Comparative study of nanoporous Ln–Cu coordination polymers containing iminodiacetate as bridging ligand

Journal of Molecular Structure 1004 (2011) 215–221

Contents lists available at SciVerse ScienceDirect

Journal of Molecular Structure

journal homepage: www.elsevier .com/ locate /molst ruc

Comparative study of nanoporous Ln–Cu coordination polymers containingiminodiacetate as bridging ligand

Julia Torres a, Paula Morales a, Sixto Domínguez b, Javier González-Platas c, Ricardo Faccio d,Jorge Castiglioni e, Alvaro W. Mombrú d, Carlos Kremer a,⇑a Cátedra de Química Inorgánica, Departamento Estrella Campos, Facultad de Química, CC 1157 UdelaR, Montevideo, Uruguayb Departamento de Química Inorgánica, Facultad de Farmacia, Universidad de La Laguna, Tenerife, Spainc Departamento de Física Fundamental II, Servicio de Difracción de Rayos X, Universidad de La Laguna, Tenerife, Spaind Laboratorio de Cristalografía, Estado Sólido y Materiales (Cryssmat-Lab), DETEMA, Facultad de Química, CC 1157 UdelaR, Montevideo, Uruguaye LAFIDESU, DETEMA, Facultad de Química, CC 1157 UdelaR, Montevideo, Uruguay

a r t i c l e i n f o

Article history:Received 27 May 2011Received in revised form 4 August 2011Accepted 4 August 2011Available online 11 August 2011

Keywords:Heterometallic coordination polymerLanthanide ionsCrystal structureChemical speciation

0022-2860/$ - see front matter � 2011 Elsevier B.V. Adoi:10.1016/j.molstruc.2011.08.007

⇑ Corresponding author. Tel.: +598 2 9240744; fax:E-mail address: [email protected] (C. Kremer).

a b s t r a c t

The stoichiometric reaction of copper(II) chloride with iminodiacetic acid (H2ida), and lanthanide(III)chloride in water yields the heteropolynuclear complexes [Ln2Cu3(ida)6]�xH2O. In this work, the synthesisand full characterization of those complexes with Ln = Ce, Ho is presented. The structures are based on[Cu(ida)2] building blocks, linked by the Ln ions via carboxylate bridges. The formation of nanochannelsalong the crystallographic c axis is verified. The comparison with analogous complexes containing otherLn ions, shows that the channels are perfectly tuneable in size along the series. These chemical systemswere also investigated in solution (25.0 �C, I = 0.5 M Me4NCl) by potentiometry. The same kind of poly-nuclear species have been found in aqueous solution.

� 2011 Elsevier B.V. All rights reserved.

1. Introduction

The design and synthesis of 3d–4f heterometallic coordinationpolymers have attracted great interest during the last years. Thecombination of a 3d-metal ion (M), a lanthanide ion (Ln) and abridging ligand, produces intriguing MOFs with exciting multidi-mensional structures and interesting properties, such as fluores-cence, chemical sensors functions, and catalysis [1–9].

Among the potential bridging ligands, carboxylate groups havedemonstrated to be very versatile, and to produce different struc-tural motifs [10]. Those complexes containing the dicarboxylic li-gand iminodiacetate, ida (ida2�, the fully deprotonated form ofH2ida, the iminodiacetic acid) as bridging ligand belong to an inter-esting series of compounds. Some of them are already known: theisostructural complexes [Ln2Cu3(ida)6]�xH2O (Ln = La, Pr, Nd, Sm,Eu, Gd, Tb, Dy, Er) have been previously reported [11–18]. Thestructures are constructed on the basis of [Cu(ida)2] buildingblocks, linked by the Ln ions via carboxylate bridges. One of themost interesting aspects of these structures is the formation ofnanochannels along the crystallographic c axis. The channels housethe lattice water molecules, which can be reversibly removed byheating without the collapse of the network.

ll rights reserved.

+598 2 9241906.

The detection of polynuclear complexes in solution with thesame stoichiometry, [Sm2M3(ida)6] (M = Co, Ni, Cu, Cd) [19], is alsoan interesting result. The polynuclear assemblies arise as potentialhomogeneous catalysts for many purposes.

With this in mind, we started a systematic study of these interest-ing chemical systems. We report here the synthesis and character-ization of two new compounds of this series: [Ln2Cu3(ida)6]�8H2O(Ln = Ce, Ho) and discuss their structural properties. The comparisonof the new compound properties together with the previously re-ported isostructural polynuclear complexes, allows the analysis ofthe influence of the Ln ion on the cavity surface and volume, andthe thermal stability of the network. We also performed an aqueoussolution study of the mixed-ion systems Ln–M–ida (Ln = La, Ce, Ho,Yb; M = Ca, Mg, Mn, Fe, Co, Ni, Cu, Zn). The selection of these four lan-thanide ions was done in order to have a broad range of ion size(from La to Yb) to compare the thermodynamic stability of suchpolynuclear complexes.

2. Materials and methods

2.1. Chemicals and equipment

All common laboratory chemicals were of reagent grade, pur-chased from commercial sources and used without further purifica-tion. CaCl2�2H2O, MgCl2�6H2O, MnCl2�4H2O, (NH4)2Fe(SO4)2�6H2O,CoCl2�6H2O, Ni(ClO4)2�6H2O, CuCl2�6H2O, ZnCl2, LaCl3�7H2O,

216 J. Torres et al. / Journal of Molecular Structure 1004 (2011) 215–221

CeCl3�6H2O, HoCl3�6H2O, and YbCl3�6H2O were employed as metalsources. The infrared spectra, as KBr pellets, were obtained from aBomen MB 102 FT-IR spectrophotometer. Elemental analysis (C, H)was performed on a Carlo Erba EA 1108 instrument. Thermal analy-sis was performed on a Shimadzu TGA-50 instrument with a TA 50Iinterface, using a platinum cell and nitrogen atmosphere. Experi-mental conditions were 0.5 �C min�1 temperature ramp rate up to80 �C, and then 1 �C min�1. Nitrogen flow rate was 50 mL min�1.

2.2. Synthesis of [Ln2Cu3(ida)6]�8H2O (Ln = Ce (1), Ho (2))

CuCl2�6H2O (0.13 g, 0.75 mmol) and LnCl3�xH2O (0.5 mmol)were dissolved in 5–10 mL of water. Iminodiacetic acid (H2ida,0.20 g, 1.5 mmol) was dissolved in 5 mL water, and the pH valuewas adjusted to 7.0 with diluted ammonia. Both solutions weremixed, and a clear blue solution was obtained. pH value was read-justed to 5.0–6.0 with diluted ammonia. After some days, blue hex-agonal prismatic crystals were obtained, separated by filtrationand washed with water. Yield 30–50%. Anal. for 1: Calc. N, 6.0; C,20.6; H, 3.3. Found: N, 5.9; C, 20.1; H, 3.1%. Anal. for 2: Calc. N,5.8; C, 19.8; H, 3.2. Found: N, 6.0; C, 19.5; H, 3.3%. IR peaks associ-ated to the ida ligand appear at ca: 3230 m(N–H), 2930 m(C–H),1620 and 1407 m(COO), 1450 m(C–N), 950 m(C–C), 1115 m(CNC)cm�1 for both compounds.

TGA diagrams show a single weight loss between room temper-ature and 90 �C, corresponding to crystallization water. Calc. 10.3(1), 9.9 (2). Found: 10.4 (1), 9.9 (2)%. Decomposition point, ca.290 �C.

2.3. X-ray data collection and refinement

The X-ray diffraction data for 1 was collected at 293(2) K withan RIGAKU AFC-7S four-circle diffractometer [20] using graphitemonochromatized Mo Ka radiation (k = 0.71069 Å). Absorptionand intensity decay corrections were applied to the diffraction dataaccording to [20]. The structures were solved by direct methodslocating most of the non-H atoms using SHELXS [21]. Fourier recy-cling and least-squares refinement were used for model comple-tion with SHELXL [21] included in the WinGX suite of programs[22]. All non-hydrogen atoms have been refined anisotropically,and all hydrogen atoms have been placed in geometrically suitablepositions and refined riding with isotropic thermal parameter re-lated to the equivalent isotropic thermal parameter of the parentatom. The geometrical analysis of interactions in the structureswas performed with PLATON [23]. Thermal ellipsoid plots were ob-tained with ORTEP3 [24]. In both structures H atoms were posi-tioned geometrically and treated as riding with C–H = 0.97 Å andN–H = 0.91 Å. H atoms bonded to tertiary C atoms and N atomswere refined with Uiso(H) = 1.2(Ueq(C)).

For structure 2, the diffraction data was collected on a AgilentSuperNova diffractometer with micro-focus X-ray with Mo radia-tion (k = 0.71073 Å). CrysAlis PRO [25] software was used to collect,index, scale and apply analytical absorption correction based onfaces of the crystal. The structure solutions were obtained by directmethods, using the SIR2008 [26] program and refined using theSHELXL-97 [21] program. All non-hydrogen atoms were refinedwith anisotropic thermal parameters using full-matrix least-squares procedures on F2. The methyl-H atoms were refined as rigidgroups, which were allowed to rotate but not to tip, with Uiso

(H) = 1.5Ueq(C). All other hydrogen atoms were allowed to ride ontheir parent atoms with Uiso(H) = 1.2Ueq(C).

Important voids appear in both structures as potential sites forfurther solvent location. The mobility of the solvent in the void re-gion avoids any attempt of successful refinement. For this reasononly one water molecule was located at the void region, withoutthe inclusion of hydrogen atoms.

Crystal data, collection procedures and refinement results aresummarized in Table 1.

2.4. Potentiometric measurements

Acidic solutions of the metal ions were prepared from commer-cial salts and standardized according to standard techniques [27].All the solutions were freed of carbon dioxide by boiling the sol-vent, and subsequent cooling under Ar atmosphere. The standardHCl solution was prepared from Merck standard ampoules. The ti-trant solution (0.1 M solution of Me4N(OH) in 0.50 M Me4NCl) wasprepared by dissolving Me4N(OH)�5H2O (from Fluka), and stan-dardized with potassium biphthalate. The protonation constantsof the ligand ida and the hydrolysis constants of La(III) and Ce(III)under the same conditions of the study, were taken from our pre-vious reports [19,28]. The hydrolysis constants of Yb(III) andHo(III), were also measured in this work, by potentiometry.

The formation constants of M(II)–ida and Ln(III)–ida specieswere also needed as a basis for the subsequent study. The behav-iour of ida in the presence of Ln(III) ions (La, Ce, Ho and Yb) orM(II) ions (Ca and Mg), for which previous data under the sameconditions were not available, was also analyzed through at leastthree potentiometric titrations per metal ion (ca. 150 experimentalpoints for each titration), at metal concentrations ranging from 1 to5 mM and Ln:ida molar ratios ranging from 1:1 to 1:3. Other for-mation constants of M(II)–ida species were taken from our previ-ous report [19]. Then, the formation of the mixed species wastested by at least three other potentiometric titrations withLn:M:ida molar ratios between 1:1:1 and 2:3:6, and different totalconcentrations of the components.

In all cases, the experimental data were collected with an auto-matic titrator Mettler Toledo T50 as previously described [19]. Theionic strength was kept constant throughout the titrations by usingsolutions containing 0.50 M Me4NCl and relatively low concentra-tions of the metal ions. The cell constants E�, and the liquid junctionpotentials were determined by means of a strong acid-strong basetitration using the GLEE program [29]. Data were analyzed usingthe HYPERQUAD program [30], and species distribution diagramswere produced using the HySS program [31]. The fit of the valuespredicted by the model to the experimental data was estimated onthe basis of the parameter r, corresponding to the scaled sum ofsquare differences between predicted and experimental values [30].

3. Results and discussion

3.1. Solid state characterization

The direct reaction of H2ida in aqueous solution with the stoi-chiometric amounts of the Cu(II) salt and the Ln(III) salt led tothe formation of complexes with the formula [Ln2Cu3(ida)6]�8H2O,according to:

2LnCl3 þ 3CuCl2 þ 6H2idaþ 8H2O

! ½Ln2Cu3ðidaÞ6� � 8H2Oþ 12HCl

The slow evaporation of the reaction solution afforded a crystal-line product. It was possible to obtain single crystals for 1 and 2which were measured by X-ray diffraction. The complexes[Ln2Cu3(ida)6]�8H2O (Ln = Ce, Ho) crystallize in the trigonal crystalsystem, space group P3c1. The two compounds are isostructuraland they will be described together using (1) as a model com-pound, see Fig. 1. Selected bond lengths and angles are given in Ta-ble 2. Unit cell packing is shown in Fig. 1a. Each Cu(II) ion iscoordinated by two ida ligands acting as tridentated. The geometryof the Cu2+ ion can be described as distorted octahedral with thetwo axial bonds elongated (Cu–O1), resulting from the Jahn–Tellereffect (Fig. 1b). This is also reflected in the bond distances depicted

Table 1Crystallographic data for compounds 1 and 2.

1 2

Empirical formula C48H92Cu6Ce4N12O62 C48H92Cu6Ho4N12O62

Formula weight 1305.40 1355.05Crystal system Trigonal TrigonalSpace group P � 3c1 P � 3c1a (Å) 13.4377(15) 13.3238(4)b (Å) 13.4377(15) 13.3238(4)c (Å) 14.853(4) 14.3368(5)a (�) 90 90b (�) 90 90c (�) 120 120V (Å3) 2322.7(7) 2204.14(11)Z 2 2Dcalc (g cm�3) 1.867 2.042l (Mo Ka) (mm�1) 3.361 5.078F(0 0 0) 1270 1370Crystal dimensions (mm) 0.20 � 0.20 � 0.20 0.1777 � 0.1436 � 0.1219Theta range for data collection (�) 2.74–27.48 3.06–29.64Limiting indices �17 6 h 6 18, �16 6 k 6 16, �13 6 l 6 19 �17 6 h 6 18, �16 6 k 6 16, �13 6 l 6 19Reflections (collected/ unique, (Rint)) 4047/1774, 0.064 6389/1879, 0.024R1

a, wR2b [F2 > 2 r (F2)] 0.047, 0.172 0.025, 0.062

Goodness-of-fit on F2 1.16 1.07Largest diff. peak and hole (e �3) 1.67/�1.71 0.80/�0.63

a R1 = R||F0| � |Fc||/R|Fc|.b wR2 = {R[wðF2

0 � F2c Þ2�=R½wðF

20Þ2

1=2.

Fig. 1. [Ln2Cu3(ida)6]�8H2O structure diagrams. Unit cell (a) and coordination geometry of Cu(II) (b) and Ce(III) (c) for compound 1. Thermal ellipsoids are shown at 50%probability. H atoms and crystallization water molecules are omitted for clarity.

J. Torres et al. / Journal of Molecular Structure 1004 (2011) 215–221 217

Table 2Selected bond lengths (Å) and angles (�) for compounds 1 and 2.

1 2

Ln–O1 2.477(5) 2.348(3)Ln–O2 2.758(6) 2.707(3)Ln–O3 2.424(6) 2.292(3)Cu–N1 2.017(6) 2.006(3)Cu–O1 2.387(5) 2.423(3)Cu–O4 1.960(5) 1.958(2)

O1–Ln–O2 49.32(16) 50.99(15)O1–Ln–O3 142.1(2) 141.15(10)O2–Ln–O3 80.61(19) 79.02(9)N1–Cu–O4 94.5(2) 94.51(11)O1–Cu–O4 90.3(2) 89.28(10)O4–C4–O3 122.2(7) 123.0(4)O1–C1–O2 121.6(7) 121.8(3)

Fig. 3. Variation of the cell parameters a and c with ionic radii of lanthanide ions.Cell parameters were taken from previously reported and our values [11,13,14,16–18]. Shannon’s radii for lanthanide ions were used [32].

218 J. Torres et al. / Journal of Molecular Structure 1004 (2011) 215–221

in Table 2. Each [Cu(ida)2] unit has four uncoordinated oxygenatoms, which are used to bind to the Ln(III) ions. This leads tothe formation of a 3D structure. The Ln ion is nine-coordinated ina distorted tricapped trigonal prism geometry, surrounded by ninecarboxylic oxygen atoms from six neighbouring ida ligands(Fig. 1c). The links are supported by three l2 anti–anti carboxylatebridges, and three l3 carboxylate bridges.

Hexagonal channels are present in the structure, running alongthe crystallographic c axis (Fig. 2). The channels are delimited bysix CuO4N2 polyhedra and six CeO9 polyhedra, and have 8.46 (1)and 8.41 (2) Å diameter. The lattice water molecules are situatedin these channels. The elemental and TGA analysis (Fig. S1) showa single weight loss at 50–90 �C which correspond to the eightcrystallization water molecules. This fact suggests that water mol-ecules are only weakly bound to the network. Chemical decompo-sition appears only above 280 �C for both complexes. For otheranalogous [Ln2Cu3(ida)6] compounds, the thermal stability is sim-ilar, with decomposition temperatures around 250–300 �C[11,13,14,16,17].

The size of the isostructural cells in the series [Ln2Cu3(ida)6]varies monotonously and according to the size of the lanthanideion (Fig. 3). In fact both a and c cell parameters decrease alongthe series. Consequently, the diameter of the channel changes from8.49 Å in [La2Cu3(ida)6] to 8.41 Å in [Er2Cu3(ida)6], which is thesmallest lanthanide compound fully characterized up to now. A

Fig. 2. Packing of compound 1 along crystal

linear relationship between the channel diameter (measured asthe shortest atomic distance across the channel) and the cellparameter a is also observed (Supplementary material, Fig. S2).

These structures exhibit a high percentage of free space with re-gard to the total volume of the unit cell. If we take the water mol-ecules out of the structures and using Van der Waal’s radii of allelements, unoccupied volume is still ca. 28%. The spacious chan-nels formed account for this void space, which again varies linearlywith the cell parameter a (Supplementary material, Fig. S3). Aswater molecules are easily removed from the channels withoutdecomposition of the structures, these tuneable size channelsmight be used to selectively store small molecules such as hydro-gen. We can also compare the inner channel surface (calculated asthe ratio of each channel free volume to the cell parameter, c) forthese compounds. A dependence on parameter a and hence tothe lanthanide ion radius is again verified (Fig. 4).

3.2. Solution chemistry

The aim of this part of the work is to explore the chemical spe-ciation in the Ln(III)–M(II)–ida ternary systems, searching for the

lographic c axis, showing the channels.

Fig. 4. Variation of the channel surface with the cell parameter a. Data were takenfrom previously reported and this work values [10,12,13,15–17].

Table 3Overall formation constants for Ln(III)–ida and M(II)–ida complexes at 25.0 �C,I = 0.50 M Me4NCl. Each log K value corresponds to the formation equilibria of thecomplex according to the following general equations: Ln3+ + rL2� + sH+

M

[LnLrHs](2r+s�3)� or M2+ + rL2� + sH+M [MLrHs](2r+s�2)�.

Ln(III)–ida [LnL]+ [LnL2]� [LnL3]3� [Ln(HL)]2+ r Reference

Ce3+ 5.28(4) 9.19(5) 12.38(5) 0.6 This workLa3+ 5.471(8) 9.43(8) 12.14(6) 10.56(4) 0.5 This workHo3+ 6.76(4) 12.15(4) 16.71(5) 11.4(1) 0.7 This workYb3+ 7.15(1) 12.79(1) 17.65(8) 10.97(1) 1.0 This work

M(II)–ida [ML] [ML2]2� [ML(HL)]� [M(HL)]+

Ca2+ 2.42(3) 4.35(5) 10.81(6) 0.4 This workMg2+ 2.87(1) 4.69(3) 10.94(4) 0.3 This workMn2+ 4.30 7.01 10.34 a [19]Fe2+ 5.78 10.15 10.54 [19]Co2+ 7.265 12.853 17.83 11.246 [19]Ni2+ 8.48 15.09 18.94 11.15 [19]Cu2+ 10.32 16.18 20.80 [19]Zn2+ 7.47 12.69 11.2 [19]Cd2+ 5.27 9.09 16.01 10.84 [19]

a For this system, b112 and b11-1 were also reported for minor species, but they areomitted for clarity.

J. Torres et al. / Journal of Molecular Structure 1004 (2011) 215–221 219

formation of analogous stoichiometries as those obtained in the so-lid state. In aqueous solution, the measurement of the correspond-ing formation constants needs the previous determination of theprotonation constants of the ligand, the hydrolysis constants ofthe metal ions (as this is an unavoidable competitive process),and the formation constants of the binary complexes Ln(III)–idaand M(II)–ida, under the same experimental conditions.

The protonation constants of ida were taken from our previousreport (KH

1 ¼ 9:267ð3Þ; bH2 ¼ 11:903ð6Þ; bH

3 ¼ 13:772ð8Þ [19]). Thehydrolysis constants for Ce, and La were taken from previously re-ported data under identical conditions (log �K1, i.e. [Ln(OH)]2+

formation:�9.3(1) for La and�5.7(1) for Ce [19,21]). The hydrolysis

Table 4Overall formation constants for Ln(III)–M(II)–ida complexes, at 25.0 �C, I = 0.50 M Me4NCl.formation equilibria of the complex according to the following general equation: pLn3+ + q

Ln M Equilibrium formation constant (log bpqrs)

[Ln2M3L6] [Ln2M3L6(OH)3] [LnML2]

Ce Ca2+ 37.65(5) 9.22(6) 12.43(2)Mg2+ 37.79(7) 9.88(7) 12.35(2)Mn2+ 37.92(4) 15.11(6) 11.3(1)Fe2+ 22.4(1)Co2+ 52.7(1)Ni2+ 56.91(5) 17.67(5)Cu2+ 60.8(2) 19.96(5)Zn2+ 53.07(7) 16.78(5)

La Ca2+ 33.4(1) 6.76(2) 11.17(3)Mg2+ 34.56(5) 7.22(2) 11.53(2)Mn2+ 37.39(3)Fe2+ 20.79(6)Co2+ 49.95(6) 15.0(1)Ni2+ 55.55(7) 17.37(6)Cu2+ 57.72(7) 18.41(3)Zn2+ 49.10(3)

Ho Ca2+ 43.99(7) 15.06(3) 14.62(3)Mg2+ 44.74(5) 16.68(4) 14.85(2)Mn2+ 44.67(3) 21.10(2) 14.25(3)Fe2+ 48.5(1) 14.75(6)Co2+ 53.81(7) 16.60(5)Ni2+ 59.53(7) 18.41(6)Cu2+ 62.07(5) 19.96(2)Zn2+ 51.97(6) 33.11(6)

Yb Ca2+ 19.78(5) 15.57(4)Mg2+ 19.6(1) 15.78(4)Mn2+ 15.01(2)Fe2+ 46.9(1) 24.79(7) 14.92(5)Co2+ 53.8(1) 16.78(7)Ni2+ 59.56(6)Cu2+ 61.1(1) 40.27(6) 19.17(2)Zn2+ 51.3(2) 32.3(1)

constant of holmium and ytterbium were measured in this workand formation of [Ho(OH)]2+ and [Yb(OH)]2+ was determined(log �K1 = �7.56(2), r = 0.5 for Ho; log �K1 = �7.7(1), r = 1.0 for Yb).

Table 3 compiles the formation constants for the binary systemsLn–ida and M(II)–ida required for this study. Some of them weremeasured in this work while others were taken from our previousreport [19]. The mixture of ida with lanthanide ions, provokes theformation of mononuclear species containing one, two or three li-gands per metal ion. These stoichiometries and the values of the

Charges of the species are omitted for clarity. Each log bpqrs value corresponds to theM2+ + rL2� + sH2O M [LnpMqLr(OH)s](2r+s�3p�2q)� + sH+.

[LnML] [LnM(OH)L] [LnM(OH)L3] r

7.04(8) 0.77.10(7) 0.7

0.88.18(4) 2.510.32(2) 15.6(1) 1.7

18.70(5) 1.714.02(3) 1.710.47(2) 16.60(6) 0.6

6.79(4) 1.17.17(2) 0.56.18(9) 7.22(2) 0.37.45(8) 1.79.31(4) 3.31(8) 13.86(6) 1.1

16.70(8) 1.412.01(4) 16.73(4) 0.5

13.69(3) 0.5

9.09(2) 9.94(3) 1.29.13(2) 10.61(2) 1.28.91(1) 11.54(2) 0.78.92(2) 0.510.10(4) 16.94(4) 1.510.83(5) 19.97(4) 1.812.95(3) 19.47(2) 0.5

1.5

9.609(8) 4.28(4) 0.79.6(1) 4.17(7) 1.68.42(5) 0.79.09(2) 1.110.24(5) 17.49(8) 1.49.9(2) 22.10(7) 2.011.7(1) 1.39.32(6) 0.8

2 4 6 8

pH

0

20

40

60

80

100

% fo

rmat

ion

rela

tive

to C

e

Ce3+

[CeCuL]3+

[Ce2Cu3L6]

[CeCuL2]+[Ce(OH)]2+

2 4 6 8pH

0

20

40

60

80

100

% fo

rmat

ion

rela

tive

to L

a

La3+

[La(HL)]2+

[LaCuL]3+

[La2Cu3L6]

[LaCuL2]+

[LaL]+

[LaCu(OH)L3]2-

2 4 6 8pH

0

20

40

60

80

100

% fo

rmat

ion

rela

tive

to H

o

[Ho2Cu3L6]

[Ho(OH)]2+

[HoCuL]3+

Ho3+

[Ho(HL)]2+

[HoCuL2]+[HoL]+

[HoCu(OH)L3]2-

2 4 6 8

pH

0

20

40

60

80

100

% fo

rmat

ion

rela

tive

to Y

b

Yb3+

[Yb(HL)]2+

[YbCuL]3+

[Yb2Cu3L6]

[Yb2Cu3(OH)3L6]3-

[YbCuL2]+[YbL]+

[YbL2]-

(a) (c)

(d)(b)

Fig. 5. Species distribution diagrams for ternary Ln(III)–Cu(II)–ida systems, at 25.0 �C, I = 0.50 M Me4NCl. [Ln3+] = 20 mM, [Cu2+] = 30 mM, [ida] = 60 mM. (a) Ln = La, (b)Ln = Ce, (c) Ln = Ho, (d) Ln = Yb.

Fig. 6. Overall formation constants logarithms of heteropolynuclear species vs.metal ion, at 25.0 �C, I = 0.50 M Me4NCl. Data of Sm(III)–M(II)–ida systems from [19]were also included for comparison.

220 J. Torres et al. / Journal of Molecular Structure 1004 (2011) 215–221

corresponding formation constants are in line with those previ-ously reported for other lanthanides [33]. Besides, in acidic mediaconditions the species [Ln(HL)]2+ is also detected for most systems.In the case of divalent ions, complexes with one or two ligands areformed, being those with alkaline earth metal ions the less stablespecies, as expected. Again, the formation of coordination com-pounds with partially protonated ligand is verified for low pHconditions.

Formation of ternary complexes Ln(III)–M(II)-carboxylic ligandsin solution have hitherto been very scarce, except for the oda (thedianion of oxydiacetic acid) containing systems [19]. Our presentresults with ida as a ligand are depicted in Table 4. Sm(III)–M(II)–ida system was the only example studied under the sameconditions [19]. Heterobimetallic species are found, if a lanthanideion, ida and a second bivalent metal ion are mixed in solution. Forall the studied systems, the neutral [Ln2M3(ida)6] and/or the spe-cies [Ln2M3(OH)3(ida)6]3� are formed. The neutral species havethe same stoichiometry as [Ln2M3(ida)6], isolated in the solid statefrom acidic solutions, while [Ln2M3(OH)3(ida)6]3� can be thoughtas an hydroxylated species relevant at basic pH values.

Taking M = Cu as example, the chemical speciations (Fig. 5)show the predominance of [Ln2Cu3(ida)6] near neutral pH valuesfor the stoichiometric ratios [Ln3+] = 20 mM, [Cu2+] = 30 mM,[ida] = 60 mM. As the lanthanide ion is varied, a higher percentageof the ternary species is formed in the neutral region due to anenhancement of the complexes stability with decreasing radiusof the 4f transition metal ion. With the aim of isolating the polynu-clear species, stoichiometric ratios Ln:M:ida 2:3:6 favour the pre-dominance of [Ln2M3ida6]; pH must be kept below ca. 5–6,where hydrolysis of the lanthanide ions and/or formation of

hydroxylated complexes begin. It should be mentioned that onlyusing Cu(II) as bivalent metal ion we have been able to isolatethese polynuclear species in the solid state. Even though, if another+2 metal ion (instead of copper) is used, polynuclear species arestill detected in solution. For the heavier transition metal ions,these complexes are also predominant while for alkaline earth

J. Torres et al. / Journal of Molecular Structure 1004 (2011) 215–221 221

and lighter transition metal ions the polynuclear species areformed in lower proportions, mostly due to the higher stabilityof M(II)–ida complexes relative to the polynuclear Ln(III)–M(II)–ida (Supplementary material, Fig. S4).

A comparison of the thermodynamic stability of the polynuclearcomplexes stability is shown in Fig. 6. The stability is mostly deter-mined by M(II), being the species which contain heavier transitionmetal ions more stable than those containing alkaline earth metalions. This is also reflected in the formed species in solution(Fig. S4). The Irving–Williams trend is verified in this series of com-plexes. The lanthanide ion has also an influence on the formationconstant of the heteropolynuclear complex, which is more notice-able for the alkaline earth metal ions and the lighter 3d transitionmetal ions. As expected, the smaller the lanthanide ion is, the high-er the stability constant.

4. Concluding remarks

This series of compounds have undoubtedly a tuneable channelsize, being mostly influenced by the lanthanide ionic radius. As aconsequence, the inner surface of the channel changes accordingly.The crystallization water molecules were reported to be reversiblyremoved from the channels without affecting the skeleton of thestructure, by examination of the dehydrated and rehydrated solidsthrough powder diffraction techniques [11,17]. With this in mind,these nanoporous solids offer interesting perspectives for thedevelopment of multifunctional heterobimetallic MOFs. The het-eropolynuclear species are also detected in solution. The stabilityis higher for those compounds containing smaller Ln ions andCu(II) as bivalent cation.

Acknowledgements

This work was partially supported by CSIC (Comisión Sectorial deInvestigación Científica, Uruguay), Project 653. J.G.P. thanks to Min-isterio de Ciencia e Innovación of Spain (MICCIN) under NationalProgram of Materials (MAT2010-21270-C04-02), Consolider-Ingenio 2010 Program (MALTA CSD2007-0045) and to the EU-FEDERfunds. The authors also thank Servicios Generales de la Universidadde La Laguna for providing X-ray facilities.

Appendix A. Supplementary material

CCDC 824825 and 824894 contain the supplementary crystallo-graphic data for this paper. These data can be obtained free of chargevia http://www.ccdc.cam.ac.uk/conts/retrieving.html (or from theCambridge Crystallographic Data Centre, 12, Union Road, Cambridge

CB2 1EZ, UK; fax: +441223 336033). Supplementary data associatedwith this article can be found, in the online version, at doi:10.1016/j.molstruc.2011.08.007.

References

[1] J.C.G. Bünzli, C. Piguet, Chem. Soc. Rev. 4 (2005) 1048.[2] D. Guo, C. Duan, F. Lu, Y. Hasegawa, Q. Meng, S. Yanagida, Chem. Commun.

(2004) 1486.[3] K. Hanaoka, K. Kikuchi, H. Kojima, Y. Urano, T. Nagano, Angew. Chem. Int. Ed.

42 (2003) 2996.[4] R. Rodríguez-Cortinas, F. Avecilla, C. Platas-Iglesias, D. Imbert, J.C.G. Bünzli, A.

de Blas, T. Rodríguez-Blas, Inorg. Chem. 41 (2002) 5336.[5] T. Gunnlauggsson, J.P. Leonard, K. Senechal, A.J. Harte, Chem. Commun. (2004)

782.[6] M. Fang, B. Zhao, Y. Zuo, J. Chen, W. Shi, J. Liang, P. Cheng, Dalton Trans. (2009)

7765.[7] Y. Zhou, M. Hong, X. Wu, Chem. Commun. (2006) 135.[8] X. Guo, G. Zhu, Z. Li, F. Sun, Z. Yang, S. Qiu, Chem. Commun. (2006) 3172.[9] S. Handa, V. Gnanadesikan, S. Matsunaga, M. Shibasaki, J. Am. Chem. Soc. 129

(2007) 4900.[10] C. Kremer, J. Torres, S. Domínguez, J. Mol. Struct. 879 (2008) 130.[11] Y.P. Ren, L.S. Long, B.W. Mao, Y.Z. Yuan, R.B. Huang, L.S. Zheng, Angew. Chem.

Int. Ed. 42 (2003) 532.[12] J.G. Mao, L. Song, J.S. Huang, Chin. J. Struct. Chem. 18 (1999) 4.[13] S.C. Manna, S. Konar, E. Zangrando, J. Ribas, N.R. Chaudhuri, Polyhedron 26

(2007) 2507.[14] Y.P. Cai, G.B. Li, Q.G. Zhan, F. Sun, J.G. Zhang, S. Gao, A.W. Xu, J. Solid State

Chem. 178 (2005) 3729.[15] J.G. Mao, L. Song, J.S. Huang, J. Chem. Crystallogr. 28 (1998) 475.[16] M.Y. Xu, F.H. Liao, J.R. Li, H.L. Sun, Z.S. Li, X.Y. Wang, S. Gao, Inorg. Chem.

Commun. 6 (2003) 841.[17] J.X. Ma, X.F. Huang, X.Q. Song, L.Q. Zhou, W.S. Liu, Inorg. Chim. Acta 362 (2009)

3274.[18] L.S. Long, Y.P. Ren, Y.Q. Jiang, L.F. Yang, R.B. Huang, L.S. Zheng, private

communication to the Cambridge Structural Database, deposition numberCCDC 199860, 2005.

[19] J. Torres, C. Kremer, S. Domínguez, Pure Appl. Chem. 80 (2008) 1303.[20] Molecular Structure Corporation, MSC/AFC Diffractometer Control Software

MSC 3200, Research Forest Drive, The Woodlands, TX 77381, USA, 1988.[21] G.M. Sheldrick, Acta Crystallogr. A64 (2008) 112.[22] L.J. Farrugia, J. Appl. Crystallogr. 32 (1999) 837.[23] A.L. Spek, Acta Crystallogr. A46 (1990) C-34.[24] W.J. Geary, Coord. Chem. Rev. 7 (1971) 81.[25] Agilent Technologies, CrysAlis PRO Agilent Technologies, Yarnton, England,

2010.[26] M.C. Burla, R. Caliandro, M. Camalli, B. Carrozzini, G.L. Cascarano, L. De Caro, C.

Giacovazzo, G. Polidori, D. Siliqi, R. Spagna, J. Appl. Crystallogr. 40 (2007) 609.[27] G. Schwarzenwach, H. Flaschka. Complexometric Titrations, Methuen, London,

1969.[28] S. Domínguez, J. Torres, F. Peluffo, A. Mederos, J. González-Platas, J. Castiglioni,

C. Kremer, J. Mol. Struct. 829 (2007) 57.[29] P. Gans, B. O’Sullivan, Talanta 51 (2000) 33.[30] P. Gans, A. Sabatini, A. Vacca, Talanta 43 (1996) 1739.[31] L. Alderighi, P. Gans, A. Ienco, D. Peters, A. Sabatini, A. Vacca, Coord. Chem. Rev.

184 (1999) 311.[32] R.D. Shannon, Acta Crystallogr. A32 (1976) 751.[33] IUPAC Stability Constants Database, IUPAC and Academic Software 1997, 2000.