Topological Diversity of Supramolecular Networks Constructed from

Jun 9, 2014 - *E-mail: [email protected]. ... extension of the structures [1D → 3D (1, 2), 1D → 2D (3), or 0D → 3D (4)] into distinct supramol...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/crystal

Topological Diversity of Supramolecular Networks Constructed from Copper(II) Aminoalcohol Blocks and 2,6-Naphthalenedicarboxylate Linkers: Self-Assembly Synthesis, Structural Features, and Magnetic Properties Sara S. P. Dias,† Vânia André,† Julia Kłak,‡ M. Teresa Duarte,† and Alexander M. Kirillov*,† †

Centro de Química Estrutural, Complexo I, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1049-001, Lisbon, Portugal ‡ Faculty of Chemistry, University of Wrocław, ul. F. Joliot-Curie 14, 50-383 Wroclaw, Poland S Supporting Information *

ABSTRACT: A series of copper(II) aminoalcohol derivatives [Cu2(μdmea)2(μ-nda)(H2O)2]n·2nH2O (1), [Cu2(μ-Hmdea)2(μ-nda)]n·2nH2O (2), [Cu2(μ-Hbdea)2(μ-nda)]n·2nH2O (3), and [Cu2(H4etda)2(μ-nda)]·nda· 4H2O (4) were generated by aqueous medium self-assembly at ∼25 °C from copper(II) nitrate, various aminoalcohols [N,N′-dimethylethanolamine (Hdmea), N-methyldiethanolamine (H2mdea), N-butyldiethanolamine (H2bdea), or N,N,N′,N′-tetrakis(2-hydroxyethyl)ethylenediamine (H4etda) for 1−4, respectively], sodium hydroxide, and 2,6-naphthalenedicarboxylic acid (H2nda). They were isolated as crystalline solids and fully characterized by IR and electron paramagnetic resonance spectroscopy, electrospray ionization mass spectrometry (ESI−MS(±)), thermogravimetric, elemental, and single-crystal X-ray diffraction analyses. The latter revealed that 1−3 are one-dimensional (1D) coordination polymers assembled from dicopper(II) aminoalcohol blocks and μ-nda linkers, whereas 4 is a discrete zerodimensional (0D) dimer composed of two [Cu(H4etda)]2+ fragments interlinked by the μ-nda moiety. In spite of some structural similarities, the main distinctive features of 1−4 arise from the different H-bonding patterns driven by the crystallization H2O molecules, thus giving rise to a further extension of the structures [1D → 3D (1, 2), 1D → 2D (3), or 0D → 3D (4)] into distinct supramolecular networks. Their topological analysis disclosed very complex multinodal nets with unique (1, 2, 4) or rare (3) topologies, which, upon further simplification, furnished a binodal 4,6-connected net with the sqc513 topology in 1, a trinodal 3,4,6-connected net with the 3,4,6L6 topology in 3, and trinodal 3,4,6- or 3,4,8-connected nets with undocumented topologies in 2 and 4, respectively. The magnetic susceptibility studies of 1−3 indicate a strong antiferromagnetic coupling between the CuII atoms through the μ-alkoxo bridges [J = −470(2), −100(2), and −590(1) cm−1, respectively], which was described by the Bleaney−Bowers dinuclear model. In contrast, 4 is devoid of any significant magnetic interaction within the dicopper(II) units.

■

including some state-of-the-art reviews.7,8 Although a multitude of topological types has been theoretically predicted, the search for compounds that adopt unusual or novel topologies continues to be an object of intensive research.6−10 In this regard, the H-bonded networks due to their high complexity represent a considerable potential not only to reveal new topological types but provide an identification of examples for the theoretically predicted topologies.9 The topological analysis methods applied to the hydrogen bonded networks can also give an insight on the prediction and rationalization of supramolecular motifs and, as a consequence, allow the design of desired structures.9a

INTRODUCTION The design of novel copper(II) coordination polymers and supramolecular frameworks has attracted an increased attention in the fields of crystal engineering, coordination and materials chemistry, especially in view of the structural diversity of such compounds and their promising applications as various functional materials.1,2 In particular, some copper(II) coordination polymers assembled from different aminoalcohol building blocks3−5 have found notable uses in oxidation catalysis3 and molecular magnetism,4,5 in addition to revealing some topologically unique metal−organic or hydrogen-bonded networks.3b,c In fact, the topological analysis and classification of novel coordination and supramolecular nets have recently seen a sweeping growth in crystal engineering research,6 as attested by a constantly increasing number of publications in this area, © XXXX American Chemical Society

Received: March 11, 2014

A

dx.doi.org/10.1021/cg500716d | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

complex multinodal nets, which, to our knowledge, feature novel (1, 2, 4) or rare (3) topologies, thus contributing to the identification of undocumented topological types in crystalline materials.

Bearing these points in mind and following our interest in the self-assembly synthesis of functional multicopper(II) coordination compounds,3,5 the main objective of the present work consisted of finding the influence of the type of aminoalcohol building block on the supramolecular structures and topological features of copper(II) coordination polymers driven by 2,6-naphthalenedicarboxylate linkers. Hence, we have chosen four aminoalcohols (Scheme 1) that bear one (Hdmea),

■

EXPERIMENTAL SECTION

Materials and Methods. All synthetic work was performed in air and at room temperature (r.t., ∼25 °C). All chemicals were obtained from commercial sources and used as received. C, H, and N elemental analyses were carried out by the Microanalytical Service of the Instituto Superior Técnico. Infrared spectra (4000−400 cm−1) were recorded on a JASCO FT/IR-4100 instrument in KBr pellets (abbreviations: vs − very strong, s − strong, m − medium, w − weak, br. − broad, sh. − shoulder). ESI−MS(±) spectra were run on a 500-MS LC Ion Trap instrument (Varian Inc., Alto Palo, CA, USA) equipped with an electrospray (ESI) ion source, using ca. 10−3 M solutions of 1−4 in MeCN or MeOH. The thermogravimetric analyses (TGA) were carried out on a Setaram Setsys TG−DTA 16 instrument by heating the crystalline samples (10−12 mg) of 1−4 under N2 at the 10 °C/min rate in the 30−750 °C temperature range. For simplicity, the term “aminoalcohol” was used throughout this study to denote a general type of the ligand in 1−4, understanding that in some cases the derived ligands are in fact aminoalcoholate(ato), being fully (1) or partially (2, 3) deprotonated. General Synthetic Procedure and Characterization for 1−4. To 10 mL of an aqueous 0.1 M solution of Cu(NO3)2·3H2O (1 mmol) was added 1 (1−3) or 0.5 (4) mL of an aqueous 1 M solution of aminoalcohol [N,N′-dimethylethanolamine (Hdmea; 1 mmol) for 1, N-methyldiethanolamine (H2mdea; 1 mmol) for 2, N-butyldiethanolamine (H2bdea; 1 mmol) for 3, and N,N,N′,N′-tetrakis(2hydroxyethyl)ethylenediamine (H4etda; 0.5 mmol) for 4] with continuous stirring at r.t. Then, 2,6-naphthalenedicarboxylic acid (H2nda; 108 mg, 0.50 mmol) and 3 mL of an aqueous 1 M solution of NaOH (3 mmol; up to pH ≈ 8) were added to the reaction mixture. The resulting solution was stirred for 1 day and then filtered off. The filtrate was left to evaporate in a beaker at r.t. Grayish blue (1), blue (2, 4), or cyan-blue (3) crystals (including those of X-ray quality) were formed in 1−2 weeks, then collected, and dried in air to furnish compounds 1−4 in ∼50% yield, based on aminoalcohol. [Cu2(μ-dmea)2(μ-nda)(H2O)2]n·2nH2O (1). Anal. Calcd for 1 + 0.5H2O: Cu2C20H34N2O10·0.5H2O (MW 598.6): C 40.13, H 5.89, N 4.68; found: C 40.02, H 5.61, N 4.48. Compound 1 is barely soluble in MeCN, Me2CO, and DMSO. IR (KBr, cm−1): 3379 (vs) ν(H2O), 3003 (w) and 2967 (w) νas(CH), 2886 (w) and 2858 (w) νs(CH), 1617 (vs) and 1585 (s) νas(COO), 1392 (vs) and 1362 (vs) νs(COO), 1497 (w), 1469 (w), 1326 (w), 1275 (w), 1247 (w), 1187 (m) and 1140 (w) ν(C−X) (X = C, N), 1085 (s) and 1071 (s) ν(C−O), 1023 (w), 1016 (w), 978 (w), 948 (m), 904 (m), 847 (w), 799 (s), 789 (s), 661 (w), 644 (w), 589 (w), 526 (w), 488 (w), 476 (w), 427 (w), and 418 (w). ESI−MS(±) (MeCN), selected fragments with relative abundance > 20%. MS(+): m/z: 373 (30%) [Cu2(dmea)2(H2O)2 + 2H2O − H]+, 330 (100%) [Cu(dmea)2 + 5H2O + H]+, 282 (60%) [Cu(dmea)2 + CH3CN + H]+ 152 (20%) [Cu(Hdmea)]+; MS(−): m/ z: 295 (30%) [Cu(H2O)(nda) − H]−, 251 (100%) [Hnda + 2H2O]−. [Cu 2 (μ-Hmdea) 2 (μ-nda)] n ·2nH 2 O (2). Anal. Calcd for 2: Cu2C22H34N2O10 (MW 613.6): C 43.06, H 5.59, N 4.57; found: C 43.51, H 5.67, N 4.50. Compound 2 is barely soluble in MeOH and EtOH. IR (KBr, cm−1): 3510 (m), 3408 (s br.) and 3189 (s br.) ν(H2O/OH), 2973 (w) νas(CH), 2861 (m) νs(CH), 1615 (s) and 1582 (s) νas(COO), 1386 (s) and 1361 (vs) νs(COO), 1496 (w), 1460 (w), 1190 (m) ν(C−X) (X = C, N), 1088 (s) ν(C−O), 1656 (w), 1139 (w), 1031 (w), 993 (m), 939 (w), 902 (m), 793 (s), 646 (m), 598 (w), 522 (w), and 479 (m). ESI−MS(±) (MeOH), selected fragments with relative abundance >20%, MS(+): m/z: 361 (20%) [Cu2(Hmdea)2 − H]+, 274 (25%) [Cu(mdea)2 + MeO]+, 120 (40%) [H2mdea + H]+; MS(−): m/z: 215 (100%) [Hnda]−. [Cu 2 (μ-Hbdea) 2 (μ-nda)] n ·2nH 2 O (3). Anal. Calcd for 1: Cu2C28H46N2O10 (MW 697.8): C 48.20, H 6.64, N 4.01; found: C 48.34, H 6.60, N 3.97. Compound 3 is barely soluble in MeOH and

Scheme 1. Structural Formulae of Aminoalcohols

two (H2mdea, H2bdea), or four (H4etda) hydroxyethyl arms, which despite their aqueous solubility, coordination versatility, and commercial availability are still poorly explored in the construction of copper(II) coordination polymers,4a,b,5 as confirmed by the search of the Cambridge Structural Database (CSD).11 Besides, the hydroxyethyl functionalities of these aminoalcohols can participate in extensive H-bonding interactions, thus making them good building blocks toward the generation of supramolecular networks. In spite of possessing two distant carboxylic groups and two conjugated aromatic rings that can give an extra rigidity and stability to resulting network structure, 2,6-naphthalenedicarboxylic acid has been scarcely applied for the synthesis of copper coordination polymers in contrast to other well explored dicarboxylate linkers (e.g., terephthalic or isophthalic acids)11 and thus was selected as a linker in the present study. Hence, we have tested the above-mentioned aminoalcohols in a series of aqueous-medium self-assembly reactions with copper(II) nitrate and 2,6-naphthalenedicarboxylic acid (H2nda), resulting in the generation of the one-dimensional (1D) coordination polymers [Cu2(μ-dmea)2(μ-nda)(H2O)2]n· 2nH2O (1), [Cu2(μ-Hmdea)2(μ-nda)]n·2nH2O (2), and [Cu2(μ-Hbdea)2(μ-nda)]n·2nH2O (3), and the discrete dimer [Cu2(H4etda)2(μ-nda)]·nda·4H2O (4), which are extended by multiple hydrogen bonds into the complex three- or twodimensional (3D or 2D) supramolecular networks with rare or unique topologies. The present study thus reports the synthesis and characterization of 1−4, as well as their structural features and magnetic properties. The novelty of the current work with respect to prior studies concerns the following points. (A) New copper(II) crystalline materials have been prepared by an easy self-assembly method in aqueous medium from different aminoalcohols and 2,6naphthalenedicarboxylic acid, which are still poorly explored building blocks, whereas their combined synthetic application has not yet been reported. (B) In spite of structural resemblance, the 1D coordination polymers 1−3 exhibit distinct magnetic properties with antiferromagnetic coupling constants within the dicopper(II) blocks ranging from −100(2) cm−1 in 2 to −590 cm−1 in 3. (C) The work also shows how a slight modification of aminoalcohol plays a key role in defining the intricate 3D or 2D H-bonded structures in 1−4, which are primarily driven by the presence of crystallization water molecules or different water clusters. (D) The topological analysis of supramolecular networks in 1−4 has disclosed very B

dx.doi.org/10.1021/cg500716d | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

Table 1. Crystal Data and Structure Refinement Details for Compounds 1−4 formula fw crystal form, color crystal size (mm) crystal syst. space group a, Å b, Å c, Å α, deg β, deg γ, deg Z V, Å3 T, K Dc, g cm−3 μ(Mo Kα), mm−1 θ range (deg) refl collected independent refl Rint R1a, wR2b [I ≥ 2σ(I)] GOF on F2 a

1

2

3

4

C10H17CuNO5 294.79 plate, blue 0.26 × 0.12 × 0.04 monoclinic P21/c 12.7095(6) 8.7854(6) 12.0498(7) 90.00 101.689(2) 90.00 4 1317.55(13) 303(2) 1.486 1.666 2.84−30.65 19 234 4080 0.0336 0.0428, 0.0973 1.069

C11H17CuNO5 306.80 block, blue 0.20 × 0.18 × 0.10 monoclinic P21/c 12.1134(10) 9.2965(8) 12.0316(10) 90.00 100.350(2) 90.00 4 1332.86(19) 273(2) 1.529 1.650 3.42−30.01 31 544 3840 0.0642 0.0398, 0.0978 1.054

C14H23CuNO5 348.87 plate, blue 0.18 × 0.10 × 0.03 triclinic P1̅ 7.7386(8) 8.5225(8) 13.6748(15) 85.096(5) 85.617(5) 63.198(4) 2 801.32(14) 293(2) 1.446 1.382 2.97−25.47 12 976 2980 0.0683 0.0567, 0.1397 1.055

C44H68Cu2N4O20 1100.12 block, blue 0.18 × 0.16 × 0.06 triclinic P1̅ 8.7497(1) 9.2024(2) 16.6861(2) 88.893(1) 79.897(1) 64.008(1) 1 1186.40(3) 293(2) 1.540 0.981 2.47−32.58 41 046 8597 0.0447 0.0434, 0.1115 1.128

R1 = Σ∥F0| − |Fc∥/Σ|F0|. bwR2 = [Σ[w(F02 − Fc2)2]/Σ[w(F02)2]]1/2.

Scheme 2. Simplified Representation of the Self-Assembly Synthesis and Structural Formulae of 1−4

MeCN. IR (KBr, cm−1): 3449 (s br), 3370 (sh.) and 3188 (m br.) ν(H2O/OH), 2953 (m) and 2922 (w) νas(CH), 2848 (m) νs(CH), 1609 (s) and 1582 (s) νas(COO), 1385 (vs) and 1352 (s) νs(COO), 1490 (w), 1462 (w), 1191 (w) ν(C−X) (X = C, N), 1088 (m) ν(C− O), 1035 (w), 979 (w), 900 (w), 783 (m), 662 (w), 582 (w), 559 (w), and 481 (w). ESI−MS(±) (MeOH), selected fragments with relative abundance >10%, MS(+): m/z: 445 (10%) [Cu2(Hbdea)2 − H]+, 384 (10%) [Cu(Hbdea)2 + H]+, 162 (100%) [H2bdea + H]+; MS(−): m/ z: 215 (100%) [Hnda]−.

[Cu 2 (H 4 etda) 2 (μ-nda)]·nda·4H 2 O (4). Anal. Calcd for 4: Cu2C44H68N4O20 (MW 1100.1): C 48.04, H 6.23, N 5.09; found: C 48.34, H 6.23, N 5.06. Compound 4 is barely soluble in MeCN, Me2CO, and DMSO. IR (KBr, cm−1): 3457 (vs br.) ν(H2O/OH), 2975 (w) νas(CH), 2902 (w) νs(CH), 1605 (s) and 1574 (s) νas(COO), 1395 (vs) and 1356 (s) νs(COO), 1494 (w), 1456 (w), 1195 (m) and 1262 (w) ν(C−X) (X = C, N), 1061 (s) ν(C−O), 1148 (w), 926 (w), 875 (w), 797 (m), 763 (w), 737 (w), 646 (w), and 487 (w). ESI−MS(±) (MeCN), selected fragments with relative C

dx.doi.org/10.1021/cg500716d | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

abundance >10%, MS(+): m/z: 362 (10%) [Cu2(Hetda)]+, 278 (100%) [Cu(nda) + H]+, 237 (35%) [H4etda + H]+; MS(−): m/z: 251 (100%) [Hnda + 2H2O]−. X-ray Crystallography. Crystals of 1−4 suitable for X-ray diffraction study were mounted with Fomblin in a cryoloop. Data were collected on a Bruker AXS-KAPPA APEX II diffractometer with graphite-monochromated radiation (Mo Kα, λ = 0.17073 Å) at 273− 303 K. The X-ray generator was operated at 50 kV and 30 mA, and the X-ray data collection was monitored by the APEX212 program. All data were corrected for Lorentzian, polarization, and absorption effects using SAINT12 and SADABS12 programs. SIR9713 and SHELXS-97 were used for structure solution and SHELXL-9714 was applied for full matrix least-squares refinement on F2. These three programs are included in the package of programs WINGX-Version 1.80.05.15 Nonhydrogen atoms were refined anisotropically. A full-matrix leastsquares refinement was used for the non-hydrogen atoms with anisotropic thermal parameters. All the hydrogen atoms were inserted in idealized positions and allowed to refine in the parent carbon or oxygen atom. TOPOS 4.06 was used for packing diagrams. PLATON16 was used for hydrogen bond interactions. Crystal data and details of data collection for 1−4 are reported in Table 1. Magnetic Studies. The magnetization of powdered samples 1−4 was measured over the 1.8−300 K temperature range using a Quantum Design SQUID-based MPMSXL-5-type magnetometer. The superconducting magnet was generally operated at a field strength ranging from 0 to 5 T. Sample measurements were made at magnetic field 0.5 T. The SQUID magnetometer was calibrated with the palladium rod sample. Corrections are based on subtracting the sample−holder signal, and the χD contribution was estimated from the Pascal’s constants.17 EPR spectra of powdered samples 1−4 were recorded at 293 and 77 K on a Bruker ESP 300 spectrometer operating at X-band equipped with an ER 035 M Bruker NMR gaussmeter and HP 5350B Hewlett-Packard microwave frequency counter.

several Cu-aminoalcohol fragments obtained upon removal of nda ligands. Thus, the heaviest [Cu2(dmea)2(H2O)2 + 2H2O − H]+ (m/z 373), [Cu2(Hmdea)2 − H]+ (m/z 361) and [Cu2(Hbdea)2 − H]+ (m/z 361) fragments with the expected isotopic distribution pattern were detected in 1−3, respectively. In these compounds, other characteristic peaks correspond to further fragmentation with the loss of one aminoalcohol moiety and/or one Cu atom. The ESI−MS(+) spectrum of 4 exhibits the characteristic [Cu2(Hetda)]+ (m/z 362) and [Cu(nda) + H]+ (m/z 278) peaks. The ESI−MS(−) plots of all compounds are less informative, showing the highly intense peaks due to the [Hnda + 2H2O]− (m/z 251 in 1 and 4) or [Hnda]− (m/z 215 in 2 and 3) fragments. Description of the Crystal Structures. The crystal structure of the linear 1D coordination polymer 1 is based on a dicopper(II) [Cu2(μ-dmea)2(H2O)2]2+ block, a μnda(2−) linker, and two crystallization H2O molecules (Figures 1a and 2a). The five-coordinate Cu1 atoms adopt a distorted {CuNO4} square-pyramidal environment (τ5 = 0.28),20 filled by the N1, O3, and O3i atoms of bidentate μ-dmea moieties [Cu1−N1 2.0417(19), Cu1−O3 1.9293(15), Cu1−O3 i 1.9063(16) Å] and the O1 atom of the μ-nda linker [Cu1− O1 1.9384(15) Å] in equatorial sites, whereas an axial position is taken by the O4 atom of the H2O ligand [Cu1−O4 2.409(2) Å]. The {Cu2(μ-O)2} core is planar and possesses a rather short Cu1···Cu1i separation of 2.9775(5) Å, while the Cu1−μnda−Cu1ii distance equals 12.9626(6) Å. In general, the bonding parameters in 1 are in agreement with those found in other structurally related dicopper(II) aminoalcohol derivatives.5,19 Similarly to 1, the structures of 2 (Figures 1b and 2b) and 3 (Figure 1c) also feature the 1D metal−organic chains assembled from the related [Cu2(μ-Hmdea)2]2+ or [Cu2(μHbdea)2]2+ blocks and μ-nda linkers. However, in these dicopper(II) blocks the axial site around the Cu atoms is not occupied by the H2O ligand, but by the OH group from an additional CH2CH2OH arm of the aminoalcohol moiety. Given the overall structural resemblance of 2 and 3 to 1, these structures are not further discussed. In contrast to the 1D coordination polymers 1−3, compound 4 reveals a discrete 0D structure composed of the dicopper(II) [Cu2(H4etda)2(μ-nda)]2+ cation, the nda(2−) anion, and four crystallization H2O molecules (Figure 1d). The cation bears two symmetry equivalent {Cu(H4etda)}2+ blocks that are interconnected by the bridging 2,6-naphthalenedicarboxylate moiety. The six-coordinate Cu1 atoms feature a distorted octahedral {CuN2O4} environment filled by the carboxylate oxygen atom (O1) of μ-nda and a pentadentate H4etda ligand having one of the hydroxyethyl groups (O3) uncoordinated. The equatorial sites around the Cu1 centers are taken by the N1, N2, and O6 atoms of H4etda [Cu1−N1 2.0860(14), Cu1− N2 2.0074(15), Cu1−O6 2.0390(14) Å] and the O1 atom of μ-nda [Cu1−O1 1.931(1) Å], whereas the apical positions are occupied by the O4 and O5 atoms of the aminoalcohol ligand [Cu1−O4 2.474(1), Cu1−O5 2.3140(14) Å]. In general, the bonding parameters in 4 are comparable to those of other Cu(II) compounds derived from H4etda.21 The Cu1···Cu1i separation of 13.3260(3) Å between the adjacent monocopper(II) blocks in 4 is slightly superior than the corresponding distances [12.8398(9)−13.1293(14) Å] between the dicopper(II) units in 1−3. All the compounds 1−4 show an extensive hydrogen bonding pattern driven by the presence of crystallization water molecules (see Table 2 for hydrogen bonding details),

■

RESULTS AND DISCUSSION Synthesis and Spectroscopic Characterization. A simple combination, in water at ∼25 °C in air, of copper(II) nitrate as a metal source with different aminoalcohols [N,N′dimethylethanolamine (Hdmea), N-methyldiethanolamine (H2mdea), N-butyldiethanolamine (H2bdea), or N,N,N′,N′tetrakis(2-hydroxyethyl)ethylenediamine (H4etda)] as main building blocks and 2,6-naphthalenedicarboxylic acid (H2nda) as a linker, followed by alkalization of the obtained reaction mixture with sodium hydroxide as a pH-regulator, gave rise to the self-assembly generation of four novel copper(II) aminoalcohol products (Scheme 2). These are 1D coordination polymers [Cu 2 (μ-dmea) 2 (μ-nda)(H 2 O) 2 ] n ·2nH 2 O (1), [Cu2(μ-Hmdea)2(μ-nda)]n·2nH2O (2), and [Cu2(μ-Hbdea)2(μ-nda)]n·2nH2O (3), as well as a discrete 0D dimer [Cu2(H4etda)2(μ-nda)]·nda·4H2O (4). They were isolated as crystalline solids and fully characterized by IR and EPR spectroscopy, ESI−MS(±), elemental, and single-crystal X-ray diffraction analyses. The IR spectra of 1−4 show comparable features due to the presence of copper(II) aminoalcohol blocks, aromatic carboxylate ligands, and crystallization water molecules.5,18,19 Hence, the characteristic vibrations include one or two ν(H2O)/ ν(OH) bands with maxima in the 3510−3185 cm−1 range due to H2O molecules and OH groups of aminoalcohol ligands, the broad character of which is indicative of intensive hydrogen bonding. The νas and νs CH vibrations are observed as two to four typically weak bands in the 3000−2845 cm−1 range. Besides, the spectra of 1−4 also reveal two pairs of strong νas(COO) [1617−1605 and 1585−1574 cm−1] and νs(COO) [1395−1385 and 1352−1362 cm−1] bands of 2,6-naphthalenedicarboxylate moieties. The ESI−MS(+) plots of 1−3 show D

dx.doi.org/10.1021/cg500716d | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

by Infantes et al.22 These {(H2O)4}n aggregates are further Hbound [O1W−H1W···O1 2.875(3), O1W−H2W···O2 2.741(3) Å] to the carboxylate groups of nda, thus giving rise to a 3D supramolecular network. To better understand the structure of this intricate H-bonded network, we have carried out its topological analysis following the concept of the simplified underlying net.6−8 Hence, all the ligands and crystallization solvent molecules have been contracted to their centroids, resulting in a tetranodal network (Figure S1, Supporting Information) that is composed of the 3- and 4connected H2O nodes (O4 and O1W, respectively), 4connected Cu1 and 6-connected nda nodes, as well as 2connected dmea linkers. The topological analysis6 of this 3,3,4,6-connected net reveals an undocumented topology described by the point symbol of (4.5.7)2(4.5.84)2(4.52)2(42.52.72.82.104.112.13), wherein the (4.5.7), (4.5.84), (4.52), and (42.52.72.82.104.112.13) notations correspond to the H2O (O4), H2O (O1W), Cu1, and nda nodes, respectively. Besides, the 3D network in 1 can be simplified further by treating the [Cu2(μ-dmea)2(H2O)2]2+ blocks as the 6-connected cluster nodes which, along with the 6-connected nda and 4-connected H2O (O1W) nodes, result in a binodal 4,6-connected net (Figure 3b) with a very rare sqc513 topology (EPINET database).23 It is described by the point symbol of (32.84)(34.42.84.94.10), wherein the (32.84) and (34.42.84.94.10) indices are those of the H2O (O1W) nodes, and topologically equivalent [Cu2(μ-dmea)2(H2O)2]2+ and nda nodes, respectively. The 3D H-bonding pattern in the structure of 2 differs from that of 1, being driven by the crystallization H2O (O1W) molecules that act as H-bond donors to the O1 and O2 atoms of nda moieties [O1W−H1W···O2 3.000(3), O1W−H2W···O1 2.822(3) Å] and H-bond acceptors from the OH groups of Hmdea ligands [O4−H001···O1W 2.688(3) Å] (Figure 4a). To get further insight into the resulting supramolecular network, we have also analyzed it from the topological viewpoint. Thus, an underlying network generated after the simplification procedure6−8 is assembled from the 3-connected H2O and topologically equivalent Cu1 and Hmdea nodes, as well as 6-connected nda nodes (Figure S2, Supporting Information). This trinodal 3,3,6-connected 3D net features an unreported topology defined by the point symbol of (4.8.10)2(42.6)4(42.84.104.124.14), wherein the (4.8.10), (42.6), and (42.84.104.124.14) notations correspond to the H2O, Cu1/ Hmdea, and nda nodes, respectively. Further simplification of this network by considering the [Cu2(μ-Hmdea)2]2+ units as the 4-connected cluster nodes gives rise to an alternative trinodal 3,4,6-connected net (Figure 4b). However, its topological analysis also reveals an undocumented topology with the point symbol of (3.8.9)2(32.84.95.102.112)(32.93.10) and the (3.8.9), (32.84.95.102.112), and (32.93.10) indices matching the H2O, nda, and [Cu2(μ-Hmdea)2]2+ nodes, respectively. The novelty of the present type of topology has been confirmed by a search of different databases.6,8,11,23 In contrast to 1 and 2, the 1D metal−organic chains of 3 are extended into a simpler 2D supramolecular network, via the intermolecular H-bonds [O1W−H1W···O2 2.741(7), O1W− H2W···O4 2.811(6), O4−H42···O1W 2.703(8) Å] involving the crystallization H2O (O1W) molecules and nda (O2) or Hbdea (O4) moieties (Figure 5a). A simplified underlying layer has been generated by reducing all the ligands and solvent molecules to their centroids (Figure S3, Supporting Information). From the topological viewpoint, it can be considered as a

Figure 1. X-ray crystal structures of 1 (a), 2 (b), 3 (c), and 4 (d) with atom numbering scheme. H atoms (except those of H2O or OH moieties), crystallization H2O molecules, and nda2− anion (in 4) are omitted for clarity. Color codes: Cu balls (green), O (red), N (blue), C (gray), and H (dark gray). Selected distances (Å) and angles (deg): 1: Cu1−N1 2.0417(19), Cu1−O1 1.9384(15), Cu1−O3 1.9293(15), Cu1−O3i 1.9063(16), Cu1−O4 2.409(2), Cu1···Cu1i 2.9775(5), Cu1···Cu1ii 12.9626(6), O3i−Cu1−O3 78.16(8), Cu1i−O3−Cu1 101.84(8); 2: Cu1−N1 2.073(2), Cu1−O2 1.9211(15), Cu1−O3 1.9338(19), Cu1−O3i 1.8968(17), Cu1−O4 2.332(2), Cu1···Cu1i 2.9341(5), Cu1···Cu1ii 12.8398(9), O3i−Cu1−O3 80.02(9), Cu1i− O3−Cu1 99.98(9); 3: Cu1−N1 2.064(3), Cu1−O1 1.928(3), Cu1− O3 1.905(3), Cu1−O3i 1.921(3), Cu1−O4 2.394(4), Cu1···Cu1i 3.0033(10), Cu1···Cu1ii 13.1293(14), O3i−Cu1−O3 76.58(14), Cu1i−O3−Cu1 103.42(14); 4: Cu1−N1 2.0860(14), Cu1−N2 2.0074(15), Cu1−O1 1.931(1), Cu1−O4 2.474(1), Cu1−O5 2.3140(14), Cu1−O6 2.0390(14), Cu1···Cu1i 13.3260(3). Symmetry codes: 1: (i) −x+1, −y+2, −z; (ii) −x, 1−y, −z; 2: (i) −x+1, −y, −z +1; (ii) −x, −y+1, −z+1; 3: −x, −y+1, −z+2; (ii) −x, −y+2, −z+1; 4: (i) −x+2, −y+1, −z+1.

leading to a further extension of the structures [1D → 3D (1, 2), 1D → 2D (3), or 0D → 3D (4)] into distinct H-bonded networks. The topological analysis of these intricate networks has revealed a number of topologically unique multinodal nets that are described below in detail. Topological Analysis of H-Bonded Networks in 1−4. An interesting feature of 1 consists in the formation of infinite 1D {(H2O)4}n water chain motifs (Figure 3a), via the repeating intermolecular H-bonds [O4−H3W···O1W 2.775(3), O4− H4W···O1W 2.990(4) Å] between the H2O ligands (O4) and crystallization water (O1W) molecules, which can be classified within the C4 type according to the systematization introduced E

dx.doi.org/10.1021/cg500716d | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

Figure 2. Structural fragments of 1 (a) and 2 (b) showing linear 1D metal−organic chains with polyhedral representation of the coordination environments around Cu atoms. H atoms and crystallization H2O molecules are omitted for clarity. Color codes are those of Figure 1.

Table 2. Hydrogen Bonding Details for Compounds 1−4 1

2

3

4

sym. op.

D−H···A

d(D−H) (Å)

d(H···A) (Å)

d(D···A) (Å)

DHA (deg)

x, 3/2−y, −1/2+z x, y, z 1−x, 2−y, −z x, 3/2−x, 1/2+z x, y, z 1−x, −1/2+y, 3/2−z 1−x, −y, 1−z 1+x, y, z 1−x, 1−y, 2−z x, y, z x, y, z 1−x, 2−y, −z x, y, z x, 1+y, z x, 1+y, z x, −1+y, z x, y, z −1+x, y, z

O1w−H1w···O1 O1w−H2w···O2 O4−H3w···O1w O4−H4w···O1w O4−H001···O1w O1w−H1w···O2 O1w−H2w···O1 O1w−H1w···O2 O1w−H2w···O4 O4−H42···O1w O6−H001···O8 O1w−H1w···O7 O5−H002···O7 O1w−H2w···O8 O3−H3A···O2 O2w−H3w···O1w O4−H40···O2 O2w−H4w···O7

0.86(2) 0.88(3) 0.87(4) 0.78(4) 0.74(5) 0.83(3) 0.85(4) 0.86 0.86 0.83(14) 0.72(3) 0.86 0.78(3) 0.86 0.82 0.87 0.82 0.86

2.09(3) 1.88(3) 1.91(3) 2.31(4) 1.96(5) 2.21(3) 1.97(4) 1.91 1.97 1.88(13) 1.88(3) 2.17 1.89(3) 2.01 2.23 2.01 1.94 2.21

2.875(3) 2.741(3) 2.775(3) 2.990(4) 2.688(3) 3.000(3) 2.822(3) 2.741(7) 2.811(6) 2.703(8) 2.588(2) 2.972(3) 2.6544(19) 2.873(3) 2.992(2) 2.873(4) 2.756(2) 3.057(4)

152(3) 167(3) 174(4) 147(4) 167(5) 159(3) 174(4) 163 163 170(18) 172(3) 157 166(4) 179 155 176 172 167

Figure 3. Structural fragments of 1. (a) H-bonding pattern showing the formation of {(H2O)4}n chain motifs (view along the a axis); H atoms apart from those of H-bonds are omitted for clarity; color codes are those of Figure 1. (b) Topological representation of the simplified binodal 4,6connected 3D net (arbitrary view) generated after treating the [Cu2(μ-dmea)2(H2O)2]2+ blocks as cluster nodes, which adopts the sqc513 topology with the point symbol of (32.84)(34.42.84.94.10); color codes: centroids of 6-connected [Cu2(μ-dmea)2(H2O)2]2+ (green) and nda (gray) nodes, centroids of 4-connected H2O (O1W) nodes (red).

(topologically equivalent), and 4-connected nda and Hbdea nodes, respectively. An alternative topological classification of this network can be performed by considering the [Cu2(μHbdea)2]2+ blocks as the 6-connected cluster nodes, thus

trinodal 3,4,4-connected net with an uncommon 3,4,4L15 topology6,24 and the point symbol of (42.6)4(42.82.102)(43.62.8)2. The (42.6), (42.82.102), and (43.62.8)2 indices correspond to the 3-connected H 2 O and Cu1 nodes F

dx.doi.org/10.1021/cg500716d | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

Figure 4. Structural fragments of 2. (a) H-bonding pattern driven by crystallization H2O (O1W) molecules (arbitrary view); H atoms apart from those of H-bonds are omitted for clarity; color codes are those of Figure 1. (b) Topological representation of the simplified trinodal 3,4,6-connected 3D net (view along the b axis) generated after treating the [Cu2(μ-Hmdea)2]2+ blocks as cluster nodes, which reveals a unique topology defined by the point symbol of (3.8.9)2(32.84.95.102.112)(32.93.10); color codes: centroids of 4-connected [Cu2(μ-Hmdea)2]2+ nodes (green), centroids of 6connected nda (gray) and 3-connected H2O (red) nodes.

Figure 5. Structural fragments of 3. (a) H-bonding pattern driven by crystallization H2O (O1W) molecules (view along the b axis); H atoms apart from those of H-bonds are omitted for clarity; color codes are those of Figure 1. (b) Topological representation of the simplified trinodal 3,4,6connected 2D net (view along the b axis) generated after treating the [Cu2(μ-Hbdea)2]2+ blocks as cluster nodes, which adopts the 3,4,6L6 topology with the point symbol of (3.4.5)2(32.42.52.62.74.82.9)(32.62.72); color codes: centroids of 6-connected [Cu2(μ-Hbdea)2]2+ nodes (green), centroids of 4-connected nda (gray) and 3-connected H2O (red) nodes.

hydrogen-bonded nets assembled from the discrete molecular units.6 Thus, the [Cu2(H4etda)2(μ-nda)]2+ dimers, nda(2−) anions, and (H2O)2 clusters have been contracted to their centroids, leading to an underlying 3D net built from the 4connected [Cu2(H4etda)2(μ-nda)]2+, 8-connected nda(2−), and 3-connected (H2O)2 nodes (Figure 6b). Its topological analysis discloses a trinodal 3,4,8-connected network with a novel topology described by the point symbol of (43)2 (46.58.67.74.83)(54.82), wherein the (43), (46.58.67.74.83), and (54.82) indices are those of the (H2O)2, nda(2−), and [Cu2(H4etda)2(μ-nda)]2+ nodes, respectively. Magnetic Studies. The magnetic properties of 1−4 were investigated over the temperature range of 1.8−300 K. Plots of magnetic susceptibility χmT product vs T (χm is the molar magnetic susceptibility for two CuII ions) are given in Figure 7. For 2, the experimental χmT value at room temperature (0.631 cm3 mol−1 K) is smaller than the theoretical one expected for the two uncoupled copper(II) ions [χmT = 2(Nβ2g2/3k)S(S +

furnishing a trinodal 3,4,6-connected net with the 3,4,6L6 topology (Figure 5b).6,25 It is defined by the point symbol of (3.4.5)2(32.42.52.62.74.82.9)(32.62.72), wherein the (3.4.5), (32.42.52.62.74.82.9), and (32.62.72) notations are those of the 3-connected H2O, 6-connected [Cu2(μ-Hbdea)2]2+, and 4connected nda nodes, respectively. The compound 4 features a very extensive H-bonding pattern driven by crystallization H2O molecules (O1W, O2W) [O1W−H1W···O7 2.972(3), O1W−H2W···O8 2.873(3), O2W−H3W···O1W 2.873(4), O2W−H4W···O7 3.057(4) Å] and three OH groups of the H4etda ligands [O3−H3A···O2 2.992(2), O5−H002···O7 2.6544(19), O6−H001···O8 2.588(2) Å], resulting in the generation of a complex 3D Hbonded framework (Figure 6a). Interestingly, the crystallization water molecules O1W and O2W form the discrete (H2O)2 clusters that are classified as the D2 type.22 For the sake of topological systematization, the H-bonded framework of 4 has been simplified using a method developed for the analysis of G

dx.doi.org/10.1021/cg500716d | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

Figure 6. Structural fragments of 4. (a) H-bonding pattern showing the interlinkage of two adjacent [Cu2(H4etda)2(μ-nda)]2+ units with nda anions and (H2O)2 clusters (view along the a axis); H atoms apart from those of H-bonds are omitted for clarity; color codes are those of Figure 1. (b) Topological representation of the simplified trinodal 3,4,8-connected 3D net (view along the a axis) generated after treating the [Cu2(H4etda)2(μnda)]2+ blocks as molecular nodes, which reveals a unique topology defined by the point symbol of (43)2(46.58.67.74.83)(54.82); color codes: centroids of 4-connected [Cu2(H4etda)2(μ-nda)]2+ nodes (green), centroids of 8-connected nda (gray) and 3-connected (H2O)2 (red) nodes.

respect to the field (H), at 2 K, also confirms the nature of the ground state in 1−4 (Figure S5, Supporting Information). In linear 1D coordination polymers 1−3, the Cu−μ-nda−Cu separations are significantly longer compared to the Cu···Cu distances within the dicopper(II) aminoalcohol units. Likewise, a long Cu−μ-nda−Cu distance in 4 points out a total magnetic isolation of the adjacent monocopper(II) blocks. Under this approach, an interaction via the Cu−μ-nda−Cu path in 1−3 is neglected, and the strong antiferromagnetic coupling is considered to take place through the μ-O-alkoxo bridges; the compounds thus behave as antiferromagnetically coupled dicopper(II) derivatives. Hence, the magnetic data of 1−3 were analyzed using the isotropic spin Hamiltonian H = −JS1⃗ · S⃗2 that results in the Bleaney−Bowers expression for two interacting spin doublets, eq 1.26,27 The paramagnetic monomeric impurity ρ was introduced in eq 1 to obtain satisfactory results.

Figure 7. Temperature dependence of experimental χmT (χm per 2 CuII atoms) for 1−4. The solid lines (for 1−3) are the calculated curves derived from eq 1.

χm =

1) = 0.749 cm3 mol−1K, wherein g = 2.1 is the spectroscopic splitting factor, N − Avogadro’s number, β − the Bohr magneton, k − Boltzmann’s constant, and S = 1/2].26 The compound 2 exhibits the typical behavior of an antiferromagnetically coupled dinuclear copper(II) derivative with a continuous decrease of the χmT value on lowering the temperature, and with a diamagnetic ground state almost completely populated below 30 K. For 1 and 3, the χmT values at 300 K are much below those expected for two magnetically isolated spin doublets, being equal to 0.371 and 0.277 cm3 mol−1 K, respectively. The χmT plots continuously decrease on lowering the temperature reaching almost zero below 100 K. These features are indicative of a strong antiferromagnetic coupling between the copper(II) ions in these compounds, for which the diamagnetic ground state is completely populated below 100 K. Additionally, the susceptibility curves for 1−3 exhibit the maximum that confirms the presence of a strong antiferromagnetic ordering with a Neel temperature (TN) about 300 K (for 1 and 3) and 80 K (for 2) (Figure S4, Supporting Information). For 4, χmT is essentially constant [∼0.774 cm3 mol−1 K] in the whole temperature range. It is consistent with two unpaired electrons in magnetically diluted copper(II) complexes.26 The variation of the magnetization (M) with

2Nβ 2g 2 Nβ 2g 2 (1 − ρ) + ρ kT[3 + exp( −J /kT ) 2kT

(1)

In eq 1, J is the intradimer exchange coupling constant, ρ stands for the molar fraction of the monomeric form, while the other parameters have their usual meaning. A least-squares fitting of the experimental data leads to the following values: J = −470(2) cm−1, g = 2.20(1), ρ = 0.075, and R = 2.14 × 10−4 for 1, J = −100(2) cm−1, g = 2.10(1), ρ = 0.052, and R = 5.38 × 10−5 for 2, and J = −590(1) cm−1, g = 2.15(2), ρ = 0.042, and R = 5.72 × 10−5 for 3. The calculated curves reproduce the magnetic data very well in the whole temperature range (Figure 7). The criterion applied to determine the best fit was based on the minimization of the sum of squares of the deviation, R = Σ(χexpT − χcalcT)2/Σ(χexpT)2. Results of the magnetic data calculations for 1−3 indicate a very strong (1, 3) or moderately strong (2) antiferromagnetic coupling between the copper(II) ions transmitted through the μ-alkoxo bridge. The EPR spectra of solid samples 1−4 at 298 and 77 K additionally confirm the properties detected by the direct magnetic measurements (Figures S6 and S7, Supporting Information). The nature and the strength of the magnetic exchange coupling in alkoxo-bridged dicopper(II) complexes are mainly affected by the Cu···Cu distance and the Cu−O−Cu bridging H

dx.doi.org/10.1021/cg500716d | Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design

Article

2,6-naphthalenedicarboxylate blocks, thus providing the first examples of Cu(II) derivatives that combine together these types of organic ligands.11 The X-ray crystal structures of 1−4 reveal multiple intermolecular H-bonding interactions that are primarily driven by the presence of crystallization H2O molecules or different water clusters, thus leading to the extension of the 1D or 0D metal−organic structures into distinct and rather complex 3D (1, 2, 4) or 2D (3) supramolecular networks. Their characteristic features are strongly influenced by the type of aminoalcohol building block (Hdmea, H2mdea, H2bdea, or H4etda), which plays a crucial structure-driven role in defining the dimensionality and topology of the resulting supramolecular assembly. In fact, the topological analysis of the underlying networks in 1−4 has revealed distinct and complex multinodal nets with novel (1, 2, 4) or rare topologies (3). These have undergone further simplification resulting in (i) a binodal 4,6-connected net with the sqc513 topology in 1, (ii) a trinodal 3,4,6connected net with the 3,4,6L6 topology in 3, and (iii) topologically unique trinodal 3,4,6- or 3,4,8-connected nets in 2 and 4, respectively. Their novelty has been confirmed by a search of different databases.6,8,11,23 Hence, the present study also contributes to the identification and classification of supramolecular networks with novel topologies, which have not been reported so far in crystalline materials.6−9 Besides, the infinite 1D {(H2O)4}n water chain motifs and discrete (H2O)2 clusters have been encountered in the crystal structures of 1 and 4, respectively, showing that such types of compounds can be suitable matrices to host various water assemblies, which, in turn, are responsible for the extension of the structures into topologically unique supramolecular networks. Moreover, apart from presenting notable topological features, some of the obtained compounds exhibit interesting and somewhat distinct magnetic properties. In fact, the magnetic study revealed that 1−3 display very strong (1, 3) or moderately strong (2) antiferromagnetic couplings that can be explained by a Bleaney−Bowers expression. The magnetic coupling constants appear to significantly depend on the type of aminoalcohol ligand used and are in good agreement with the general trend for alkoxo-bridged dicopper(II) derivatives. In summary, this work contributes toward (i) broadening the diversity of copper(II) coordination polymers assembled from aminoalcohol building blocks and aromatic dicarboxylate linkers, namely, furnishing the first examples of Cu(II) compounds that simultaneously bear any aminoalcohol ligand and 2,6-naphthalenedicarboxylate linker, (ii) discerning unprecedented topological types in the resulting supramolecular networks, and (iii) studying their magnetic properties. Further exploration of various aminoalcohols toward the design and application of novel copper based coordination polymers or metal−organic frameworks is currently in progress.

angle (α).28,29 In general, the antiferromagnetic coupling is observed for large Cu−O−Cu bond angles and long copper··· copper separations, whereas the ferromagnetic interaction is favored for small Cu−O−Cu bond angles (