doi: 10.1002/ejic.201101334lamosa/nova_nmr/rel_fct2012/art12.pdf · (2-ethoxycarbonyl)ethyl moiety...

FULL PAPER

DOI: 10.1002/ejic.201101334

Tris(phosphonomethyl) Cyclen Derivatives: Synthesis, Acid–Base Propertiesand Complexation Studies with Cu2+ and Zn2+ Ions

Luís M. P. Lima,[a] Catarina V. Esteves,[a] Rita Delgado,*[a] Petr Hermann,[b] Jan Kotek,[b]

Romana Sevcíková,[c] and Premysl Lubal*[c,d]

Keywords: Macrocyclic ligands / Copper / Thermodynamics / Kinetics / Phosphonate complexes

Three compounds that are based on cyclen and containthree methylphosphonate pendant arms � (1,4,7,10-tetraaza-cyclododecane-1,4,7-triyl)tris(methylene)triphosphonic acid(H6do3p), 3-{4,7,10-tris[(dihydroxyphosphoryl)methyl]-1,4,7,10-tetraazacyclododecan-1-yl}propanoic acid (H7do3p1pr) and[10-(3-hydroxypropyl)-1,4,7,10-tetraazacyclododecane-1,4,7-triyl]tris(methylene)triphosphonic acid (H6do3p1ol) � weresynthesized and characterized. X-ray crystal structures weredetermined for H6do3p and for the complex [Cu(H2O)6]2+-[Cu(H2O)(H4dotp)]2– of a related ligand H8dotp [H8dotp =(1,4,7,10-tetraazacyclododecane-1,4,7,10-tetrayl)tetrakis-(methylene)tetraphosphonic acid]. They show the copper(II)centre coordinated only to the four amines of the macrocycleand one water molecule in the apical position. The acid–base

Introduction

Tetraaza macrocycles are a prominent class of organicligands due to their remarkable ability to bind strongly toa variety of metal cations with diverse coordination prefer-ences, thereby taking advantage of the preorganized butsufficiently flexible nature of its skeleton. Derivatives ofcyclam (1,4,8,11-tetraazacyclotetradecane) and cyclen(1,4,7,10-tetraazacyclododecane) with N-appended coordi-nating arms are by far the most studied macrocycles of suchtype, especially those based on the structure of the tetraace-tic acid derivative of cyclen, H4dota (Scheme 1).[1–3] Thislatter compound and its tetrakis(methylphosphonic acid)analogue, H8dotp, exhibit high thermodynamic stability

[a] Instituto de Tecnologia Química e Biológica, UniversidadeNova de Lisboa,Av. da República, 2780-157 Oeiras, PortugalFax: +351-214411277E-mail: [email protected]

[b] Department of Inorganic Chemistry, Faculty of Science,Univerzita Karlova (Charles University),Hlavova 2030, 12840 Prague 2, Czech Republic

[c] Department of Chemistry, Faculty of Science, MasarykUniversity,Kotlarská 2, 61137 Brno, Czech Republic

[d] CEITEC, Masaryk University,Brno, Czech RepublicE-mail: [email protected] information for this article is available on theWWW under http://dx.doi.org/10.1002/ejic.201101334.

Eur. J. Inorg. Chem. 2012, 2533–2547 © 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 2533

properties of the three compounds were studied in aqueoussolution by potentiometry and 31P NMR spectroscopy. All li-gands exhibit very high basicity and their protonationschemes are dominated by proton relocations between thebasic sites and intramolecular hydrogen bonding. Thethermodynamic stability constants for complexes of the threeligands with Cu2+ and Zn2+ metal ions were determined bypotentiometry and exhibit very high values for the complexesof Cu2+. UV/Vis spectroscopy was used to assess the acid-assisted dissociation of the Cu2+ complexes and showed thatthe dissociation rates are faster than for the correspondingcomplex of H8dotp, whereas the [Cu(do3p1ol)]4– complex isthe most inert one in this series.

and kinetic inertness as complexes of biologically relevanttransition-metal ions such as Zn2+ and especially Cu2+.[4–6]

Thus, numerous analogues of H4dota and H8dotp havebeen developed over the past few decades for the efficientchelation of transition-metal ions, including those for po-tential medical applications. For example, complexes ofCu2+ can be used as radiopharmaceuticals for nuclear medi-cine in the form of chelates of some of its radionuclides,such as 64Cu for positron emission tomography (PET) or67Cu for radioimmunotherapy (RIT).[7–10] Such uses requirechelates with high thermodynamic stability and kinetic in-ertness as well as with fast chelate formation, and effortshave been made to design so-called bifunctional chelators(BFC) that can be conjugated to a targeting biomolecule bymeans of a suitable reactive moiety on the chelator.[7,8,11–13]

One of the methods for improving the properties of cy-clen-based ligands is appending a different number of pen-dant arms that contain carboxylate and phosphonategroups in the macrocyclic skeleton and, more recently, dif-ferent combinations of both types of groups. Analogues ofH4dota that contain an increasing number of acetate pen-dant arms replaced by methylphosphonate ones, such asone in H5do3a1p,[14–17] two in trans-H6do2a2p[18–20] andthree in H7do3p1a,[21,22] have been developed and studied.A cyclen derivative with only three methylphosphonate pen-dant arms, H6do3p, has also been previously disclosed andbriefly studied; it has been shown to form very stable com-

R. Delgado, P. Lubal et al.FULL PAPER

Scheme 1. Structures of the compounds discussed in this work.

plexes with Cu2+[23] and also stable complexes with Zn2+

and Fe3+.[24] In this work we report on the synthesis andstudy of two novel compounds, H7do3p1pr and H6do3p1ol,as well as their precursor compound, H6do3p, in which thenew compounds have an additional functional group ap-pended at the secondary amine of the precursor H6do3p.The thermodynamic and kinetic properties for complexesof the ligands with Cu2+ and Zn2+ are described.

Results and Discussion

Synthesis of the Macrocyclic Ligands

The compounds H6do3p, H7do3p1pr and H6do3p1olwere synthesized, the latter two for the first time in thiswork, by a parallel methodology starting from cyclen asoutlined in Scheme 2. Initially, mono-N-alkylation of cyclenwas performed by adaptation of a selective reaction pro-cedure from the literature.[25] This method required the useof a four- to fivefold excess amount of cyclen, which couldbe nonetheless easily recovered by chloroform extraction ofthe basified aqueous phases that originate from the reactionworkup. In the case of H6do3p1ol, the reactive alcoholfunction of the commercially available alkylation reactantwas previously protected with an acetyl group. Next, methyl-phosphonic ester groups were appended by a Mannich-type reaction with paraformaldehyde and triethyl phosphiteunder anhydrous conditions. Finally, hydrolysis of the carb-oxylate and/or methylphosphate esters in strongly acidicmedia yielded the final compounds, but it was necessary toadopt different reaction and workup conditions depending

www.eurjic.org © 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Eur. J. Inorg. Chem. 2012, 2533–25472534

on the target compound. Indeed, N-(2-carboxyethyl) moie-ties are known to be partially eliminated under strong acidand/or strong heating by a retro-Michael reaction mecha-nism.[26] In our case, this dealkylation occurred to a variableextent whenever any of the compounds that contained the(2-ethoxycarbonyl)ethyl moiety (1 and 2) were treated withconcentrated acid under heating, but such side reactionscould be avoided as long as that ester moiety was first hy-drolyzed at room temperature to the corresponding acid.Thus, after careful isolation of the fully esterified precursorof H7do3p1pr (2), its hydrolysis in 20% aqueous HCl firstat room temp. with subsequent heating to a gentle refluxafforded H7do3p1pr without noticeable occurrence of de-alkylation. Furthermore, and by taking advantage of thementioned reactivity features, we could obtain the pureH6do3p by hydrolysis (and simultaneous dealkylation)of the same precursor (i.e., the ester of H7do3p1pr, 2) in20% aqueous HBr under prolonged reflux. To obtainH6do3p1ol, the hydrolysis was performed in 20 % aqueousHCl under reflux, in which the ester hydrolysis was ac-companied by a partial formation of a N-(3-chloropropyl)derivative. Thus, the chloro derivative had to be reverted tothe target H6do3p1ol by a final reflux in diluted aqueousNaOH solution.

Purification of all final compounds was done by cationic-exchange chromatography. For H6do3p and H6do3p1ol, thecrude products from the hydrolysis step were first loaded ina strong cationic exchanger, nonaminic impurities wereeluted with water and the macrocycles were then eluted withaqueous ammonia in the form of ammonium salts. The am-monium salts of H6do3p and H6do3p1ol, or the crudeproduct that contained H7do3p1pr from the acidic hydroly-

Tris(phosphonomethyl) Cyclen Derivatives

Scheme 2. Synthetic route for the compounds.

sis step, were loaded in a weak cationic exchanger andeluted with water, which isolated the relevant compoundsin zwitterionic form free from other macrocyclic byproductsthat were more retained on the column. Finally, precipi-tation from a concentrated aqueous solution yieldedH6do3p as a stable solid, whereas evaporation and vacuumdrying of aqueous solutions yielded H6do3p1ol as a slightlyhygroscopic solid and H7do3p1pr as a very hygroscopic so-lid.

X-ray Crystal Structure of H6do3p

Single crystals were obtained in the form of a non-stoichiometric hydrate of formula H6do3p·3.8H2O by slowevaporation of a saturated aqueous solution, and the solv-ate water content was determined on the basis of X-rayanalysis. A similar crystal structure of the compound hasbeen previously reported.[24] However, our structure showsdifferent cell dimensions and belongs to a different spacegroup, besides having significantly better refinement param-eters. Additionally, the protonation pattern of the reportedstructure is also slightly different, although that can be jus-

Eur. J. Inorg. Chem. 2012, 2533–2547 © 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjic.org 2535

tified by the fact that all protons are involved in strong hy-drogen-bonding interactions in both structures.

In the crystal structure obtained, shown in Figure 1, thenitrogen atom of the secondary amine N10 and the transan-nular amine with a methylphosphonate pendant arm N4are both protonated. The remaining four protons are dis-tributed on the oxygen atoms of the pendant arms, withone proton in each of the phosphonate moieties appendedto nonprotonated amines N1 and N7, and the two remain-ing protons in the phosphonate moiety appended at theprotonated amine N4. The four nitrogen atoms form aplane and the whole macrocyclic ring is in a square(3,3,3,3)-B conformation according to the Dale formal-ism,[27] which is the most usual one for twice-protonatedsubstituted cyclen rings. This macrocyclic conformation isstabilized by a system of intramolecular N–H···N hydrogenbonds with N···N distances in the range 2.88–3.05 Å andN–H···N angles in the range 101–113°. All three pendantarms are pointing in the same direction with respect to themacrocyclic plane; the two pendant arms from amines N1and N7 are folded above the macrocyclic cavity, whereas theremaining pendant arm at N4 is turned away; an analogous

R. Delgado, P. Lubal et al.FULL PAPERarrangement was found in the literature for the crystalstructure of H8dotp.[28] The molecular conformation is sta-bilized by strong cooperative intramolecular hydrogenbonds that involve one protonated tertiary amino groupand two oxygen atoms from the pendant phosphonategroups, N4–H···O33 with a N···O distance of 2.81 Å andO11–H···O33 with an O···O distance of 2.54 Å, respectively.The ligand molecules are connected by means of strong hy-drogen bonds between protonated and unprotonated phos-phonate oxygen atoms with O···O distances in the range2.46–2.60 Å. The solvate water molecules cross-link thestructure with a system of medium-strong hydrogen bonds

Figure 1. Molecular structure of H6do3p found in the crystal struc-ture of H6do3p·3.8H2O. Hydrogen atoms bound to carbon atomsare omitted for the sake of clarity.

Figure 2. The molecular structure of the [Cu(H2O)(H4dotp)]2– anion found in the solid-state structure of [Cu(H2O)6]2+-[Cu(H2O)(H4dotp)]2–. Hydrogen atoms bound to carbon atoms are omitted for the sake of clarity.


between water hydrogen and the phosphonate oxygenatoms with O···O distances in the range of 2.8–2.9 Å (seeFigure S1 in the Supporting Information).

X-ray Crystal Structure of [Cu(H2O)6]2+[Cu(H2O)-(H4dotp)]2–

Single crystals of [Cu(H2O)6]2+[Cu(H2O)(H4dotp)]2–

were obtained randomly during crystallization attempts atlow pH with a nonpurified H6do3p ligand batch that con-tained H8dotp as impurity. In the solid-state structure ofthe complex, the independent unit is formed by one half ofthe formula unit, with molecules of the macrocyclic com-plex anion and of a hexaaqua copper(II) cation lyingon the twofold symmetry axes. The central Cu atom in the[Cu(H2O)(H4dotp)]2– anion is coordinated to four macro-cyclic nitrogen atoms (dCu–N ≈ 2.03 Å) to form a planarbase, and to a water molecule in a distant apical position(dCu–O ≈ 2.27 Å). The central Cu atom lies 0.455(3) Å abovethe N4 plane, whereas all pendant arms are monoproton-ated and uncoordinated. Both transannular N–Cu–N#

angles are very similar (≈154°) and, as all N–Cu–O1Wangles are also very similar (ca. 103°), the coordinationsphere has an almost regular square-pyramidal geometry.This structural motif is frequently found in Cu2+ complexesof cyclen derivatives. The Cu–N distances in the title com-plex are very similar to those in the literature (nitrogenatoms substituted with Me,[29] Bn[30] or propionitrile[31]),but the Cu–Oax distance is significantly longer here than inthe other complexes (2.126–2.156 Å) and the copper atomis located closer to the N4 plane here than in its analogues(0.493–0.499 Å). This structural data correlates with thehigher basicity of nitrogen atoms (and thus to stronger


binding to the central ion) of the (H4dotp)4– anion. Thestructure of the [Cu(H2O)6]2+ cation is six-coordinate, withapical virtual shortening (dCu–O ≈ 1.99 Å compared to2.15 Å observed in the equatorial base) due to the li-brationally disordered Jahn–Teller structure.[32] The wholestructure is stabilized by an extensive network of intermo-lecular hydrogen bonds, with very strong interaction be-tween protonated and unprotonated phosphonate oxygenatoms from neighbouring units (dO···O ≈ 2.55 Å). Selectedgeometric parameters are listed in Table 1 and the molecu-lar structure of the macrocyclic complex anion is shown inFigure 2. A related structure was found for a complex ofH8dotp with Ni2+,[33] but in that case with two oxygenatoms of opposite pendant arms cis-coordinated to the Niatom to give octahedral geometry.

Table 1. Selected geometric parameters found in the crystal struc-ture of [Cu(H2O)6]2+[Cu(H2O)(H4dotp)]2–. The hash symbol (#)denotes a symmetrically associated atom (twofold axis).

[Cu(H2O)(H4dotp)]2– [Cu(H2O)6]2+

Distances [Å]

Cu1–N1 2.038(4) Cu2–O1B 1.985(4)Cu1–N4 2.021(4) Cu2–O2B 2.152(5)Cu1–O1A 2.265(6) Cu2–O3B 2.153(4)

Angles [°]

N1–Cu1–N4 87.5(2) O1B–Cu2–O2B 89.7(2)N1–Cu1–N1# 154.3(2) O1B–Cu2–O3B 91.0(2)N1–Cu1–N4# 86.8(2) O1B–Cu2–O1B# 178.6(2)N1–Cu1–O1A 102.8(1) O1B–Cu2–O2B# 89.3(2)N4–Cu1–N4# 153.9(2) O1B–Cu2–O3B# 90.1(2)N4–Cu1–O1A 103.1(1) O2B–Cu2–O3B 90.2(1)– – O2B–Cu2–O2B# 90.0(2)– – O2B–Cu2–O3B# 179.7(2)– – O3B–Cu2–O3B# 89.6(2)

Acid–Base Behaviour

The acid–base properties of H7do3p1pr, H6do3p andH6do3p1ol were studied by potentiometric and 31P NMRspectrometric titrations in aqueous solution. The overallprotonation constants obtained are collected in Table S1 ofthe Supporting Information. Calculated stepwise constantsare presented in Table 2, together with literature values

Table 2. Stepwise protonation constants (log KiH) for the discussed ligands in aqueous solution; T = 298.2 K; I = 0.10 m in

[N(CH3)4]NO3.

Equilibrium quotient[a] H7do3p1pr H6do3p H6do3p1ol H4dota H8dotp

[HL]/[L][H] 14.0[b] 13.9[b], 12.9,[c] 13.24[d] 13.9[b] 12.09[e] 14.65[f]

[H2L]/[LH][H] 11.8[b] 11.7[b], 11.4,[c] 11.3[d] 11.9[b] 9.76[e] 12.40[f]

[H3L]/[LH2][H] 9.00 9.02, 8.69,[c] 8.47[d] 8.65 4.56[e] 9.28[f]

[H4L]/[LH3][H] 7.64 7.40, 7.09,[c] 7.1[d] 7.19 4.09[e] 8.09[f]

[H5L]/[LH4][H] 6.06 5.51, 5.53,[c] 5.3[d] 5.39 – 6.12[f]

[H6L]/[LH5][H] 4.64 2.04, 1.42[c] 1.90 – 5.22[f]

[H7L]/[LH6][H] 2.04 – – – –[H4L]/[L][H]4 42.44 42.02 41.63 30.50[e] 44.42[f]

[a] Charges of equilibrium species are omitted for clarity. [b] This work, determined by 31P NMR spectroscopy and corrected by theDavies equation. [c] Ref.[23] with I = 0.1 m in KCl. [d] Ref.[24] with I = 0.1 m in (NBu4)NO3. [e] Ref.[34] [f] Ref.[35]


for H6do3p[23,24] and related H4dota[34] and H8dotp.[35]

Four protonation constants were found for H6do3p andH6do3p1ol and five for H7do3p1pr, as the remaining basiccentres have constants that fall outside the range of logKa

= 2–11.5 that can be accurately determined by potentiome-try. The trend of tetraazamacrocycles with N-methylphos-phonate pendant arms to exhibit very high protonationconstants for the first (one or two) ring amine(s) is wellknown.[5,16,18,19,34–41] It has been justified by two differentphenomena: one is the effect of spreading the high electrondensity of the fully deprotonated phosphonate group to thenearest nitrogen atom;[5] another is the formation of strongintramolecular hydrogen bonds between a protonatedamine and a neighbouring oxygen atom from a phos-phonate moiety or between protonated and nonprotonatedring amines.[36] For this reason, the highest protonationconstants of all three ligands were determined from “in-cell” 31P NMR spectroscopic titrations in aqueous solutionin the pH 9–14 region.

The two highest protonation constants are quite similarfor the studied ligands, being somewhat lower than thosefound for H8dotp. It was not possible to determine the pro-tonation constant for the propanol moiety of H6do3p1ol,thus indicating that such deprotonation occurs above pH14 as expected for a primary alcohol. The following threeprotonation constants of each compound are common forphosphonate moieties and show a trend of decreasing val-ues from H7do3p1pr to H6do3p1ol. Next, H7do3p1pr pres-ents a constant for the carboxylate somewhat higher thancould be expected for such functionality appended on aprotonated amine. Finally, the lowest constants for eachcompound can be associated with further protonation ontheir phosphonate moieties. All compounds have very highoverall basicity that increases in the order H6do3p1ol �H6do3p � H7do3p1pr. For H6do3p, the protonation con-stants now determined compare reasonably well with thosepreviously reported,[23,24] considering the different experi-mental conditions used. Species distribution diagrams forthe three studied ligands are shown in Figures S2–S4 of theSupporting Information.

Additional 31P NMR spectroscopic titrations of eachcompound were performed in aqueous solution in the pH–1 to 9 range. This allowed plotting of the pH dependenceof 31P NMR spectroscopic chemical shifts throughout

R. Delgado, P. Lubal et al.FULL PAPERthe whole pH range (see Figure 3 and Figure S5 in the Sup-porting Information) to tentatively assign protonationsites of the compounds. Data reported for otherpolyazamacrocycles with methylphosphonate pen-dants[16,18,19,21,24,36,37,39,41–44] were used here as the basis ofthe following discussion.

Figure 3. 31P NMR spectroscopic titration of H7do3p1pr (top) andH6do3p1ol (bottom). For the resonance assignment, see Schemes 3and 4.

For H7do3p1pr (Scheme 3), starting from the fully de-protonated species (do3p1pr)7–, there is a significant upfieldshift of δa whereas δb is mostly unchanged (pH � 12). Thisclearly indicates that protonation starts at the amine N7.Next, there is an upfield shift of δb and also of δa (12 � pH� 10), but these shifts are smaller than those expected forthe protonation of an amine with an appended methylphos-phonate moiety. This concurrent change can be understoodby admitting the protonation of amine N1 (due to electro-static repulsion) with formation of hydrogen bonds betweenthe protonated amines and the remaining ones (N4 andN10), thereby delocalizing the positive charge throughoutall the ring amines. Next, there is a strong downfield shiftof δb and a smaller upfield shift of δa (10 � pH � 6.5), in


a region in which only protonation of the phosphonategroups is expected. Protonation of the phosphonates “b” istherefore undisputed, but the shift of δa is indicative ofmore protonation of amine N7, something that can be ex-plained by the dissociation of the hydrogen bonds in thering, which leads to an increase of the positive charge onN7 and also N1. Next there is a small downfield shift of δa

and still of δb (6.5 � pH � 4.5), which must then indicatethe protonation of phosphonate “a”. After this, there is aregion in which both δa and δb remain unchanged (4.5 �pH � 2) that can correspond to the protonation of thecarboxylate group, which is not expected to influence thephosphorus signals due to the separation between the carb-oxylate and the phosphonate moieties. The final downfieldshift of both δa and δb (pH � 2) indicates further proton-ation of all phosphonate groups.

For H6do3p (Scheme S1 of the Supporting Information),the pH dependence of the 31P NMR spectroscopic chemicalshifts is similar to that of H7do3p1pr throughout the wholepH range, with only two exceptions. At first, the absence ofa carboxylate moiety in H6do3p, which, in any case, is notreflected in the experimental data for the reason pointedabove. Then, at the most acidic range (pH � –1) there is asmall upfield shift of δb combined with a downfield shift ofδa, which suggests that protonation of the remaining aminesof the ring (N4 and N10) could start simultaneously withthe end of protonation of the phosphonate groups. In thiscase, the proposed protonation scheme is in agreement withthe protonation mode in the crystal structure for the zwit-terionic species of H6do3p found in this work and reportedelsewhere.[24]

For H6do3p1ol (Scheme 4), which starts from the depro-tonated species (do3p1ol)6–, there is an upfield shift of δa

and in a smaller degree also of δb (pH � 12). This seems toindicate the protonation of amine N7 with probable hydro-gen bonds to N4 and N10. Next, there is a strong upfieldshift of δb and a minor one of δa (12 � pH � 10), whichcould indicate that the protonation of a second amine inthe ring forces the relocation of the protons to N4 and N10whereas hydrogen bonds should now involve N7. It is likelythat N1 is, at this point, also involved in hydrogen bondingwith N4 and N10, although it cannot be proven from thesedata. Next there is a very substantial downfield shift of δb

and a noticeable upfield shift of δa (10 � pH � 6.5). Thesesignificant and opposing shifts in a region in which onlyprotonation of phosphonate moieties is expected can onlybe understood if a new relocation of protons on the ringamines is considered. Thus, the downfield shift of δb shouldbe explained by protonation of the phosphonate groups “b”and simultaneous deprotonation of N4 and N10, whereasthe upfield shift of δa should be explained by the reproton-ation of N7. As a consequence of the redistribution of ringprotons, N1 would also become protonated at this point. Inthe pH � 6.5 range, the change of the chemical shifts ofH6do3p1ol is very similar to the one seen for the H6do3p,so the same assignments are proposed. In particular, at themost acidic range (pH � –1) there is a more visible upfieldshift of δb combined with a downfield shift of δa, which


Scheme 3. Tentative protonation sequence for H7do3p1pr.

Scheme 4. Tentative protonation sequence for H6do3p1ol.

again points to the protonation of the remaining amines ofthe ring (N4 and N10) that starts simultaneously with theend of protonation of the phosphonate groups.

Thermodynamic Stability of Metal Complexes

The complexes of H7do3p1pr, H6do3p and H6do3p1olwith Cu2+ and Zn2+ ions were studied by potentiometrictitrations in aqueous solution with ionic strength kept at0.10 m in [N(CH3)4]NO3. Equilibria were attained quicklyin all cases, thus allowing the study of these Cu2+ and Zn2+

complexes by direct “in-cell” titrations. The stability con-stants obtained are collected in Table S2 of the SupportingInformation. The calculated stepwise stability constants forthe complexes of the studied ligands, together with litera-ture values for the related H4dota and H8dotp,[23,34,40,45–48]

are presented in Table 3. Only mononuclear complex spe-cies were found to be present under our experimental condi-tions. Dinuclear metal complexes are likely to be formed in


the presence of more than one equivalent of metal ions, ashas been found for complexes of similar ligands,[16,18,19] butthe stability constants of such species are generally low.

The complexes of all the studied ligands with Cu2+ havevery high stability constants, with the highest one beingfound for H7do3p1pr. It is noteworthy that the higherlogKML values correlate well with the increase of overallbasicity of the ligands due to the presence of severalaminomethylphosphonate groups. The stability constantsfor the studied complexes are much higher than the onereported for the H4dota complex, but are surprisinglyhigher than that for the H8dotp complex. The latter factcan only be explained by a possible underestimation of thestability constant reported for H8dotp, which is suggestedby the noticeable underestimation of the two highest pro-tonation constants of this ligand.[47] The complexes of Zn2+

present stability constants rather similar for all ligands,which are slightly higher than that reported for H4dota butlower than that for H8dotp. Additionally, the constants for

R. Delgado, P. Lubal et al.FULL PAPERTable 3. Stepwise stability constants (log KMHiL) for complexes of the discussed ligands in aqueous solution; T = 298.2 K; I = 0.10 m in[N(CH3)4]NO3.

Ion Equilibrium quotient[a] H7do3p1pr H6do3p H6do3p1ol H4dota H8dotp

Cu2+ [ML]/[M][L] 28.77 28.43, 26.9[b] 27.86 22.25[c] 25.4[d]

[MHL]/[ML][H] 7.57 7.45 7.47 3.78[c] 7.41[d]

[MH2L]/[MHL][H] 6.88 6.45 6.39 3.70[c] 6.42[d]

[MH3L]/[MH2L][H] 5.38 4.51 4.66 – 6.16[d]

[MH4L]/[MH3L][H] 4.19 – – – 4.58[d]

[ML]/[MLOH][H] 12.09 11.84 11.62 – –

Zn2+ [ML]/[M][L] 22.72 22.34, 21.23[e] 22.27 21.10[c] 24.8[d]

[MHL]/[ML][H] 8.08 8.17, 7.8[e] 8.21 4.18[c] 7.21[d]

[MH2L]/[MHL][H] 7.05 6.71, 6.5[e] 6.65 3.51[c] 6.72[d]

[MH3L]/[MH2L][H] 5.80 5.42, 5.1[e] 5.23 – 5.06[d]

[MH4L]/[MH3L][H] 4.71 – 3.38 – 4.78[d]

[ML]/[MLOH][H] 11.93 12.11 11.79 – –

[a] Charges of equilibrium species are omitted for clarity. [b] Ref.[23] with I = 0.1 m in KCl. [c] Ref.[34] [d] Ref.[47] with I = 1.0 m in KNO3.[e] Ref.[24] with I = 0.1 m in (NBu4)NO3.

the Zn2+ complexes are 5–6 orders of magnitude lower thanthose for the Cu2+ ones, thereby indicating a good degreeof selectivity for the latter cation. On the other side, thestability constants found for the protonated species of bothcomplexes with Cu2+ and Zn2+ seem to indicate that pro-tonation happens only on the pendant arms, as the valuesfor the protonation constants of the MHL and MH2L spe-cies are lower than the first two protonation constants as-signed to the phosphonate moieties on the correspondingfree ligands. Similar features are observed for H8dotp, andthe same effect is verified in the solid state as demonstratedabove by the structure of [Cu(H2O)(H4dotp)]2– in Figure 2.This is understandable in view of the known preference ofthe metal cations for amine donors. The stability constantsobtained for the complexes of H6do3p also compare wellwith those previously reported.[23,24]

The pM values (–log [M]) were calculated for the com-plexes of all discussed ligands at pH 7.4 (physiological),from the thermodynamic constants in Tables S1 and S2 inthe Supporting Information or from literature ones, and arepresented in Table 4. These values are the correct way toevaluate the complexation efficiency of ligands that exhibitdifferent basicity. Overall, the pM values confirm that allthree studied ligands are very efficient chelators for bothmetal cations. Still, their efficiency to chelate Cu2+ is re-markably high, even higher than that of H4dota andH8dotp. In contrast, the chelation efficiency for Zn2+ islower than that of the latter ligands. However, the compari-son of our results with the ones available in the literaturefor H4dota and H8dotp must be taken cautiously, as the

Table 4. The pM[a] values calculated for metal complexes of thediscussed ligands at pH 7.4.

Ion H7do3p1pr H6do3p H6do3p1ol H4dota H8dotp

Cu2+ 16.18 16.05 15.74 15.20[b] 15.57[c]

Zn2+ 10.57 10.52 10.73 14.05[b] 14.89[c]

[a] Values calculated for 100% molar excess amount of ligand overmetal ion with cM = 1.00�10–5 m, based on the protonation andstability constants of Tables S1 and S2 in the Supporting Infor-mation or on literature values. [b] Ref.[34] [c] Ref.[47] with I = 1.0 min KNO3.


experimental methods and procedures are sometimes signif-icantly different from ours, thus entailing non-negligible in-consistencies. Globally, H7do3p1pr is the most efficientamong the studied chelators for both metal cations, al-though there are only slight differences between them.

Examples of distribution diagrams for the complexes ofH7do3p1pr with Cu2+ and Zn2+ are shown in Figure 4,

Figure 4. Species distribution diagrams for the complexes ofH7do3p1pr with Cu2+ (top) and Zn2+ (bottom), at cM = cL =1.0�10–3 m. Charges of species were omitted for simplicity.


whereas diagrams for the other systems are quite similar.These diagrams show several protonated species that existthroughout the whole pH range, as expected from the highnumber of protonation equilibria allowed by the severalfunctional groups present on the ligands, notably the phos-phonate moieties. It can be seen that complexes exist atphysiological pH mainly as unprotonated, monoprotonatedand diprotonated species, whereas the complexes are fullyformed from below neutral pH for Zn2+ or even below pH4 for Cu2+.

Dissociation Kinetics Study of the Cu2+ Complexes

Kinetic inertness is commonly considered the decisiveparameter for the in vivo stability of metal complexes usedin medicine or molecular biology.[49,50] The measurement ofkinetics of the acid-assisted dissociation of metal complexesis a commonly used way of determining the kinetic inert-ness of the complexes. It has been shown that rates of disso-ciation for complexes of Cu2+ with macrocyclic ligands canspan over several orders of magnitude.[51] Thus, we decidedto follow the dissociation kinetics of the investigated Cu2+

complexes in detail.The UV/Vis spectra of Cu2+ complexes of the studied

ligands in aqueous solution exhibit absorption band max-ima at the 320–330 nm range in the UV region and at620 nm in the visible region (Figures S6 and S7 in the Sup-porting Information), thus indicating that these complexeshave the same structure. The acid-assisted dissociation ofthe Cu2+ complexes was studied in the presence of high HClconcentrations (see examples in Figure 5 and Figure 6), andit was found that the Cu2+ complex of H6do3p1ol is morekinetically inert than those of H6do3p and H7do3p1pr.There is no shift in the absorption maximum of the com-plexes in both UV and visible regions (Figures S8–S12 in

Figure 5. Comparison of the pseudo-first-order dissociation rateconstants for Cu2+ complexes of the studied ligands at 298.2 K.The lines represent fitting according to Equation (1).


the Supporting Information) during the time course of thedissociation reaction and there is only one isosbestic pointin the UV range, thereby indicating that there is no detect-able kinetic intermediate under our experimental condi-tions.

Figure 6. Dependence of the pseudo-first-order rate constants forthe dissociation of the Cu2+ complex of H6do3p1ol as a functionof acidity at different temperatures. The experimental points werefitted by Equation (1) with the parameters given in Table 5.

Assuming the reaction mechanism proposed inScheme 5, the following empirical rate law [Equation (1)]was derived.[52]

Scheme 5. Proposed reaction mechanism for acid-assisted dissoci-ation of the Cu2+ complexes of H6do3p, H6do3p1ol (m = 3, n = 4)and H7do3p1pr (m = 4, n = 5).

The measured pseudo-first-order rate constants at dif-ferent solution acidities, kd,obs, were fitted according toEquation (1) and the results are presented in Table 5. It isinteresting to note that the dependences illustrated in Fig-ures 5 and 6 can be considered a composition of two simpli-fied limiting cases [Equations (2a) and (2b), see below].

These two equations were also used for a fitting of somemeasurements. Activation parameters estimated from thedata obtained at different temperatures are also presentedin Table 5.

In general, the parameters are rather similar for all thecomplexes, thereby suggesting that the reaction mechanism

R. Delgado, P. Lubal et al.FULL PAPERTable 5. Kinetic parameters[a] for acid-assisted dissociation of thecomplexes of Cu2+; I = 3.0 m (Na,H)Cl.

Temp. [K] Rate constants Protonation

k1�103 k2�103 k2/k1 K[s–1] [m–1s–1] [m–1]

H6do3p

288.2[b] 0.75(2) 0.525(9) 0.70(2) n.d.293.2[b] 1.2(1) 0.79(8) 0.7(1) n.d.298.2 1.9(1) 1.12(9) 0.59(6) 10(3)308.2 6.3(3) 2.9(3) 0.46(6) 8(1)318.2 14.2(5) 6.9(5) 0.48(4) 10(1)ΔH� [kJmol–1] 75(3) 63(2) – –ΔS� [Jmol–1K–1] –47(9) –88(8) – –

H6do3p1ol

293.2[b] 0.466(3) 0.172(1) 0.369(3) n.d.298.2 0.91(2) 0.23(2) 0.25(2) 16(2)303.2[b] 1.6(1) 0.34(5) 0.21(3) n.d.308.2 2.81(9) 0.62(4) 0.22(2) 16(2)318.2 7.8(2) 1.9(2) 0.24(2) 16(2)328.2 15.4(3) 5.1(2) 0.33(1) 37(5)ΔH� [kJmol–1] 78(3) 78(5) – –ΔS� [Jmol–1K–1] –42(10) –55(15) – –

H7do3p1pr

288.2 0.69(4) 0.31(3) 0.46(5) 7(1)298.2 1.9(1) 0.9(1) 0.47(8) 8(2)308.2 4.9(2) 2.4(2) 0.49(5) 11(2)318.2 13.2(6) 6.2(1) 0.47(2) 11(2)328.2 27.3(7) 11.8(5) 0.43(2) 13(1)ΔH� [kJmol–1] 70(1) 70(2) – –ΔS� [Jmol–1K–1] –61(4) –70(7) – –

[a] The standard deviations are given in parentheses. [b] Data at thetemperatures were fitted according to Equation (2a).

of the dissociation reactions should be similar and shouldcorrespond to Scheme 5. The species with one proton oneach acidic pendant arm (phosphonate and carboxylate)are thermodynamically stable (Table 3) and their struc-tures are expected to be analogous to that of the[Cu(H2O)(H4dotp)]2– complex (Figure 2). The next proton-ation induces the complex dissociation and should takeplace on a phosphonate group.[52–54] Protonation constantvalues for the kinetically active species of the investigatedcomplexes, log K ≈ 1, are similar to those for kineticallyactive species of Cu2+ complexes of other phosphonate-con-taining macrocyclic ligands.[52,54] A transfer of this proton(constant k1) to a ring nitrogen atom and/or a direct attackto nitrogen atom(s) (constant k2) lead to complex decompo-sition.

Both k1 and k2 rate constants for the complex ofH6do3p1ol are systematically 2–4 times lower than thosefor the complexes of H6do3p and H7do3p1pr (Table 5 andalso illustrated in Figure 5). The higher kinetic inertness ofthe former complex in solutions with higher acidities isgiven mainly by the lower k2 values (k2/k1 ratio in Table 5).It suggests that the proton transfer from protonated pen-dant arms or the direct proton attack to ring amines mighthave a different efficiency in these complexes; the presenceof a more basic secondary amine (H6do3p) or distant acidicgroup on a pendant arm (H7do3p1pr) accelerates the trans-


fer and/or direct nitrogen attack if compared with the com-plex of ligand with the distant alcohol group (H6do3p1ol).This can be caused by a different intramolecular hydrogen-bond network commonly supposed to be present in com-pounds with protonated phosphonate groups. Water mole-cule(s) that originate from the bulk solvent can also partici-pate significantly in the system as the values of the acti-vation parameters point to a strongly associative mecha-nism. Similar values of the activation parameters were de-termined for Cu2+ complexes of other phosphonate-con-taining macrocycles,[52,54] thus indicating that the complexdissociation mechanisms can be analogous.

As the Cu2+ complexes of cyclen and its derivatives aresupposed to have similar structures in acidic solutions(tetragonal pyramidal arrangement with N4 base plane andwater molecule in the apical position), their kinetic inert-ness can be directly compared. A practical comparison ofthe kinetic inertness of the complexes (Table 6), by meansof their half-life values and in spite of the different experi-mental conditions, suggests that complexes of the title li-gands are less kinetically inert than the Cu2+ complex ofH4dota, somewhat less inert than the complex of cyclen andcomparably inert to the complex of H8dotp. Nonetheless,as the in vivo stability of such complexes does not alwaysfollow the trend of acid-assisted dissociation mechanisms,the kinetic inertness of the reported complexes might stillbe suitable for possible applications in vivo.

Table 6. Comparison of kinetic inertness of the Cu2+ complexes ofcyclen-based ligands in acidic medium ([H+] = 1.0 m, 298.2 K).

Ligand Ionic medium Half-life (τ1/2) Reference[min]

H6do3p 3.0 m (Na,H)Cl 3.8 this workH6do3p1ol 3.0 m (Na,H)Cl 10.1 this workH7do3p1pr 3.0 m (Na,H)Cl 4.1 this workcyclen 5.0 m (Na,H)ClO4 54.2 [55]

5.0 m (Na,H)NO3 16.5 [56]

H4dota 5.0 m (Na,H)ClO4 1914 [57]

H8dotp 5.0 m (Na,H)ClO4 2.1 [57]

Conclusion

Three macrocyclic compounds based on cyclen and thatcontain three methylphosphonate pendant arms � H6do3p,H7do3p1pr and H6do3p1ol � were synthesized and charac-terized, the latter two for the first time. All three com-pounds show a very high basicity, as expected, due to thepresence of three methylphosphonate pendant arms. Theirprotonation sequence could be tentatively assigned from thepH dependence of the 31P NMR spectroscopic chemicalshifts, which proved the occurrence of proton relocationsbetween the various basic sites and also suggested the pres-ence of strong hydrogen-bonding interactions between ringamines even at very high pH. All ligands were shown toform complexes of Zn2+ with high stability, whereas com-plexes of Cu2+ notably exhibit a very high stability. Thetrend in the increase of the complex stability constants forboth metal ions follows the trend in the increase of ligand


basicity. The kinetic study of acid-assisted dissociation ofthe complexes of Cu2+ showed that the [Cu(do3p1ol)]4–

complex is more kinetically inert than the complexes withthe other two ligands. This seems to indicate that the prop-anol group of H6do3p1ol is less involved in the key pro-ton-transfer step of the dissociation reaction relative to the[Cu(do3p1pr)]5– complex that has the propionate group. Onthe other hand, the kinetic inertness of the [Cu(do3p)]4–

complex is decreased due to the presence of the secondaryamine. Although the pendant arms are probably not coordi-nated to the central metal ion in these complexes, their na-ture can influence the kinetic behaviour of the complexes.Thus, it is feasible to use the hydroxy function ofH6do3p1ol, through a suitable chemical modification, inthe development of bifunctional chelators for the efficientcomplexation of Cu2+ ions.

Experimental SectionGeneral: Cyclen (1,4,7,10-tetraazacyclododecane) was obtainedfrom CheMatech; paraformaldehyde, ethyl 3-bromopropionate andtriethyl phosphite were obtained from Sigma–Aldrich; and 3-chlo-ropropan-1-ol and acetyl chloride were obtained from Merck. Allreagents obtained from commercial sources were used as received.Organic solvents were dried by standard methods.[58] Elementalanalyses and electrospray mass spectra (ESI-MS) were performedby the Analytical Services Unit of ITQB-UNL/IBET. The 1H,13C{1H} and 31P{1H} NMR spectra for the ligand characterizationand the pH titration studies were recorded with a Bruker AvanceIII 400 spectrometer (1H at 400.14 MHz, 13C at 100.61 MHz and31P at 161.99 MHz) at a probe temperature of 298.2 K. Chemicalshifts (δ) are given in ppm and coupling constants (J) in Hz. Tet-ramethylsilane (TMS) or tert-butyl alcohol were used as internalreferences for 1H and 13C NMR spectra in CDCl3 or D2O, respec-tively, whereas H3PO4 was used as external reference (inner-capil-lary method) for 31P NMR spectra. The resonance assignments arebased on peak integration and multiplicity, and on 2D homo- andheteronuclear correlation experiments.

Ethyl 3-(1,4,7,10-Tetraazacyclododecan-1-yl)propanoate (1): Cyclen(5.34 g, 31 mmol) was dissolved in chloroform (120 mL) and trieth-ylamine (1.29 mL, 9.3 mmol) was added. Ethyl 3-bromopropionate(0.99 mL, 7.75 mmol) was then added and the solution was gentlyheated at reflux under nitrogen for 15 h. The reaction mixture wascooled and extracted with aq. NaOH (1 m, 3�40 mL) and water(3 �40 mL). The organic phase was separated, evaporated to dry-ness, redissolved in chloroform (40 mL) and extracted again withaq. NaOH (1 m, 2�40 mL) and water (2 �40 mL). The organicphase was then dried with anhydrous sodium sulfate, filtered, evap-orated to dryness and the residue was dried under vacuum to yieldpure mono-alkylated cyclen as a colourless oil (1.32 g, 63%). 1HNMR (CDCl3): δ = 1.18 (t, 3 H, O–CH2–CH3), 2.41 (t, 2 H, CH2–CO), 2.45–2.75 (m, 18 H, ring and pendant N–CH2), 4.07 (q, 2 H,O–CH2–CH3) ppm. 13C{1H} NMR: δ = 14.1 (1 C, O–CH2–CH3),32.6 (1 C, CH2–CO), 44.8, 45.9, 46.8, 51.1 (4 �2 C, ring N–CH2),49.8 (1 C, pendant N–CH2), 60.2 (1 C, O–CH2–CH3), 172.5 (1 C,C=O) ppm.

Ethyl 3-{4,7,10-Tris[(diethoxyphosphoryl)methyl]-1,4,7,10-tetraaza-cyclododecan-1-yl}propanoate (2): Compound 1 (1.28 g, 4.7 mmol)was dissolved in triethyl phosphite (2.58 mL, 14.8 mmol). Para-formaldehyde was then added (0.49 g, 15.5 mmol) in small portions


over 1 h. The solution was finally stirred at room temp. for an ad-ditional 4 d. The volatiles were evaporated and the residue was co-evaporated with toluene (3 � 50 mL) and dried under vacuum forseveral hours to yield a clear oil (3.34 g). This crude product, whichalso contained a minor amount of a diethyl hydroxymethylphos-phonate byproduct (around 8% based on 31P NMR spectroscopicpeak integration; overall yield of 2 was 90%), was used withoutfurther purification in the following hydrolysis step to remove bothcarboxylate and phosphonate ethyl esters. 1H NMR (CDCl3): δ =1.21 (t, 3 H, C–O–CH2–CH3), 1.28 (t, 18 H, P–O–CH2–CH3), 2.38(t, 2 H, CH2–CO), 2.54 (m, 4 H, ring N–CH2), 2.71 (t, 2 H, N–CH2–CH2–CO), 2.80 (m, 12 H, ring N–CH2), 2.87 (d, 2JH,P =8.4 Hz, 6 H, N–CH2–P), 4.07 (m, 14 H, C–O–CH2–CH3 and P–O–CH2–CH3) ppm. 13C{1H} NMR: δ = 13.5 (1 C, C–O–CH2–CH3),15.7 (d, 3JC,P = 5.5 Hz, 6 C, P–O–CH2–CH3), 31.6 (1 C, CH2–CO),49.8, 49.9 (d, 1JC,P = 154.3 Hz, 1+2 C, N–CH2–P), 50.2 (1 C, N–CH2–CH2–CO), 51.9 (2 C, ring N–CH2), 52.5, 52.7, 52.8 (d, 3JC,P

= 6.3 Hz, 3 �2 C, ring P–CH2–N–CH2), 59.3 (1 C, C–O–CH2–CH3), 60.8 (d, 2JC,P = 6.6 Hz, 6 C, P–O–CH2–CH3), 171.7 (1 C,C=O) ppm. 31P{1H} NMR: δ = 26.99 (2 P), 27.02 (1 P) ppm.

3-[4,7,10-Tris(phosphonomethyl)-1,4,7,10-tetraazacyclododecan-1-yl]-propanoic Acid (H7do3p1pr): Compound 2 (1.34 g, around1.7 mmol at 92% purity) was dissolved in aq. HCl (20%, 50 mL)and the solution was stirred at room temp. for 12 h and then heatedto a gentle reflux for an additional 36 h. The suspension was fil-tered, volatiles from the filtrate were evaporated and the residuewas co-evaporated with water (3 � 50 mL) to remove the excessamount of HCl. The residue was dissolved in a small amount ofwater, loaded in a weak cationic exchange column (Amberlite CG-50, 20 �4.5 cm, H+-form) and the column was eluted with water(800 mL) to collect fractions of around 20 mL. The fractions thatcontained only the desired compound were combined, treated withcharcoal (0.3 g), filtered and the solution was evaporated to obtaina hygroscopic transparent oil which, after drying under vacuum,yielded pure H7do3p1pr in the zwitterionic form as very hygro-scopic solid (0.52 g, 58%). ESI-MS: m/z = 527.1 [M + H]+. 1HNMR (DCl in D2O, pD = 1.2): δ = 2.39 (t, 2 H, CH2-CO), 2.72(d, 2JH,P = 11.2 Hz, 4 H, N–CH2–P), 2.76, 2.93, 3.10, 3.39 (m, 16H, ring N–CH2), 3.46 (d, 2JH,P = 11.7 Hz, 2 H, N–CH2–P), 3.62(t, 2 H, N–CH2–CH2–CO) ppm. 13C{1H} NMR: δ = 36.6 (1 C,CH2–CO), 42.8, 49.6 (2 � 2 C, ring N–CH2), 49.9 (d, 1JC,P =137.9 Hz, 1 C, N–CH2–P), 50.3 (d, 1JC,P = 132.4 Hz, 2 C, N–CH2–P), 51.1, 51.4 (2 �2 C, ring N–CH2), 57.1 (2 C, N–CH2–CH2–CO),176.3 (1 C, C=O) ppm. 31P{1H} NMR: δ = 6.97 (1 P), 21.64 (2 P)ppm.

(1,4,7,10-Tetraazacyclododecane-1,4,7-triyl)tris(methylene)triphos-phonic Acid (H6do3p): Compound 2 (1.89 g, around 2.4 mmol at92% purity) was dissolved in aq. HBr (20%, 50 mL) and the solu-tion was heated at reflux for 3 d. The suspension was filtered, vola-tiles from the filtrate were evaporated and the residue was co-evap-orated with water (3 �50 mL) to remove the excess amount of HBr.The product was dissolved in a small amount of water, the solutionwas loaded on a strong cationic exchange column (Dowex 50WX4,20 � 2.5 cm, H+-form), and the column was eluted with water(300 mL) followed by 5 % aq. ammonia solution (500 mL). Thewater fractions contained only inorganic impurities, whereas thecombined ammonia fractions that contained macrocyclic com-pounds were evaporated to dryness and co-evaporated with water(3 � 50 mL) to remove the excess amount of ammonia. The re-sulting oil was dissolved in a minimal volume of water, loaded ina weak cationic exchange column (Amberlite CG-50, 20 �4.5 cm,H+-form) and eluted with water (600 mL) in fractions of around20 mL. The fractions that contained only the desired compound

R. Delgado, P. Lubal et al.FULL PAPERwere combined and the solution was treated with charcoal (0.4 g),filtered and concentrated in the evaporator. After cooling and over-night standing a white solid precipitated, which after filtration anddrying under vacuum yielded pure H6do3p in the zwitterionic form(0.71 g, 65%). C11H29N4O9P3 (454.29): calcd. C 29.08, H 6.43, N12.33; found C 29.10, H 6.73, N 12.30. ESI-MS: m/z = 455.1 [M +H]+. 1H NMR (KOD in D2O, pD = 11.6): δ = 2.73 (d, 2JH,P =9.9 Hz, 4 H, N–CH2–P), 2.90, 2.98, 3.04 (t, 3 �4 H, ring N–CH2),3.08 (d, 2JH,P = 10.8 Hz, 2 H, N–CH2–P), 3.24 (t, 4 H, ring N–CH2) ppm. 13C{1H} NMR: δ = 44.4, 50.5, 51.0 (3 �2 C, ring N–CH2), 51.1 (d, 1JC,P = 125.4 Hz, 1 C, N–CH2–P), 52.2 (2 C, ringN–CH2), 53.2 (d, 1JC,P = 136.1 Hz, 2 C, N–CH2–P) ppm. 31P{1H}NMR: δ = 10.42 (1 P), 15.68 (2 P) ppm.

3-Chloropropyl Acetate (3): A solution of acetyl chloride (1.43 mL,20 mmol) in dichloromethane (10 mL) was added dropwise(30 min) to a solution of 3-chloropropan-1-ol (1.67 mL, 20 mmol)and triethylamine (3.46 mL, 25 mmol) in dichloromethane (15 mL)while cooling the mild exotherm with an ice/water bath (around5 °C). The reaction was then continued at room temp. for another2 h. The resulting solution was added to water (25 mL), vigorouslystirred for 10 min and then extracted with dichloromethane. Theorganic phase was dried with anhydrous sodium sulfate, filteredand the solvents evaporated. The clear yellowish liquid obtainedwas dried under limited reduced pressure (�50 Torr) to remove theremaining volatile impurities and yield the title compound as a li-quid (2.62 g, 96%). 1H NMR (CDCl3): δ = 1.97 (s, 3 H, CO–CH3),2.02 (m, 2 H, CH2–CH2–CH2), 3.55 (t, 2 H, Cl–CH2–CH2), 4.13(t, 2 H, CH2–CH2–O) ppm. 13C{1H} NMR: δ = 20.6 (1 C, CO–CH3), 31.4 (1 C, CH2–CH2–CH2), 41.1 (1 C, Cl–CH2–CH2), 60.9(1 C, CH2–CH2–O), 170.5 (1 C, C=O) ppm.

3-(1,4,7,10-Tetraazacyclododecan-1-yl)propyl Acetate (4): Cyclen(3.44 g, 20 mmol) was dissolved in chloroform (50 mL) and trieth-ylamine (0.83 mL, 6 mmol) was added. 3-Chloropropyl acetate(compound 3, 0.96 g, 5 mmol) was then added and the solutionwas gently heated at reflux under nitrogen for 3 d. The reactionmixture was cooled to room temp. and extracted with aq. NaOH(1 m, 3 �20 mL) and water (3 �20 mL). The organic phase wasseparated, evaporated to dryness, the residue was redissolved inchloroform (30 mL) and the solution was extracted again with aq.NaOH (1 m, 2 �20 mL) and water (2 �20 mL). The organic phasewas then dried with anhydrous sodium sulfate, filtered, evaporatedto dryness and the residue was dried under vacuum to yield puremono-alkylated cyclen as yellowish oil (0.71 g, 52 %). 1H NMR(CDCl3): δ = 1.74 (m, 2 H, CH2–CH2–CH2), 1.96 (s, 3 H, CO–CH3), 2.40–2.75 (m, 18 H, ring and pendant N–CH2), 4.07 (t, 2 H,CH2–CH2–O) ppm. 13C{1H} NMR: δ = 20.7 (1 C, CO–CH3), 26.2(1 C, CH2–CH2–CH2), 44.9, 45.9, 46.8, 51.3 (4 �2 C, ring N–CH2),50.4 (1 C, pendant N–CH2), 62.2 (1 C, CH2–CH2–O), 170.6 (1 C,C=O) ppm.

3-{4,7,10-Tris[(diethoxyphosphoryl)methyl]-1,4,7,10-tetraazacyclo-dodecan-1-yl}propyl Acetate (5): Compound 4 (0.66 g, 2.4 mmol)was dissolved in triethyl phosphite (1.32 mL, 7.6 mmol). Paraform-aldehyde was then added (0.24 g, 7.9 mmol) in small portions over1 h. The solution was finally stirred at room temp. for an additional3 d. The volatiles were evaporated, and the residue was co-evapo-rated with toluene (3 �50 mL) and dried under vacuum to yield aclear oil (1.72 g). This crude product, which also contained a minoramount of a diethyl hydroxymethylphosphonate byproduct (around7 % based on 31P NMR spectroscopic peak integration, overallyield of 5 was 92%), was used without further purification in thefollowing hydrolysis step to remove all ethyl ester protecting groups.1H NMR (CDCl3): δ = 1.26 (t, 18 H, P–O–CH2–CH3), 1.69 (m, 2


H, CH2–CH2–CH2), 1.97 (s, 3 H, CO–CH3), 2.38 (s, 2 H, pendantN–CH2), 2.51, 2.78 (t, 4+12 H, ring N–CH2), 2.86 (d, 2JH,P =8.4 Hz, 6 H, N–CH2–P), 3.97 (t, 2 H, CH2–CH2–O), 4.06 (m, 12H, P–O–CH2–CH3) ppm. 13C{1H} NMR: δ = 15.5 (6 C, P–O–CH2–CH3), 19.9 (1 C, CO–CH3), 25.5 (1 C, CH2–CH2–CH2), 49.8,49.9 (d, 1JC,P = 154.4 Hz, 2+1 C, N–CH2–P), 51.2 (1 C, pendantN–CH2), 51.6, 52.1, 52.3, 52.4 (4 �2 C, ring N–CH2), 60.3 (d, 2JC,P

= 5.7 Hz, 6 C, P–O–CH2–CH3), 61.9 (1 C, CH2–CH2–O), 169.9 (1C, C=O) ppm. 31P{1H} NMR: δ = 25.70 (1 P), 25.76 (2 P) ppm.

[10-(3-Hydroxypropyl)-1,4,7,10-tetraazacyclododecane-1,4,7-triyl]-tris(methylene)triphosphonic Acid (H6do3p1ol): Compound 5(1.63 g, around 2.1 mmol at 93 % purity) was dissolved in aq. HCl(20%, 50 mL) and the solution was heated at reflux for 2 d. Thevolatiles were evaporated and the residue was co-evaporated withwater (3�50 mL) to remove the excess amount of HCl. The residuewas dissolved in diluted aq. NaOH (0.2 m, 50 mL) and the solutionwas heated to reflux for another 12 h. The suspension was filtered,volatiles from the filtrate were evaporated and the oil obtained wasthen purified by ion-exchange chromatography as described forH6do3p. The fractions from the weak cationic-exchange columncontaining only the desired compound were combined, treated withcharcoal (0.5 g), filtered, evaporated and the product was dried un-der vacuum to yield pure H6do3p1ol in the zwitterionic form as ahygroscopic off-white solid (0.76 g, 71%). C14H35N4O10P3 (512.37):calcd. C 32.82, H 6.89, N 10.93; found C 32.50, H 6.88, N 10.98.MS: m/z 513.1 [M + H]+. 1H NMR (DCl in D2O, pD = 1.8): δ =1.79 (m, 2 H, CH2–CH2–O), 2.65 (d, 2JH,P = 11.1 Hz, 4 H, N–CH2–P), 2.79 (t, 2 H, pendant N–CH2), 2.89, 3.16, 3.22 (t, 6+4+2H, ring N–CH2), 3.35 (t, 2 H, CH2–CH2–O), 3.39 (d, 2JH,P =12.3 Hz, 2 H, N–CH2–P), 3.45 (t, 4 H, ring N–CH2) ppm. 13C{1H}NMR: δ = 24.4 (1 C, CH2–CH2–O), 49.6 (1 C, pendant N–CH2),50.2 (d, 1JC,P = 132.4 Hz, 1 C, N–CH2–P), 50.6 (d, 1JC,P =136.1 Hz, 2 C, N–CH2–P), 49.7, 49.9, 51.1, 51.3 (6 C, ring N–CH2),51.8 (1 C, CH2–CH2–O), 58.3 (2 C, ring N–CH2) ppm. 31P{1H}NMR: δ = 6.68 (1 P), 20.35 (2 P) ppm.

X-ray Crystallography: Single crystals of H6do3p·3.8H2O were ob-tained in the form of small colourless prisms upon leaving a con-centrated aqueous solution of the free ligand standing over a fewdays. Single crystals of [Cu(H2O)6]2+[Cu(H2O)(H4dotp)]2– were ob-tained at random as small blue prisms upon standing over severalmonths a concentrated acidic solution (pH 2–3) that contained anexcess amount of Cu(ClO4)2 and a mixture of H7do3p and H8dotpthat came from preliminary synthetic work. Selected crystals weremounted on a glass fibre in random orientation and cooled to150(1) K. The diffraction data were collected with a Nonius KappaCCD diffractometer (Enraf–Nonius) by using Mo-Kα (λ =0.71073 Å) at 150(1) K (Cryostream Cooler Oxford Cryosystem)and analyzed by using the HKL DENZO program package[59,60]

for H6do3p·3.8H2O and Bruker APEX/SAINT software[61] for[Cu(H2O)6]2+[Cu(H2O)(H4dotp)]2–. The structures were solved bydirect methods and refined by full-matrix least-squares techniques(SIR92,[62] SHELXS97[63] and SHELXL97[63]). The used scatteringfactors for neutral atoms were included in the SHELXL97 pro-gram. In the case of [Cu(H2O)6]2+[Cu(H2O)(H4dotp)]2–, a semiem-pirical absorption correction was carried out using SADABS.[64]

Selected experimental data are listed in Table 7. CCDC-855329(for H6do3p·3.8H2O) and -855328 (for [Cu(H2O)6]2+-[Cu(H2O)(H4dotp)]2–) contain the supplementary crystallographicdata for this paper. These data can be obtained free of charge fromThe Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif. The crystal structure of H6do3p·3.8H2Owas refined as a racemic twin with near 1:1 enantiomeric ratio. Allnon-hydrogen atoms were refined anisotropically. The independent


unit contains a ligand molecule and several water solvate molecules.Two molecules were refined with full occupancy, and one of theremaining water molecules was best refined disordered in two posi-tions with relative 55:45 occupancy. However, some other weakmaxima of residual electronic density close to each other pointed toanother disordered water molecule. This disorder cannot be reliablyfitted as there were several maxima with a low intensity, so thiselectronic density was squeezed off by using PLATON.[65] By theelectronic count, the squeezed density corresponded to approxi-mately 0.8 water molecules. All hydrogen atoms were found in theelectronic difference map except those that belonged to the disor-dered water molecule; however, they were located in theoretical po-sitions (C–H) or were kept in their original positions (O–H, N–H)to keep a low number of refined parameters. Some O–H distancesare unusually long due to localization of the hydrogen atom nearits acceptor in the very short hydrogen-bond system (dO···O ≈2.45 Å). In the crystal structure of [Cu(H2O)6]2+-[Cu(H2O)(H4dotp)]2–, all non-hydrogen atoms were refined an-isotropically. All hydrogen atoms were found in the electronic dif-ference map, and were placed in theoretical positions (C–H) orwere kept in their original positions (O–H). The highest differencemaxima (1.07 and 0.75 eÅ–3) lay very close to copper(II) cations(ca. 1.1 Å) and thus they cannot be real atoms and are attributedto artefacts.

Table 7. Experimental crystallographic data for the reported crystalstructures.

H6do3p·3.8H2O [Cu(H2O)6]2+

[Cu(H2O)(H4dotp)]2–

Formula C11H36.6N4O12.8P3 C12H42Cu2N4O19P4

Mr 522.75 797.46Color and shape colourless prism blue prismCrystal size [mm] 0.15�0.18�0.25 0.13�0.25�0.36Crystal system orthorhombic orthorhombicSpace group Pna21 (no. 33) Pcca (no. 54)a [Å] 16.6902(2) 15.7326(11)b [Å] 8.0162(5) 11.2581(7)c [Å] 16.7793(4) 15.6342(11)V [Å3] 2244.94(15) 2769.1(3)Z 4 4Dcalcd. [gcm–3] 1.547 1.913μ [mm–1] 0.334 1.858F(000) 1112 1648GOF on F2 1.046 1.258Total reflections 4185 3033Obsd. reflections [I�2σ(I)] 3818 2471Parameters 281 187R; R�[I�2σ(I)] 0.0396; 0.0329 0.0712; 0.0563wR; wR�[I�2σ(I)] 0.0814; 0.0776 0.1328; 0.1281Largest diff. peak/hole [eÅ–3] 0.221/–0.287 1.069/–0.684

Potentiometric Measurements: Purified water was obtained with aMillipore Milli-Q demineralization system. Stock solutions ofH6do3p, H7do3p1pr and H6do3p1ol were prepared at around2�10–3 m. The [N(CH3)4]NO3 was prepared by neutralization of acommercial [N(CH3)4]OH solution with HNO3. Metal-ion solu-tions were prepared in water at 0.025–0.050 m from analytical-grade nitrate salts of the metal ions and standardized by titrationwith Na2H2edta.[66] Carbonate-free solutions of the titrant[N(CH3)4]OH were obtained at around 0.10 m by treating freshlyprepared silver oxide with a solution of [N(CH3)4]I under nitrogen.These solutions were standardized by application of Gran’smethod[67] and were discarded as soon as the concentration of car-bonate reached approximately 1% of the total amount of base. A0.100 m standard solution of HNO3 prepared from a commercial


ampoule was used for back-titrations. The potentiometric setupused for conventional titrations has been described before.[37] Mea-surements were carried out at (298.2 � 0.1) K in solutions withionic strength kept at (0.10 �0.01) m with [N(CH3)4]NO3, and at-mospheric CO2 was excluded from the cell during titrations bypassing purified N2 across the top of the experimental solutions.The [H+] of the solutions was determined by measurement of theelectromotive force of the cell, E = E°�+ Qlog [H+] + Ej. The termpH is defined as –log [H+]. E°� and Q were determined by titratinga solution of known hydrogen-ion concentration at the same ionicstrength in the acid pH region. The liquid-junction potential, Ej,was found to be negligible under the experimental conditions used.The value of Kw = [H+][OH–] was found to be equal to 10–13.80 bytitrating a solution of known hydrogen-ion concentration at thesame ionic strength in the alkaline pH region, considering E°� andQ valid for the entire pH range. Measurements during conventionaltitrations were carried out with around 0.05 mmol of ligand in atotal volume of around 30 mL, in the absence of metal ions and inthe presence of each metal ion at around 0.9 equiv. relative to theligand. A back-titration was always performed at the end of eachdirect complexation titration to check if equilibrium was attainedthroughout the full pH range. Each titration curve consisted typi-cally of 70–90 points in the pH 2.5–11.5 region and a minimum oftwo replicate titrations were performed for each system.

NMR Spectroscopic Measurements: For determination of the twohighest protonation constants of H6do3p, H7do3p1pr andH6do3p1ol, 31P NMR spectra of aqueous solution of the ligandswere recorded, approximately 20 points per titration over the rangeof pH 9 to 14 at 298.2 K. A ligand stock solution was prepared at0.010 m; the titrant was a fresh 25% (w/w) aqueous solution of[N(CH3)4]OH (Aldrich) and was standardized by titration with a1.0 m HNO3 solution. The titration was carried out in a closedtitration cell and the titrant was added with a Crison microBU2031 automatic burette. The pH was measured with an Orion 420Ameasuring instrument fitted with a Metrohm 6.0210.100 combinedelectrode after calibration of the electrode with commercial buffersolutions of standard pH (at 298.2 K) of 7.96 and 11.88. Atmo-spheric CO2 was excluded from the cell during the titration by pass-ing purified nitrogen across the top of the experimental solution.The measurements were carried out with 0.03 mmol of ligand in atotal volume of around 5 mL and the ionic strength of the titrationsolution was kept at 0.50 m with [N(CH3)4]NO3. Following eachaddition of titrant, the pH was measured after equilibration and asample of solution was placed in a 5 mm NMR spectroscopy tubeadapted with an internal capillary tube that contained D2O andH3PO4 for locking and referencing purposes. After recording each31P NMR spectrum, the sample volume was returned to the ti-tration cell. In the pH range of –1 to 9, 31P NMR spectroscopictitrations similar to the ones above were performed for each ligandwith a 1.0 m HNO3 solution as titrant to obtain titrations curvesin the full pH range.

Calculation of Thermodynamic Equilibrium Constants: The datafrom the 31P NMR spectroscopic titrations in the pH 9–14 regionwere used to calculate two highest protonation constants ofH6do3p, H7do3p1pr and H6do3p1ol at an ionic strength 0.50 m byrefinement with the HYPNMR program.[68] The 31P NMR spectraof each compound present two resonances, and all determinationswere done using both signals. Values of the first stepwise proton-ation constants thus obtained (log K) were 13.59 and 11.43 forH6do3p; 13.61 and 11.45 for H7do3p1pr; and 13.61 and 11.66 forH6do3p1ol. These values were then corrected to ionic strength0.10 m using the Davies equation[69] and the adjusted values werefinally used as constants in the determination of the remaining pro-

R. Delgado, P. Lubal et al.FULL PAPERtonation constants of the ligand. The calculation of overall equilib-rium constants βi

H and βMmHhLl(βMmHhLl

= [MmHhLl]/[M]m[H]h[L]l

and βMH–1L = βML(OH) �Kw) was done by refinement of thepotentiometric data from with the HYPERQUAD program.[70]

Species distribution diagrams were plotted from the calculated con-stants with the HYSS program.[71] Differences in log units betweenthe values of protonated (or hydrolyzed) and nonprotonated con-stants provide the stepwise (log K) reaction constants (KMmHhLl

=[MmHhLl]/[MmHh–1Ll][H]). The errors quoted are the standard de-viations of the overall stability constants calculated by the fittingprogram from all the experimental data for each system.

Kinetic Measurements: The dissociation kinetics of Cu2+ complexes(cCuL ≈ 1.0 �10–4 m) was studied under pseudo-first-order experi-mental conditions ([H+]tot �� cCuL) at an ionic strength of 3.00 m

(Na,H)Cl. The temperature was changed in the 288–328 K rangeto estimate the activation parameters. The complexes of Cu2+ wereprepared by mixing CuCl2 and ligand solutions in 1:1.2 molar ratiofollowed by stepwise neutralization of the stock solution of com-plex formed at pH 7.5–9.0. The absorbance kinetic measurementswere carried out on double-beam UV4 (PYE UNICAM, UK) anddiode-array HP8453A (Agilent, USA) spectrophotometers.

Supporting Information (see footnote on the first page of this arti-cle): Hydrogen bonding in the crystal structure of H6do3p·3.8H2O;overall protonation and stability constants with standard devia-tions; 31P NMR spectroscopic titration curve and tentative proton-ation sequence for H6do3p; representative speciation diagrams insolution, UV/Vis spectra of the Cu2+ complexes; rate constants andtime course for the acid-assisted dissociation of the Cu2+ com-plexes.

Acknowledgments

The authors acknowledge Fundação para a Ciência e a Tecnologia(FCT) (project PTDC/QUI/67175/2006) with co-participation ofthe European Community (funds FEDER, POCI, QREN andCOMPETE) for the financial support. The NMR spectrometerused is part of the National NMR Network and was purchased inthe framework of the National Program for Scientific Re-equip-ment, contract REDE/1517/RMN/2005, with funds from POCI2010 (FEDER) and FCT. The staff of the Analytical Services Unitof ITQB-UNL/IBET and M. C. Almeida are acknowledged forproviding data for elemental analysis and ESI-MS. L. M. P. L.thanks Fundação para a Ciência e a Tecnologia for the PhD fellow-ship (SFRH/BD/18522/2004). P. L. thanks the Ministry of Educa-tion, Youth and Sports of the Czech Republic (projects ME09065and LC06035), GRICES for a travel grant, and the EuropeanUnion (EU) (CEITEC CZ.1.05/1.1.0/02.0068). P. H. and J. K.thank the Grant Agency of the Czech Republic (grant number 207/11/1437), the Ministry of Education, Youth and Sports of theCzech Republic (MSM0021620857), and the European Union(EU) (COST TD 1004 and CM802).

[1] E. K. Barefield, Coord. Chem. Rev. 2010, 254, 1607–1627.[2] M. Meyer, V. Dahaoui-Gindrey, C. Lecomte, R. Guilard, Co-

ord. Chem. Rev. 1998, 178–180, 1313–1405.[3] K. P. Wainwright, Coord. Chem. Rev. 1997, 166, 35–90.[4] G. Anderegg, F. Arnaud-Neu, R. Delgado, J. Felcman, K. Po-

pov, Pure Appl. Chem. 2005, 77, 1445–1495.[5] I. Lukes, J. Kotek, P. Vojtísek, P. Hermann, Coord. Chem. Rev.

2001, 216–217, 287–312.[6] P. Lubal, M. Kývala, P. Hermann, J. Holubová, J. Rohovec, J.

Havel, I. Lukes, Polyhedron 2001, 20, 47–55.


[7] T. J. Wadas, E. H. Wong, G. R. Weisman, C. J. Anderson,Chem. Rev. 2010, 110, 2858–2902.

[8] M. Shokeen, C. J. Anderson, Acc. Chem. Res. 2009, 42, 832–841.

[9] T. J. Wadas, E. H. Wong, G. R. Weisman, C. J. Anderson, Curr.Pharm. Des. 2007, 13, 3–16.

[10] P. J. Blower, J. S. Lewis, J. Zweit, Nucl. Med. Biol. 1996, 23,957–980.

[11] L. Lattuada, A. Barge, G. Cravotto, G. B. Giovenzana, L. Tei,Chem. Soc. Rev. 2011, 40, 3019–3049.

[12] B. M. Zeglis, J. S. Lewis, Dalton Trans. 2011, 40, 6168–6195.[13] S. Liu, Adv. Drug Delivery Rev. 2008, 60, 1347–1370.[14] P. Táborský, I. Svobodová, P. Lubal, Z. Hnatejko, S. Lis, P.

Hermann, Polyhedron 2007, 26, 4119–4130.[15] J. Rudovský, P. Cígler, J. Kotek, P. Hermann, P. Vojtísek, I.

Lukes, J. A. Peters, L. Vander Elst, R. N. Muller, Chem. Eur. J.2005, 11, 2373–2384.

[16] P. Táborský, P. Lubal, J. Havel, J. Kotek, P. Hermann, I. Lukes,Collect. Czech. Chem. Commun. 2005, 70, 1909–1942.

[17] P. Vojtísek, P. Cígler, J. Kotek, J. Rudovský, P. Hermann, I.Lukes, Inorg. Chem. 2005, 44, 5591–5599.

[18] M. P. C. Campello, S. Lacerda, I. C. Santos, G. A. Pereira,C. F. G. C. Geraldes, J. Kotek, P. Hermann, J. Vanek, P. Lubal,V. Kubícek, É. Tóth, I. Santos, Chem. Eur. J. 2010, 16, 8446–8465.

[19] F. K. Kálmán, Z. Baranyai, I. Tóth, I. Bányai, R. Kíraly, E.Brücher, S. Aime, X. Sun, A. D. Sherry, Z. Kovács, Inorg.Chem. 2008, 47, 3851–3862.

[20] M. P. Campello, F. Marques, L. Gano, S. Lacerda, I. Santos,Radiochim. Acta 2007, 95, 329–334.

[21] M. P. Campello, M. Balbina, I. Santos, P. Lubal, R. Sevcík, R.Sevcíková, Helv. Chim. Acta 2009, 92, 2398–2413.

[22] L. Gano, F. Marques, M. P. Campello, M. Balbina, S. Lacerda,I. Santos, Q. J. Nucl. Med. Mol. Imaging 2007, 51, 6–15.

[23] X. Sun, M. Wuest, Z. Kovacs, A. D. Sherry, R. Motekaitis, Z.Wang, A. E. Martell, M. J. Welch, C. J. Anderson, J. Biol. In-org. Chem. 2003, 8, 217–225.

[24] U. Kreher, M. T. W. Hearn, L. Spiccia, Aust. J. Chem. 2009,62, 1583–1592.

[25] J. Massue, S. E. Plush, C. S. Bonnet, D. A. Moore, T.Gunnlaugsson, Tetrahedron Lett. 2007, 48, 8052–8055.

[26] E. K. Barefield, K. A. Foster, G. M. Freeman, K. D. Hodges,Inorg. Chem. 1986, 25, 4663–4668.

[27] J. Dale, Acta Chem. Scand. 1973, 27, 1115–1158.[28] I. Lázár, D. C. Hrncir, W.-D. Kim, G. Kiefer, A. D. Sherry, In-

org. Chem. 1992, 31, 4422–4424.[29] R. L. Webb, M. L. Mino, E. L. Binn, A. A. Pinkerton, Inorg.

Chem. 1993, 32, 1396–1402.[30] D. Kong, X. Huang, Y. Xie, Inorg. Chim. Acta 2002, 340, 133–

138.[31] D. Kong, C. Chin, Y. Xie, Bioorg. Med. Chem. Lett. 1999, 9,

1087–1092.[32] M. A. Halcrow, Dalton Trans. 2003, 4375–4384.[33] D. Kong, D. G. Medvedev, A. Clearfield, Inorg. Chem. 2004,

43, 7308–7314.[34] S. Chaves, R. Delgado, J. J. R. Fraústo da Silva, Talanta 1992,

39, 249–254.[35] F. Marques, L. Gano, M. P. Campello, S. Lacerda, I. Santos,

L. M. P. Lima, J. Costa, P. Antunes, R. Delgado, J. Inorg. Bio-chem. 2006, 100, 270–280.

[36] J. Kotek, P. Vojtísek, I. Císarová, P. Hermann, P. Jurecka, J.Rohovec, I. Lukes, Collect. Czech. Chem. Commun. 2000, 65,1289–1316.

[37] L. M. P. Lima, R. Delgado, M. G. B. Drew, P. Brandão, V.Félix, Dalton Trans. 2008, 6593–6608.

[38] R. Delgado, V. Félix, L. M. P. Lima, D. W. Price, Dalton Trans.2007, 2734–2745.

[39] L. Burai, J. Ren, Z. Kovács, E. Brücher, A. D. Sherry, Inorg.Chem. 1998, 37, 69–75.


[40] R. Delgado, L. C. Siegfried, T. A. Kaden, Helv. Chim. Acta1990, 73, 140–148.

[41] L. M. P. Lima, R. Delgado, J. Plutnar, P. Hermann, J. Kotek,Eur. J. Inorg. Chem. 2011, 527–538.

[42] M. Försterová, I. Svobodová, P. Lubal, P. Táborský, J. Kotek,P. Hermann, I. Lukes, Dalton Trans. 2007, 535–549.

[43] K. P. Guerra, R. Delgado, L. M. P. Lima, M. G. B. Drew, V.Félix, Dalton Trans. 2004, 1812–1822.

[44] C. F. G. C. Geraldes, A. D. Sherry, W. P. Cacheris, Inorg. Chem.1989, 28, 3336–3341.

[45] A. D. Sherry, J. Ren, J. Huskens, E. Brücher, É. Tóth,C. F. C. G. Geraldes, M. M. C. A. Castro, W. P. Cacheris, Inorg.Chem. 1996, 35, 4604–4612.

[46] R. Delgado, J. J. R. Fraústo da Silva, Talanta 1982, 29, 815–822.

[47] I. M. Kabachnik, T. Y. Medved, F. I. Belskii, S. A. Pisareva,Izv. Akad. Nauk Ser. Khim. 1984, 4, 844–849.

[48] W. P. Cacheris, S. K. Nickle, A. D. Sherry, Inorg. Chem. 1987,26, 958–960.

[49] E. Brücher, Top. Curr. Chem. 2002, 221, 103–122.[50] M. Port, J.-M. Idée, C. Medina, C. Robic, M. Sabatou, C. Co-

rot, Biometals 2008, 21, 469–490.[51] K. S. Woodin, K. J. Heroux, C. A. Boswell, E. H. Wong, G. R.

Weisman, W. Niu, S. A. Tomellini, C. J. Anderson, L. N. Zakh-arov, A. L. Rheingold, Eur. J. Inorg. Chem. 2005, 4829–4833.

[52] J. Kotek, P. Lubal, P. Hermann, I. Císarová, I. Lukes, T. God-ula, I. Svobodová, P. Táborský, J. Havel, Chem. Eur. J. 2003,9, 233–248.

[53] S. Füzerová, J. Kotek, I. Císarová, P. Hermann, K. Binnemans,I. Lukes, Dalton Trans. 2005, 2908–2915.

[54] I. Svobodová, J. Havlícková, J. Plutnar, P. Lubal, J. Kotek, P.Hermann, Eur. J. Inorg. Chem. 2009, 3577–3592.

[55] R. W. Hay, M. P. Pujari, Inorg. Chim. Acta 1985, 100, L1–L3.


[56] W.-J. Lan, C.-S. Chung, J. Chem. Soc., Dalton Trans. 1994,191–194.

[57] P. Lubal, P. Hermann, unpublished results.[58] D. D. Perrin, W. L. F. Armarego, Purification of Laboratory

Chemicals, Pergamon, Oxford, UK, 3rd ed., 1988.[59] Z. Otwinovski, W. Minor, HKL Denzo and Scalepack Program

Package, Nonius BV, Delft, The Netherlands, 1997.[60] Z. Otwinovski, W. Minor, Methods Enzymol. 1997, 276, 307–

326.[61] APEX and SAINT, Bruker AXS Inc., Madison, WI, USA,

2004.[62] A. Altomare, M. C. Burla, M. Camalli, G. Cascarano, C. Gia-

covazzo, A. Guagliardi, G. Polidori, J. Appl. Crystallogr. 1994,27, 435–435.

[63] G. M. Sheldrick, SHELXS97 and SHELSL97, A ComputerPrograms for Solution and Refining of Crystal Structures, Uni-versity of Göttingen, Germany, 1997.

[64] G. M. Sheldrick, SADABS, Bruker AXS Inc., Madison, WI,USA, 2004.

[65] A. L. Spek, PLATON – A Multipurpose Crystallographic Tool,Utrecht University, The Netherlands, 2005.

[66] G. Schwarzenbach, W. Flaschka, Complexometric Titrations,Methuen & Co, London, 1969.

[67] F. J. Rossotti, H. J. Rossotti, J. Chem. Educ. 1965, 42, 375–378.[68] C. Frassineti, S. Ghelli, P. Gans, A. Sabatini, M. S. Moruzzi,

A. Vacca, Anal. Biochem. 1995, 231, 374–382.[69] L. D. Pettit, I. Sukhno, V. Buzko, Aq. Solutions Programs Suite,

Academic Software, IUPAC, 2004.[70] P. Gans, A. Sabatini, A. Vacca, Talanta 1996, 43, 1739–1753.[71] L. Alderighi, P. Gans, A. Ienco, D. Peters, A. Sabatini, A.

Vacca, Coord. Chem. Rev. 1999, 184, 311–318.Received: November 28, 2011

Published Online: March 30, 2012

doi: 10.1002/ejic.201101334lamosa/nova_nmr/rel_fct2012/art12.pdf · (2-ethoxycarbonyl)ethyl moiety...

Documents