Comment

The rational design and construction of novel functional metal complexes have attracted much attention, owing to their intriguing architectures and topologies (Ye & Tong et al., 2005; Hill et al., 2005; Hong, 2007) and their potential applications in many fields, including ferroelectricity, magnetism, electrical conductivity, molecular adsorption, heterogeneous catalysis and fluorescence (Zhang et al., 2009; Zhang & Xiong, 2012; Nouar et al., 2008; Li et al., 2009; Yao et al., 2008). In particular, noncentrosymmetric bulk materials or polar crystals possess technologically useful properties, such as ferroelectricity, pyroelectricity, piezoelectricity, triboluminescence and nonlinear optical (NLO) functionality (especially second harmonic generation, SHG) (Marks & Ratner, 1995; Desiraju, 1989; Xiong & You, 2002), and SHG, piezoelectric and ferroelectric properties are of particular importance in areas such as telecommunications, optical storage, information processing and mechanical energy transfer. However, the rational design and synthesis of metal complexes crystallizing as acentric and homochiral solids still remain long-term challenges. Recent progress in the synthesis (or self-assembly) and design of novel materials based on metal-organic coordination compounds has shown that acentric and homochiral metal-organic solids can be achieved through the following protocols: (i) using asymmetric, flexible or racemic organic ligands to self-assemble with metal ions; (ii) using racemic organic ligands coordinating to metal ions to result in self-resolution and so obtain chiral compounds; (iii) obtaining asymmetric coordination compounds through self-assembly of metal ions and homochiral organic ligands (Ye & Wu et al., 2005; Xiong et al., 2000; Ye et al., 2008; Zhao et al., 2004). We are interested in hydrothermal or solvothermal reactions because they allow in situ reactions to occur without exerting synthetic control, and we have obtained a fascinating variety of novel acentric metal-organic coordination compounds through these processes (Ye, Wu et al., 2005; Ye et al., 2008; Ye & Wang et al., 2005).

A series of cadmium coordination polymers containing carboxylate ligands has been reported (Yang et al., 2009; Zhang et al., 2012; Zhou et al., 2011; Nie & Wang, 2011; Han et al., 2012; Wang et al., 2012) and, compared with the above-mentioned cases, seven-coordinated complexes crystallizing in noncentrosymmetric space groups are rare. We report here the synthesis and crystal structure of an acentric seven-coordinated tetranuclear cadmium coordination compound tetrakis[-2-(quinolin-8-yloxy)acetato]tetrakis[-2-(quinolin-8-yloxy)acetonitrile]tetracadmium tetrakis(perchlorate) dihydrate, (I).

Compound (I) has an asymmetric unit containing two crystallographically unique Cd^II cations, two 2-(quinolin-8-yloxy)acetate anions, two 2-(quinolin-8-yloxy)acetonitrile molecules, two perchlorate anions and a water molecule (Fig. 1). X-ray analysis of (I) reveals that half of the cyano groups of 2-(quinolin-8-yloxy)acetonitrile were hydrolyzed in situ, leading to the formation of carboxylate groups. Three O atoms (O14, O15 and O16) in one of the perchlorate anions are disordered. The local coordination environments around the two Cd^II cations are similar and are best described as capped octahedra, with coordination from two N atoms and five O atoms from one 2-(quinolin-8-yloxy)acetonitrile and two 2-(quinolin-8-yloxy)acetate ligands. The 2-(quinolin-8-yloxy)acetonitrile ligand coordinates to one Cd^II atom through the O atom and ring N atom to form a stable five-membered chelate ring (Cd1/O4/C17/C16/N2 and Cd2/O8/C28/C27/N4), while the cyano group of the 2-(quinolin-8-yloxy)acetonitrile ligand is not coordinated.

The coordination mode of the 2-(quinolin-8-yloxy)acetate ligand is more complex than that of the 2-(quinolin-8-yloxy)acetonitrile ligand. In addition to the O and N atoms chelating the Cd^II centre, two carboxylate O atoms are also coordinated to the Cd^II centre to form two five-membered chelate rings (Cd1/O1/C10/C11/O2 and Cd1/N1/C5/C6/O1, and Cd2/O5/C43/C44/O6 and Cd2/N5/C38/C39/O5) and one four-membered chelate ring (Cd1/O2^ix/C11^ix/O3^ix and Cd2/O6^xi/C44^xi/O7^xi; symmetry codes as in Fig. 1). Thus, the seven-coordinated Cd^II coordination environment is fulfilled by one O atom and one N atom from one 2-(quinolin-8-yloxy)acetonitrile ligand, two O atoms of the carboxylate group of one 2-(quinolin-8-yloxy)acetate ligand, and two O atoms and one N atom from another 2-(quinolin-8-yloxy)acetate ligand, as shown in Fig. 1.

Although the two unique Cd^II centres show similar coordination environments, the coordination bond lengths and angles are significantly different. The Cd1-O and Cd1-N bond lengths are in the ranges 2.190 (4)-2.698 (5) and 2.254 (5)-2.270 (4) Å, respectively, while the Cd2-O and Cd2-N bond lengths are in the ranges 2.219 (4)-2.687 (4) and 2.269 (4)-2.283 (4) Å, respectively. The Cd-O bond lengths are comparable with those found in [KCd₂(edta)I]_n (H₄edta = ethylenediaminetetraacetic acid; Wang et al., 2012), which range from 2.208 (3) to 2.518 (3) Å, but the Cd-N distances are shorter than those found in [KCd₂(edta)I]_n [2.382 (3)-2.416 (3) Å].

The 2-(quinolin-8-yloxy)acetate anion acts as a tetradentate linker, connecting two Cd^II centres through the ring N atom and three O atoms of the oxyacetate group. Thus, the Cd^II cations involving Cd1 at (x, y, z), (y, -x + 1, -z + 2), (-x + 1, -y + 1, z) and (-y + 1, x, -z + 2) are linked by the 2-(quinolin-8-yloxy)acetate ligand to form a tetranuclear unit (Fig. 2), which is like a distorted `#' symbol, the fourfold rotoinversion center sitting at the centre of the Cd₄ core. A similar Cd₄ core is constructed from the Cd2 atoms. The oxyacetate group bridges the two nearest Cd^II cations: one of the O atoms from the oxyacetate group links the Cd^II cations via an O-Cd bond, and two linked Cd^II cations are strengthened by the oxyacetate group with a Cd-O-C-C-O-Cd bridge.

In the structure of (I), numerous intermolecular hydrogen bonds (O1W-H1WO9, O1W-H2WO15 and O1W-H2WO15B; Table 1) play an important role in stabilizing the structure, linking the ions of the asymmetric unit, the water molecules and the perchlorate anions. In addition, there are numerous nonclassical C-HO hydrogen bonds (Table 1) in the structure, with C-HO angles in the range 106-170°, and CO and HO distances in the ranges 2.982 (8)-3.495 (10) and 2.29-2.58 Å, respectively. The weak C23-H23O12 and C43-H43BO11 intramolecular hygrogen bonds link one of the perchlorate anions and part of the tetranuclear unit in the asymmetric unit. All the tetranuclear units are linked by the perchlorate anions and water molecules via weak C2-H2O11ⁱ, C8-H8O13ⁱⁱ, C10-H10AO3ⁱⁱⁱ, C12-H12O16B^iv, C13-H13O1W^iv, C21-H21BO3^v, C32-H32AO7^vi, C41-H41O9^vii and C43-H43AO7^viii hydrogen bonds, forming the three-dimensional structure (Fig. 3; see Table 1 for symmetry codes).

In summary, (I) crystallizes fortuitously in an acentric space group stabilized by hydrogen-bonding interactions. The hydrothermal reaction leads to in situ ligand synthesis and the formation of this novel tetranuclear Cd^II coordination compound. In addition, the Cd^II cation has a capped octahedral geometry, and coordination compounds with such a geometry for a Cd^II centre are extremely rare in the Cambridge Structural Database (Version 5.33; Allen 2002).

Figure 1 The molecular structure of (I), showing the atom-numbering scheme. Displacement ellipsoids are drawn at the 30% probability level. All H atoms have been omitted for clarity. [Symmetry codes: (ix) -y + 1, x, -z + 2; (x) y, -x + 1, -z + 2; (xi) -y + , x + , -z + ; (xii) y - , -x + , -z + .]

Figure 2 A view of the tetranuclear unit of (I).

Figure 3 The crystal structure of (I), viewed along the c axis. All H atoms have been omitted, except for those on atom O1W. Dashed lines indicate intermolecular O1W-HO hydrogen bonds.

Experimental

The hydrothermal reaction of 2-(quinolin-8-yloxy)acetonitrile (1 mmol) and Cd(ClO₄) (1 mmol) in the presence of water (1.5 ml) in a sealed Pyrex tube at 383 K for several days gave colourless block-shaped crystals of (I). Crystals were isolated from the solution and then washed and dried.

Crystal data

[Cd4(C11H8NO3)4(C11H8N2O)4](ClO4)4·2H2O Mr = 2428.94 Tetragonal, a = 27.838 (4) Å c = 11.779 (2) Å V = 9128 (2) Å3 Z = 4 Mo K radiation = 1.13 mm-1 T = 293 K 0.2 × 0.2 × 0.2 mm

Data collection

Rigaku Mercury2 diffractometer Absorption correction: multi-scan (CrystalClear; Rigaku, 2005) Tmin = 0.795, Tmax = 0.810 47881 measured reflections 10477 independent reflections 8258 reflections with I > 2(I) Rint = 0.075 180 standard reflections every 3 reflections intensity decay: none

Refinement

R[F2 > 2(F2)] = 0.048 wR(F2) = 0.103 S = 1.00 10477 reflections 668 parameters 111 restraints H-atom parameters constrained max = 0.56 e Å-3 min = -0.34 e Å-3 Absolute structure: Flack (1983), with 4980 Friedel pairs Flack parameter: 0.00 (2)

Table 1 Hydrogen-bond geometry (Å, °) D-HA D-H HA DA D-HA O1W-H2WO15B 0.82 2.38 3.078 (15) 143 O1W-H2WO15 0.82 2.67 3.06 (2) 111 O1W-H1WO9 0.82 2.16 2.948 (10) 162 C2-H2O11i 0.93 2.49 3.212 (8) 135 C8-H8O13ii 0.93 2.38 3.285 (9) 165 C10-H10AO3iii 0.97 2.45 3.308 (7) 147 C12-H12O16Biv 0.93 2.29 3.155 (17) 154 C13-H13O1Wiv 0.93 2.47 3.258 (10) 142 C21-H21BO3v 0.97 2.57 2.982 (8) 106 C23-H23O12 0.93 2.47 3.233 (9) 139 C32-H32AO7vi 0.97 2.47 3.131 (8) 126 C41-H41O9vii 0.93 2.58 3.495 (10) 170 C43-H43AO7viii 0.97 2.34 3.236 (7) 153 C43-H43BO11 0.97 2.57 3.344 (9) 137 Symmetry codes: (i) ; (ii) ; (iii) -x+1, -y+1, z; (iv) ; (v) -y+1, x, -z+2; (vi) ; (vii) ; (viii) -x+1, -y+2, z.

Three O atoms of the perchlorate anion were modelled as disordered over two sites. The site-occupancy factor of the major component refined to 0.55 (4). All atom lengths in the anions were restrained to reasonable distances, and restraints were also applied to the three disordered O atoms to ensure that reasonable displacement parameters were obtained. Constraints were added to give equivalent displacement parameters for the corresponding atoms of the minor and major disorder components.

H atoms on C atoms were included in calculated positions and were refined using a riding model, with C-H = 0.93 or 0.97 Å for aromatic or methylene C atoms, respectively, and with U_iso(H) = 1.2U_eq(C). The H atoms bonded to atom O1W were discernible in difference electron-density maps. These atoms were placed as found and initially allowed to refine, with the O-H distances restrained to 0.85 (1) Å and HH distances restrained to 1.39 (1) Å, but for the final refinement cycles, these H atoms were constrained to ride on the water O atom. For water H atoms, U_iso(H) = 1.5U_eq(O).

Data collection: CrystalClear (Rigaku, 2005); cell refinement: CrystalClear; data reduction: CrystalClear; program(s) used to solve structure: SHELXS97 (Sheldrick, 2008); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008); molecular graphics: PLATON (Spek, 2009) and DIAMOND (Brandenburg & Putz, 2005); software used to prepare material for publication: SHELXTL (Sheldrick, 2008).

Supplementary data for this paper are available from the IUCr electronic archives (Reference: WQ3022 ). Services for accessing these data are described at the back of the journal.