIIcom- bined and concentrated in vacuo. Microcrystallinic H₄DOTP^H·

plexes of H₄DOTA amides.^[46]

0.5H₂O (0.66 g, 23%) was obtained after slow evaporation of the

solvent from an aqueous ligand solution. ¹H NMR (D₂O): δ =

The observed relaxivities of all [Gd(DOTP^R)]^–complexes

fall into the narrow region 2.1–2.8 s^–1m^–1(20 MHz,

37 °C). The relaxivity of Gd^IIIcomplexes with DOTA-like

ligands having one inner-sphere water molecule is higher

(≈ 4),^[22,54]although in these cases the inner-sphere contri-

butions are about 50% of the final value, whereas the sec-

ond-sphere water molecules of the presently studied com-

plexes have a contribution of only 10–15% to the total re-

laxivity. The ligands studied differ in the size and hydrophi-

lic/hydrophobic character of the substituents on the phos-

phorus atoms and in the isomeric distributions of their

complexes in solution. However, this has only a marginal

2

2.53–2.55 (m, J_P,H= 8.8 Hz, 24 H, CH₂P, NCH₂CH₂N), 7.04 (d,

¹J_P,H= 504.0 Hz, 4 H, PH) ppm. ³¹P NMR (D₂O): δ = 22.5 (dt,

2

¹J_P,H= 504, J_P,H= 9.2 Hz) ppm. ³¹P{¹H} NMR (D₂O): δ = 22.5

(s) ppm. C₁₂H₃₂N₄O₈P₄·0.5H₂O (493.31): calcd. C 29.22, H 6.74,

N 11.36; found C 29.29, H 6.62, N 11.38.

1,4,7,10-Tetraazacyclododecane-1,4,7,10-tetrakis[(hydroxymethyl)-

(methyl)phosphinic acid] (H₄DOTP^hm): Paraformaldehyde (0.93 g,

31 mmol) was added to an aqueous solution of H₄DOTP^Hhydrate

(1.00 g, 2.03 mmol), and the mixture was stirred and heated under

reflux for 1 d. The mixture was then concentrated to dryness under

reduced pressure and redissolved in a small amount of water. The

solution was purified on a cation exchange column (Dowex 50, H⁺-

influence on the relaxivity. This indicates that the second- form, 300 mL, elution with water). The first fractions containing

the product were combined, and the solvent was evaporated. The

residue was dissolved in a small amount of water, and the product

was precipitated by addition of acetone. The solid was filtered off

and dried with P₂O₅. A white microcrystalline solid was isolated in

51 % yield (0.65 g). ¹H NMR (D₂O): δ = 2.70–2.60 (m, 24 H,

CH₂OH, NCH₂CH₂N), 3.46 (d, ²J_P,H= 5.6 Hz, 8 H, NCH₂P) ppm.

³¹P{¹H} NMR (D₂O): δ = 38.2 (s) ppm. C₁₆H₄₀N₄O₁₂P₄·1.5H₂O

(631.43): calcd. C 30.44, H 6.86, N 8.87; found C 30.50, H 6.68, N

8.66.

sphere water molecules contributing to the relaxivity (i.e.

residing close to the central Gd^IIIion) are mainly connected

–

with the oxygen atoms of electronegative PO₂moiety

through hydrogen bonds, independently on the orientation

of the side chains and, consequently, on the local arrange-

ment of the second-sphere water molecules.

Experimental Section

Monoethyl 1,4,7,10-Tetraazacyclododecane-1,4,7,10-tetrakis-

(methylphosphonate) (H₄DOTP^OEt):^[32]Cyclen (1.00 g, 5.80 mmol)

was treated with paraformaldehyde (0.87 g, 29 mmol) and triethyl

phosphite (9.8 g, 59 mmol, as solvent) to produce the octaethyl es-

ter Et₈DOTP. The mixture was stirred at 40 °C for 3 d and then

loaded onto a cation exchange column (Dowex 50, H⁺-form,

300 mL). Di- and triethyl phosphite were eluted with EtOH, and

Et₈DOTP was washed out with a 1:3 (v:v) mixture of 25% aq.

NH₃/EtOH. The fractions containing the desired ester were con-

centrated in vacuo, after which the purity and the identity of this

compound was verified by NMR spectroscopy [¹H NMR (CDCl₃):

General: Cyclen (1,4,7,10-tetraazacyclododecane, Strem or CheMa-

Tech), hypophosphorous acid (Fluka), triethyl phosphite (Fluka),

LnCl₃hydrates (Strem or Aldrich), D₂O (99.95%D, Chemtrade)

and Dowex 50 (ϫ4, 100–200 mesh, Fluka) were used without fur-

ther purification. Paraformaldehyde was filtered from an old form-

aldehyde solution and dried over P₂O₅in vacuo. Anhydrous EtOH

was obtained from Lachema (Czech Republic). H₄DOTP^Etand

H₄DOTP^OBuwere synthesized according to literature pro-

cedures.^{[13,33] 1}H (400 MHz), ¹³C (101 MHz) and ³¹P (162 MHz)

NMR spectra were recorded with a Varian INOVA 400 spectrome-

ter in CDCl₃, D₂O or H₂O solutions at 25 °C, or with a Varian

Unity 500 spectrometer (¹H: 499.82 MHz; ³¹P: 202.33 MHz). For

measurements in D₂O, the internal standard was tBuOH; in

CDCl₃, the internal standard was TMS; for ³¹P, 85 % H₃PO₄was

used as an external standard. The elemental analyses were carried

out at the Institute of Macromolecular Chemistry (Academy of

Sciences of the Czech Republic, Prague). Luminescence measure-

ments were performed with an Aminco Bowman^®Series 2 spec-

trometer by using the excitation at the ⁵L₆ǟ⁷F₀band (396 nm).

3

δ = 1.27 (t, J_{H , H}= 7.2 Hz, 24 H, CH₃), 2.82 (s, 16 H,

NCH₂CH₂N), 2.94 (d, ²J_P,H= 9.2 Hz, 8 H, NCH₂P), 4.03–4.11 (m,

16 H, OCH₂) ppm. ¹³C NMR (CDCl₃): δ = 16.5 s (CH₃), 50.6 (d,

¹J_C,P= 151.5 Hz, NCH₂P), 53.4 (s, NCH₂CH₂N), 61.5 (s, OCH₂)

ppm. ³¹P(¹H) NMR (CDCl₃): δ = 26.7 (s) ppm. The target ligand

was obtained after hydrolysis (1 d, 50 °C) of this ester in aq. NaOH

(10%, 25 mL). The sodium ions were completely removed by pass-

ing the reaction mixture through a cation exchange resin

(Dowex 50, H⁺-form, 300 mL, elution with water). Evaporation of

the water from the eluate resulted in a colourless oil. The ligand

was obtained as a white microcrystalline solid after precipitation

from its aqueous solution by slow addition of acetone, filtering and

drying over P₂O₅(3.91 g, 92% based on cyclen). ¹H NMR (D₂O): δ

= 1.18 (t, ³J_H,H= 6.8 Hz, 12 H, CH₃), 3.20–3.27 (m, 24 H, NCH₂P,

NCH₂CH₂N), 3.89–3.96 (m, 8 H, OCH₂) ppm. ³¹P{¹H} NMR

(D₂O): δ = 19.4 (s) ppm. C₂₀H₄₈N₄O₁₂P₄·4H₂O (732.57): calcd. C

32.79, H 7.70, N 7.65; found C 32.65, H 6.94, N 7.63.

7

Emission lifetimes were measured for the F₂ǟ⁵D₀transition (615

and 621 nm). In general, the samples prepared for NMR measure-

ments were also used for luminescence measurements. UV/Vis spec-

5

tra of the D₀ǟ⁷F₀transition were acquired with a Perkin–Elmer

Lambda 19 spectrometer, in data steps of 0.01 nm. The concentra-

tion of the samples was around 0.02 .

1,4,7,10-Tetraazacyclododecane-1,4,7,10-tetrakis(methylphosphinic

acid) (H₄DOTP^H): Cyclen (1.00 g, 5.80 mmol) and 50% aqueous

132

www.eurjic.org

Eur. J. Inorg. Chem. 2009, 119–136

Lanthanide(III) Complexes of Phosphorus Acid Analogues of H₄DOTA

1,4,7,10-Tetraazacyclododecane-1,4,7,10-tetrakis(methylphosphonic

acid) (H₈DOTP): The target ligand was obtained after acid hydrol-

ysis (aq. HCl 1:1, 2 d, reflux) of the ester obtained as above. The

reaction mixture was concentrated, and the residue was recrys-

tallized from boiling water. The ligand, in the form of a white

microcrystalline solid, was collected by filtration and dried in air

(95% based on cyclen). Spectroscopic data were identical with

those reported in the literature.^[35]C₁₂H₃₂N₄O₁₂P₄·1.5H₂O

(575.33): calcd. C 25.05, H 6.13, N 9.74; found C 25.01, H 6.04, N

9.78.

H₄DOTP^OEt·H₂O were refined anisotropically except for some dis-

ordered ethyl groups of the pendant arms. The hydrogen atoms

belonging to the carbon atoms were fixed in the theoretical posi-

tions by using the riding model with U_eq(H) = 1.2U_eq(C). Some

hydrogen atoms attached to amino groups or oxygen atoms were

located in the electron density difference map. In the case of the

amino groups, the protons were fixed in the theoretical positions,

and the protons of the oxygen atoms were refined in the original

positions with thermal parameters U_eq(H) = 1.2U_eq(X). All non-

hydrogen atoms in the structure of Li[Gd(DOTP^H)]·6H₂O were re-

fined anisotropically, and all hydrogen atoms were fixed in their

theoretical (C–H and P–H) or original (O–H) positions by using

U_eq(H) = 1.2U_eq(X). One pendant arm was best refined when

turned into two positions (with R and S configuration on the phos-

phorus atom, respectively), and the lithium counterion was fitted

in two positions as well (as it is coordinated by an oxygen atom of

a partially occupied oxygen atom of a disordered phosphinate). All

non-hydrogen atoms in the structure of Li[Gd(DOTP^OEt)]·

0.5Me₂CO·8H₂O were refined anisotropically except for one disor-

dered ethyl group of the pendant arms and one disordered water

solvate molecule. The hydrogen atoms belonging to carbon atoms

were fixed in their theoretical positions by using the riding model

with U_eq(H) = 1.2U_eq(C). Some of the hydrogen atoms attached to

the solvate molecules were localized in the difference Fourier map

and were treated in their original positions with U_eq(H) =

1.2U_eq(O). CCDC-704357 (H₄DOTP^OEt·H₂O), -704358 {Li[Gd-

(DOTP^OEt)]·0.5Me₂CO·8H₂O}, and -704359 {Li[Gd(DOTP^H)]·

6H₂O} contain the supplementary crystallographic data for this

paper. These data can be obtained free of charge from The Cam-

bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/

data_request/cif.

Lanthanide(III) Complexes of the Ligands: Samples of the Ln^III

complexes for NMR measurements were prepared by dissolving

LnCl₃·6H₂O and a 10% molar excess of the ligand in water, in-

creasing the pH to about 7 with an aq. KOH solution and leaving

the mixture for 1 h; ¹⁷O-enriched water was added to the samples

for ¹⁷O variable-temperature and LIS NMR measurements to a

concentration of about 0.5% ¹⁷O. The samples for structural stud-

ies were prepared analogously, but in D₂O, to yield 0.1  solutions

of the complexes in all cases. The samples for ³¹P NMR LIS were

prepared in 20% D₂O. The pH of all resulting solutions was ad-

justed to 7 with either dilute HCl (DCl) or KOH (KOD). The ab-

sence of free Ln^IIIions was verified by using xylenol orange as

indicator in acetate buffer at a pH of about 5.5. Samples of the

Eu^IIIcomplexes for UV/Vis spectroscopy were prepared analo-

gously, and the samples originally prepared for NMR studies were

used for luminescence experiments. The samples for measurements

1

of H NMRD profiles were prepared by exact dilution of the sam-

ples for ¹⁷O NMR measurements to the final concentration of

5 m.

X-ray Studies: Single crystals of all studied compounds were ob-

tained by slow vapour diffusion of acetone into aqueous solutions

of the compounds. For the complexes, dilute aq. LiOH was used

for pH adjustment. The diffraction data were collected by using a

Nonius Kappa CCD diffractometer (Enraf–Nonius) at 150(1) K

with Mo-K_αradiation (λ = 0.71073 Å) and analysed by using the

HKL program package.^[55]The structures were solved by direct

methods and refined by full-matrix least-squares techniques

(SIR92^[56]and SHELXL97^[57]). Scattering factors for neutral atoms

were included in the SHELXL97 program. Experimental data are

given in Table 7. All non-hydrogen atoms in the structure of

Potentiometric Measurements: Potentiometric titrations were run,

and the data obtained were treated according to a previously pub-

lished procedure.^[58]Protonation/dissociation and stability con-

stants of H₄DOTP^Hor H₄DOTP^hmand their complexes were de-

termined in 0.1  Me₄NCl at 25 °C. The stability constants of Gd^III

complexes were obtained by the out-of-cell method with a waiting

time for equilibrium of three weeks (some solutions were checked

after six weeks and found to give the same data). Titrations were

performed in the pH range 1.8–6.0 with around 30 data points per

Table 7. Crystallographic parameters of the studied compounds.

Compound

H₄DOTP^OEt·H₂O

Li[Gd(DOTP^H)]·6H₂O

Li[Gd(DOTP^OEt)]·0.5Me₂CO·8H₂O

Empirical formula

M_r

Colour,

C₂₀H₅₀N₄O₁₃P₄

678.52

colourless

rod

C₁₂H₄₀GdLiN₄O₁₄P₄

752.55

colourless

rod

C_21.5H₆₃GdLiN₄O_20.5P₄

993.83

colourless

prism

Habit

Crystal system

Space group

a [Å]

b [Å]

c [Å]

rhombic

Pca2₁

19.4095(2)

13.4920(2)

24.5857(3)

90

6438.3(1)

8

1.400

0.298

12198

10090

0.0535

0.0697

0.1348

0.1468

monoclinic

C2/c

rhombic

P2₁2₁2₁

14.2862(1)

24.2628(2)

25.0544(2)

90

8684.4(1)

8

1.520

1.748

19853

17491

0.0395

17.5713(3)

6.9336(1)

22.3551(4)

101.177(1)

2671.92(8)

4

1.871

2.790

3055

2965

0.0235

0.0246

0.0581

0.0589

β [°]

V [Å³]

Z

D_calcd.[gcm^–3]

µ [mm^–1]

Total refl.

Obsd. refl. [I Ͼ 2σ(I)]

R

RЈ [I Ͼ 2σ(I)]

wR

wRЈ [I Ͼ 2σ(I)]

0.0503

0.0926

0.0987

Eur. J. Inorg. Chem. 2009, 119–136

www.eurjic.org

133

I. Lukesˇ, J. A. Peters et al.

FULL PAPER

titration and three titrations per system. The pre-formed gadolini-

um(III) complexes in solution were obtained by mixing a known

amount of ligand (5% molar excess) with GdCl₃(as a defined stock

solution) in a glass ampoule. A standard Me₄NOH solution (just

to neutralize the ligand amount) was slowly added, and the am-

poule was sealed and left at 80 °C overnight to fully complex the

metal ion. The ampoules were opened, and aliquots of the solu-

tions of the gadolinium(III) complexes were transferred to a ti-

tration vessel. Excess of HCl and Me₄NCl solutions were added

(to reach a pH of about 2 and 0.1  Me₄NCl in the final solution)

TIST^®for WINDOWS^TMby Micromath^®(ref.^[62]) using a set of

equations published previously.^[27a]Molecular sizes of the com-

plexes were estimated by constructing molecular models with Hy-

perChem^®, Version 7.5.^[63]

Supporting Information (see footnote on the first page of this arti-

cle): Molecular structures of two independent units found in the

structure of H₄DOTP^OEt·H₂O; UV/Vis spectra for the ⁵D₀ǟ⁷F₀

transition of [Eu(DOTP^H)]^–and [Eu(DOTP^OEt)]^–complexes;

³¹P{¹H} NMR spectra of Nd^III, Eu^IIIand Yb^IIIcomplexes

of H₄DOTP^H, H₄DOTP^hm

,

H₄DOTP^Et, H₄DOTP^OEtand

and the solution was immediately titrated with

a standard

H₄DOTP^OBu; an example of a ³¹P EXSY NMR spectrum for the

[Yb(DOTP^H)]^–complex; temperature dependence of the ³¹P NMR

spectra of the [Yb(DOTP^OEt)]^–, [Eu(DOTP^OEt)]^–and [Yb-

(DOTP^hm)]^–complexes; ¹H NMR spectra of the [Eu(DOTP^hm)]^–

and [Eu(DOTP^OEt)]^–complexes; schematic representation of the

McConnel cone in the complexes; variable-temperature ¹⁷O re-

duced relaxation rates and reduced angular frequencies of Gd^III

complexes of H₄DOTP^H, H₄DOTP^hm, H₄DOTP^OEtand H₈DOTP;

dependence of the relaxivity of the [Gd(DOTP^hm)]^–complex on

pH; luminescence lifetimes (τ_ex) and the corresponding hydration

numbers of selected Eu^IIIcomplexes.

Me₄NOH solution up to a pH of about 12 at 25 °C, acquiring

around 45 data points for each of three titrations.

EPR Measurements: X-band EPR spectra of aqueous solutions of

the Gd^IIIchelates (1 m, pH of about 7.0) were obtained at 25 °C

by using a Bruker ESP-300E spectrometer operating at 9.43 GHz

(0.34 T) with a quartz flat cell. Spectral acquisition parameters:

sweep width: 40 mT; microwave power: 0.632 mW; modulation am-

plitude: 0.32 mT; time constant: 20.48 ms.

NMR and Relaxation Studies of the Lanthanide(III) Complexes: ¹H

(300.0 MHz), ¹⁷O (40.7 MHz) and ³¹P (121.5 MHz) NMR spectra

were acquired with a Varian VNMRS 300 spectrometer by using a

1

5 mm broadband probe. H (400.0 MHz), ¹⁷O (54.2 MHz) and ³¹P

Acknowledgments

(161.9 MHz) NMR spectra were also acquired with a Varian Unity

INOVA 400 spectrometer and ¹H (500.0 MHz) and ³¹P

(202.3 MHz) NMR spectra with a Varian Unity 500 spectrometer

using the same type of probe. BMS effects were corrected by using

the internal D₂O lock signals. Water was used as solvent for the

variable-temperature ¹⁷O NMR shifts and LIS measurements,

therefore no frequency lock could be applied. In this case, the cor-

rection for the BMS shifts was performed with the ¹H resonance

of the internal standard tBuOH. H₃PO₄(85%) was used as an ex-

ternal chemical shift reference for ³¹P spectra. COSY spectra of

some of the diamagnetic complexes were obtained by using

2048ϫ2048 data points in F₁and F₂. Pure absorption mode EXSY

spectra were recorded by using the conventional NOESY 90°–t₁–

90°–τ_mix–90°–acq. phase-sensitive pulse sequence in D₂O as sol-

vent. The ³¹P EXSY data were acquired with a variable mixing

time between 2 and 150 ms and processed with the program Mes-

trec EXSY CALC v. 1.0.^[59]Longitudinal relaxation rates (1/T₁)

were obtained by the inversion-recovery method, and transverse

relaxation rates (1/T₂) were measured by the Carr–Purcell–Mei-

boom–Gill spin-echo technique. At each temperature, the spectral

parameters were measured for both the sample with gadolini-

um(III) complex and the sample of acidified water (pH 6) under

exactly the same conditions. No frequency lock was applied. The

1/T₁nuclear magnetic relaxation dispersion (¹H NMRD) profiles

of water protons at several temperatures (5, 25 and 37 °C) were

obtained with a Stelar Master FFC-2000 relaxometer by using the

field-cycling method and covering a continuum of magnetic fields

from 2.35ϫ10^–4to 0.35 T (corresponding to a proton Larmor fre-

quency range from 0.01 MHz to 15 MHz). The spin-lattice proton

relaxation rates at 0.47 (20) and 1.41 T (60 MHz) were measured

with a Bruker Minispec mq-20 and a Bruker Minispec mq-60 spin

analyzer, respectively. Relaxivity at different pH values (2–12) was

measured with a Bruker Minispec mq-20 at 37 °C. The Ln^IIIcon-

tents in all samples were estimated from the bulk magnetic suscep-

tibility (BMS) shift.^[60]

We thank Prof. É. Tóth (CNRS, Orléans, France) for allowing us

to measure UV/Vis spectra. This work was supported by the Grant

Agency of the Czech Republic (grant no. 203/06/0467), the Czech

Academy of Sciences (grant no. KAN201110651) and the Ministry

of Education, Youth and Sport of the Czech Republic (grant no.

MSM0021620857) and by the Foundation of Science and Technol-

ogy (F.C.T.), Portugal (project PTDC/QUI/70063/2006 and grant

SFRH/BD/9685/2002 to G. A. P.) and FEDER. This work was car-

ried out in the framework of the COST D38 Action D38 “Metal-

Based systems for Molecular Imaging Applications” and EU-FP6

“Network of Excellence” EMIL (grant no. LSHC-2004-503569)

and DiMI (grant no. LSHB-2005-512146) projects.

[1] The Chemistry of Contrast Agents in Medical Magnetic Reso-

nance Imaging (Eds.: A. E. Merbach, É. Tóth), Wiley, Chiches-

ter, England, 2001.

[2] Topics in Current Chemistry, Springer Verlag, Heidelberg, 2002,

vol. 221.

[3] P. Caravan, J. J. Ellison, T. J. McMurry, R. B. Laufer, Chem.

Rev. 1999, 99, 2293–2352.

[4] S. Aime, M. Botta, E. Terreno, Adv. Inorg. Chem. 2005, 57,

173–237.

[5] P. Hermann, J. Kotek, V. Kubícˇek, I. Lukesˇ, Dalton Trans.

2008, 3027–3047.

[6] P. Caravan, Chem. Soc. Rev. 2006, 35, 512–523.

[7] A. Nonat, C. Gateau, P. H. Fries, M. Mazzanti, Chem. Eur. J.

2006, 12, 7133–7150.

[8] A. Borel, J. F. Bean, R. B. Clarkson, L. Helm, L. Moriggi,

A. D. Sherry, M. Woods, Chem. Eur. J. 2008, 14, 2658–2667.

[9] M. Botta, Eur. J. Inorg. Chem. 2000, 399–407.

[10] F. Avecilla, J. A. Peters, C. F. G. C. Geraldes, Eur. J. Inorg.

Chem. 2003, 4179–4186.

[11] S. Aime, M. Botta, E. Terreno, P. L. Anelli, F. Uggeri, Magn.

Reson. Med. 1993, 30, 583–591.

[12] P. Caravan, M. T. Greenfield, X. Li, A. D. Sherry, Inorg. Chem.

2001, 40, 6580–6587.

[13] C. F. G. C. Geraldes, A. D. Sherry, I. Lázár, A. Miseta, P.

Bogner, E. Berenyi, B. Sumegi, G. E. Kiefer, K. McMillan, F.

Maton, R. N. Muller, Magn. Reson. Med. 1993, 30, 696–703.

[14] M. Murru, D. Parker, G. Williams, A. Beeby, J. Chem. Soc.,

Chem. Commun. 1993, 1116–1118.

Data Evaluation: Potentiometric data were evaluated with the pro-

gram package OPIUM.^[61]Relaxometric data were evaluated with

Origin^TM, Version 6.0. All calculations using relaxometric data

were carried out with least-squares fitting by the program SCIEN-

134

www.eurjic.org

Eur. J. Inorg. Chem. 2009, 119–136

Lanthanide(III) Complexes of Phosphorus Acid Analogues of H₄DOTA

ˇ

[15] J. Rohovec, P. Vojtísek, P. Hermann, J. Mosinger, Z. Zák, I.

[36] S. Aime, A. Barge, J. I. Bruce, M. Botta, J. A. K. Howard, J. M.

Moloney, D. Parker, A. S. de Sousa, M. Woods, J. Am. Chem.

Soc. 1999, 121, 5762–5771.

ˇ

Lukesˇ, J. Chem. Soc., Dalton Trans. 1999, 3585–3592.

[16]

a) S. Aime, A. S. Batsanov, M. Bota, J. A. K. Howard, D. Par-

ker, K. Senanayake, G. Williams, Inorg. Chem. 1994, 33, 4696–

4706; b) S. Aime, A. S. Batsanov, M. Botta, R. S. Dickins, S.

Faulkner, C. E. Foster, A. Harrison, J. A. K. Howard, J. M.

Moloney, T. J. Norman, D. Parker, L. Royle, J. A. G. Williams,

J. Chem. Soc., Dalton Trans. 1997, 3623–3636; R. L. Luck,

C. L. Maupin, D. Parker, J. P. Riehl, J. A. G. Williams, Inorg.

Chim. Acta 2001, 317, 331–337.

S. Aime, M. Botta, M. Fasano, M. P. M. Marques, C. F. G. C.

Geraldes, D. Pubanz, A. E. Merbach, Inorg. Chem. 1997, 36,

2059–2068.

R. L. Luck, C. L. Maupin, D. Parker, J. P. Riehl, J. A. G. Wil-

liams, Inorg. Chim. Acta 2001, 317, 331–337.

[37] I. Lukes, J. Kotek, P. Vojtísek, P. Hermann, Coord. Chem. Rev.

ˇ

2001, 216–217, 287–312.

[38] J. Rohovec, M. Kývala, P. Vojtísek, P. Hermann, I. Lukes, Eur.

ˇ

J. Inorg. Chem. 2000, 195–203.

W. D. Kim, G. E. Kiefer, J. Huskens, A. D. Sherry, Inorg. Chem.

1997, 36, 4128–4134.

J. Rohovec, I. Lukesˇ, P. Vojtísˇek, I. Císarˇová, P. Hermann, J.

Chem. Soc., Dalton Trans. 1996, 2685–2691.

I. Lázár, A. D. Sherry, R. Ramasamy, E. Brücher, R. Király,

Inorg. Chem. 1991, 30, 5016–5019.

R. Delgado, J. Costa, K. P. Guerra, L. M. P. Lima, Pure Appl.

Chem. 2005, 77, 569–579.

G. Anderegg, F. Arnaud-Neu, R. Delgado, J. Felcman, K. Po-

pov, Pure Appl. Chem. 2005, 77, 1445–1495.

A. D. Sherry, J. Ren, J. Huskens, E. Brücher, É. Tóth,

C. F. G. C. Geraldes, M. M. Castro, W. P. Cacheris, Inorg.

Chem. 1996, 35, 4604–4612.

[39]

[40]

[41]

[42]

[43]

[44]

[17]

[18]

[19]

[20]

C. F. G. C. Geraldes, A. D. Sherry, G. E. Keifer, J. Magn. Re-

son. 1992, 97, 290–304.

J. Rudovský, P. Cígler, J. Kotek, P. Hermann, P. Vojtísˇek, I.

Lukesˇ, J. A. Peters, L. Vander Elst, R. N. Muller, Chem. Eur. J.

2005, 11, 2373–2384.

[21]

[22]

a) J. Rudovský, J. Kotek, P. Hermann, I. Lukesˇ, V. Mainero, S.

Aime, Org. Biomol. Chem. 2005, 3, 112–117; b) J. Rudovský,

M. Botta, P. Hermann, A. Koridze, S. Aime, Dalton Trans.

2006, 2323–2333.

P. Lebdusˇková, P. Hermann, L. Helm, É. Tóth, J. Kotek, K.

Binnemans, J. Rudovský, I. Lukesˇ, A. E. Merbach, Dalton

Trans. 2007, 493–501.

P. Vo j t ísˇek, P. Cígler, J. Kotek, J. Rudovský, P. Hermann, I.

Lukesˇ, Inorg. Chem. 2005, 44, 5591–5599.

J. Kotek, J. Rudovský, P. Hermann, I. Lukesˇ, Inorg. Chem.

2006, 45, 3097–3102.

a) S. Aime, E. Gianolio, D. Corpillo, C. Cavallotti, G. Palmis-

ano, M. Sisti, G. B. Giovenzana, R. Pagliarin, Helv. Chim. Acta

2003, 86, 615–631; b) S. Aime, M. Botta, L. Frullano, S. Genin-

atti Crich, G. Giovenzana, R. Pagliarin, G. Palmisano, F. R.

Sirtori, M. Sisti, J. Med. Chem. 2000, 43, 4017–4024; c) S.

Aime, M. Botta, S. Geninatti Crich, G. Giovenzana, R. Paglia-

rin, M. Sisti, E. Terreno, Magn. Reson. Chem. 1998, 36, S200–

S208.

a) J. Kotek, P. Lebdusˇková, P. Hermann, L. Vander Elst, R. N.

Muller, T. Maschmeyer, I. Lukesˇ, J. A. Peters, Chem. Eur. J.

2003, 9, 5899–5915; b) S. Aime, Z. Baranyai, E. Gianolio, K.

Ramalingam, R. Swenson, R. Ranganathan, E. Brücher, S.

Aime, Contrast Media Mol. Imaging 2007, 2, 94–102.

a) P. Lebdusˇková, J. Kotek, P. Hermann, L. Vander Elst, R. N.

Muller, I. Lukesˇ, J. A. Peters, Bioconjugate Chem. 2004, 15,

881–889; b) P. Lebdusˇková, A. Sour, L. Helm, É. Tóth, J. Ko-

tek, I. Lukesˇ, A. E. Merbach, Dalton Trans. 2006, 3399–3406.

a) J. Rudovský, M. Botta, P. Hermann, K. I. Hardcastle, I.

Lukesˇ, S. Aime, Bioconjugate Chem. 2006, 17, 975–987; b) J.

Rudovský, P. Hermann, M. Botta, S. Aime, I. Lukesˇ, Chem.

Commun. 2005, 2390–2392.

[45]

[46]

A. Bianchi, L. Calabi, C. Giorgi, P. Losi, M. Palma, P. Paoli,

P. Rossi, B. Valtancoli, M. Virtuani, J. Chem. Soc., Dalton

Trans. 2000, 697–705.

a) S. Aime, M. Botta, M. Fasano, E. Terreno, Acc. Chem. Res.

1999, 32, 941–949; b) S. Aime, E. Gianolio, A. Barge, D. Kos-

takis, I. C. Plakatouras, N. Hadjiljadis, Eur. J. Inorg. Chem.

2003, 2045–2048; c) F. K. Kálmán, M. Woods, P. Caravan, P.

Jurek, M. Spiller, G. Tircsó, R. Király, E. Brücher, A. D.

Sherry, Inorg. Chem. 2007, 46, 5260–5270.

W. W. De Horrocks Jr, D. R. Sudnick, J. Am. Chem. Soc. 1979,

101, 334–340.

A. Beeby, I. M. Clarkson, R. S. Dickins, S. Faulkner, D. Parker,

L. Royle, A. S. de Sousa, J. A. G. Williams, M. Woods, J.

Chem. Soc. Perkin Trans. 2 1999, 493–503.

a) N. Graeppi, D. H. Powell, G. Laurenczy, L. Zékány, A. E.

Merbach, Inorg. Chim. Acta 1995, 235, 311–326; b) É. Tóth,

O. M. Ni Dhubhghaill, G. Besson, L. Helm, A. E. Merbach,

Magn. Reson. Chem. 1999, 37, 701–708; c) F. Yerly, F. A. Dun-

and, É. Tóth, A. Figueirinha, Z. Kovács, A. D. Sherry,

C. F. G. C. Geraldes, A. E. Merbach, Eur. J. Inorg. Chem. 2000,

1001–1006; d) F. A. Dunand, S. Aime, S. G. Crich, G. B. Giov-

enzana, A. E. Merbach, Magn. Reson. Chem. 2002, 40, 87–92;

e) M. Mato-Iglesias, C. Platas-Iglesias, K. Djanashvili, J. A.

Peters, É. Tóth, E. Balogh, R. N. Muller, L. Vander Elst, A.

de Blas, T. Rodríguez-Blas, Chem. Commun. 2005, 4729–4731.

a) S. Aime, M. Botta, D. Parker, J. A. G. Williams, J. Chem.

Soc., Dalton Trans. 1995, 2259–2266; b) S. Aime, M. Botta,

R. C. Dickins, C. L. Maupin, D. Parker, J. P. Riehl, J. A. G.

Williams, J. Chem. Soc., Dalton Trans. 1998, 881–892.

a) J. A. Peters, J. Huskens, D. J. Raber, Prog. NMR Spectrosc.

1996, 28, 283–350; b) K. Djanashvili, J. A. Peters, Contrast Me-

dia Mol. Imaging 2007, 2, 67–71.

E. Zitha-Bovens, R. N. Muller, S. Laurent, L. Vander Elst,

C. F. G. C. Geraldes, H. van Bekkum, J. A. Peters, Helv. Chim.

Acta 2005, 88, 618–632.

L. Vander Elst, A. Sessoye, S. Laurent, R. N. Muller, Helv.

Chim. Acta 2005, 88, 574–587.

D. H. Powell, O. M. N. Dhubhghaill, D. Pubanz, L. Helm, Y. S.

Lebedev, W. Schlaepfer, A. E. Merbach, J. Am. Chem. Soc.

1996, 118, 9333–9346.

Z. Otwinowski, W. Minor, HKL Denzo and Scalepack Program

Package by Nonius BV, Delft, 1997; Z. Otwinowski, W. Minor,

Methods Enzymol. 1997, 276, 307–326.

[23]

[24]

[25]

[47]

[48]

[49]

[26]

[27]

[28]

[50]

[51]

[52]

[29]

[30]

A. Borel, L. Helm, A. E. Merbach, Chem. Eur. J. 2001, 7, 600–

610.

a) E. Balogh, M. Mato-Iglesias, C. Platas-Iglesias, E. Tóth, K.

Djanashvili, J. A. Peters, A. de Blas, T. Rodríguez-Blas, Inorg.

Chem. 2006, 45, 8719–8728; b) M. Mato-Iglesias, E. Balogh, C.

Platas-Iglesias, É. Tóth, A. de Blas, T. Rodríguez-Blas, Dalton

Trans. 2006, 5404–5415.

[53]

[54]

[31]

[32]

K. Bazakas, I. Lukesˇ, J. Chem. Soc., Dalton Trans. 1995, 1133–

1137.

[55]

[56]

M. Försterová, Z. Jandurová, F. Marques, L. Gano, P. Lubal,

J. Vaneˇk, P. Hermann, I. Santos, J. Inorg. Biochem. 2008, 102,

1531–1540.

L. Burai, R. Király, I. Lázár, E. Brücher, Eur. J. Inorg. Chem.

2001, 813–820.

M. Meyer, V. Dahaoui-Ginderey, C. Lecomte, R. Guilard, Co-

ord. Chem. Rev. 1998, 178–180, 1313–1405.

I. Lázár, D. C. Hrncir, W.-D. Kim, C. E. Kiefer, A. D. Sherry,

Inorg. Chem. 1992, 31, 4422–4424.

SIR92, Program for Automatic Solution of Crystal Structures

by Direct Methods: A. Altomare, G. Cascarano, C. Giacovazzo,

A. Guagliardi, M. C. Burla, G. Polidori, M. Camalli, J. Appl.

Crystallogr. 1994, 27, 435–435.

G. M. Sheldrick, SHELXL97, Program for Crystal Structure

Refinement from Diffraction Data, University of Göttingen,

Göttingen 1997.

[33]

[34]

[35]

[57]

Eur. J. Inorg. Chem. 2009, 119–136

www.eurjic.org

135

I. Lukesˇ, J. A. Peters et al.

FULL PAPER

[58] a) P. Táborský, P. Lubal, J. Havel, J. Kotek, P. Hermann, I.

Lukesˇ, Collect. Czech. Chem. Commun. 2005, 70, 1909–1942;

b) M. Försterová, I. Svobodová, P. Lubal, P. Táborský, J. Ko-

tek, P. Hermann, I. Lukesˇ, Dalton Trans. 2007, 535–549.

[59] J. C. Cobas, M. Martín-Pastor, Universidade de Santiago

de Compostella, 2004.

[60] D. M. Corsi, C. Platas-Iglesias, H. van Bekkum, J. A. Peters,

Magn. Reson. Chem. 2001, 39, 723–726.

[61] M. Kývala, I. Lukesˇ, Chemometrics ’95, Abstract book, Pardu-

bice, Czech Republic, 1995, p. 63; full version of OPIUM pro-

gram package is available (free of charge) on http://ww w .natur.-

cuni.cz/~kyvala/opium.html.

Scientist for Windows version 2.01, Micromath Inc., Salt Lake

City, UT, 1995.

[62]

[63] HyperChem. Release 7.5 for Windows, Molecular Modeling

System, Hypercube Inc., Ontario, Canada, 2002.

Received: August 25, 2008

Published Online: November 7, 2008

136

www.eurjic.org

Eur. J. Inorg. Chem. 2009, 119–136

Article Doi

Lanthanide(III) complexes of phosphorus acid analogues of H4DOTA as model compounds for the evaluation of the second-sphere hydration

DOI: 10.1002/ejic.200800859

Source and publish data:

Authors:

Article abstract of DOI:10.1002/ejic.200800859

Full text of DOI:10.1002/ejic.200800859

Products guided by the article

R&D Labs maybe for 91987-74-5

Relevant to this article

Hot Product