PAPER www.rsc.org/dalton | Dalton Transactions

Combination of lacunary polyoxometalates and high-nuclear transition-metalclusters under hydrothermal conditions: first 65·8 CdSO4-type 3-D frameworkbuilt by hexa-CuII sandwiched polyoxotungstates†

Jun-Wei Zhao,a Shou-Tian Zheng,a Zhao-Hui Lib and Guo-Yu Yang*a

Received 3rd June 2008, Accepted 7th November 2008First published as an Advance Article on the web 16th January 2009DOI: 10.1039/b809338e

A 65·8 CdSO4-type 3-D framework built by hexa-CuII sandwiched polyoxotungstates [Cu(en)2]2-[Cu(deta)(H2O)]2[Cu6(en)2(H2O)2(B-a-GeW9O34)2]·6H2O (1) (en = ethylenediamine and deta =diethylenetriamine) has been hydrothermally synthesized and structurally characterized by elementalanalysis, IR spectrometry, thermogravimetric analysis and single-crystal X-ray diffraction. To the bestof our knowledge, 1 represents the first 3-D TMSP with 65·8 CdSO4 topology in polyoxometalatechemistry. Magnetic measurements performed on 1 indicate the occurrence of ferromagneticinteractions within the copper ions and the coupling constants can be determined.

Introduction

Designed synthesis of high-nuclear transition-metal substitutedpolyoxometalates (TMSPs) is currently an intensive interestin polyoxometalate (POM) chemistry, because of not onlytheir potential applications in catalysis, biology, medicine andmagnetochemistry,1 but also some original topologies based onTMSPs and exotic bridging groups.2 Since the first sandwich-typeTMSP and its ferromagnetic functionality were discovered,3 in-vestigations on the functionalization of TMSPs constructed fromlacunary POMs and different TM cations have been extensivelyperformed. So far, trivacant Keggin or Dawson POMs have provento be the most versatile synthons in making novel TMSP magneticfunctional clusters that are generally synthesized by conventionalaqueous solution methods.4 However, functional TMSPs made byhydrothermal reactions are very rare,2,5 which not only provides uswith great challenges and opportunities, but also gives us a greatimpetus to explore this domain.

Since 2005, we have concentrated on exploring the reactionsystem containing lacunary POM precursors and TM cationsin the presence of N-ligands under hydothermal conditions.Under the guidance of our synthetic concept that lacunarysites of XW9O34 (X = PV/SiIV/GeIV) fragments act as structure-directing agents to induce the formation of larger oligomers ofTM clusters, and multidentate amines act as structure-stabilizingagents to capture and stabilize TM oligomers formed in-situ,to make novel TMSPs,2a,5a a family of novel organic-inorganichybrid TMSPs containing tera-,2c,5b,c hexa-,2a,5a,d–f hepta-,2a

octa-,2a,b and higher-nuclear TM clusters2a,5g have been successfully

aState Key Laboratory of Structural Chemistry, Fujian Institute of Researchon the Structure of Matter and Graduate School of the Chinese Academy ofSciences, Fuzhou, Fujian 350002, P. R. China. E-mail: ygy@fjirsm.ac.cn;Fax: (+86) 591-83710051bCollege of Chemistry and Chemical Engineering, Fuzhou University,Fuzhou, Fujian 350002, P. R. China† Electronic supplementary information (ESI) available: Some related 4-connected topologies, IR spectrum, TG analysis and related magneticcurves. CCDC reference numbers 682314. For ESI and crystallographicdata in CIF or other electronic format see DOI: 10.1039/b809338e

isolated by us. Notably, we find that the same TMSP unitscan be self-polymerized by terminal oxygen atoms of POMs orexogenous bridging groups to further generate larger aggregatesor extended frameworks.2,5a,b,e These appealing findings encourageus to further exploit this subject. To construct novel extendedframeworks with a new topological net, the CuII ions are employedchiefly due to their flexible coordination modes and obviousJahn–Teller distortion. Fortunately, we have made two inorganic-organic hybrid 3-D TMSP-based frameworks, [{Cu6(m3-OH)3

(en)3(H2O)3}(B-a-PW9O34)]·7H2O2a and [Cu(H2O)2]H2[Cu8(dap)4

(H2O)2(B-a-GeW9O34)2]2b with 6-connected 4966 and (3,6)-connected (4·62)(42·64·87·102) topology, respectively. Continuingour work, we again obtained a novel 4-connected 3-D frame-work with 65·8 CdSO4 topology built by hexa-CuII sandwichedpolyoxotungstates [Cu(en)2]2[Cu(deta)(H2O)]2[Cu6 (en)2(H2O)2(B-a-GeW9O34)2]·6H2O (1), which represents the first 3-D TMSP with65·8 CdSO4 topology in POM chemistry. In the present paper, wereport the synthesis, crystal structure and related characterizationsof 1.

Experimental

Materials and methods

All chemicals were commercially purchased and used withoutfurther purification. The precursor K8Na2[A-a-GeW9O34]·25H2O6

was synthesized according to the literature and characterized byIR spectroscopy. Elemental analyses (C, H and N) were performedusing a PE 2400 II elemental analyzer. IR spectra were obtainedon an ABB Bomen MB 102 spectrometer with pressed KBrpellets in the range of 4000–400 cm-1. Thermo-gravimetric (TG)analysis was performed on a Mettler TGA/SDTA851 thermalanalyzer in a flowing air atmosphere in the temperature regionof 30–800 ◦C with a heating rate of 10 ◦C min-1. Magneticsusceptibility measurements were carried out with a QuantumDesign MPMS-5 magnetometer in the temperature range of2–300 K. The susceptibility data were corrected from the diamag-netic contributions as deduced by using Pascal’s constant tables.

1300 | Dalton Trans., 2009, 1300–1306 This journal is © The Royal Society of Chemistry 2009

Publ

ishe

d on

16

Janu

ary

2009

. Dow

nloa

ded

by U

nive

rsity

of

New

cast

le o

n 10

/08/

2013

11:

48:2

0.

View Article Online / Journal Homepage / Table of Contents for this issue

The electron spin resonance (EPR) spectrum was recorded at roomtemperature on a Bruker ER200-D-SRC spectrometer operatingat X-band.

Synthesis

[Cu(en)2 ]2 [Cu(deta)(H2O)]2 [Cu6 (en)2 (H2O)2 (B-a-GeW9O34 )2 ]·6H2O (1). A mixture of K8Na2[A-a-GeW9O34]·25H2O (0.246 g,0.08 mmol), CuCl2·2H2O (0.085 g, 0.50 mmol), deta (0.05 mL,0.047 mmol), KOH (1 M, 0.10 mL) and H2O (8 mL, 444 mmol)was stirred for 4 hours, sealed in a Teflon-lined bomb (20 mL),kept at 130 ◦C for 5 days and then cooled to room temperature.Dark green prismatic crystals were obtained by filtration, washedwith distilled water and dried in air. Yield: ca. 42% (based on K8

Na2[A-a-GeW9O34]·25H2O). Anal. Calcd (%) for C20H92Cu10 Ge2

N18O78W18 (1): C, 4.06; H, 1.57; N, 4.26. Found: C, 3.98; H, 1.75;N, 4.35. IR (KBr pellets, cm-1): 3436, 3284, 3255, 1592, 1459, 1392,1272, 1145, 1099, 1044, 940, 877, 771, 740, 695, 506, 466.

X-Ray crystallography

A prismatic single-crystal of 1 was mounted on a glass fiberfor indexing, and intensity data were collected at 293 K on aRigaku Mercury 70 CCD/AFC diffractometer with graphite-monochromated Mo Ka radiation (l = 0.71073 A). Directmethods were used to solve the structure and to locate theheavy atoms using the SHELXTL-97 program package.7 Theremaining atoms were found from successive full-matrix least-squares refinements on F 2 and Fourier syntheses. Routine Lorentzpolarization corrections and empirical absorption correction wereapplied to intensity data. Anisotropic thermal parameters wereused to refine all non-H atoms except for C1, C2, C5, N1, N2and O5W. No H atoms associated with water molecules werelocated from the difference Fourier map. H atoms attached toC and N atoms were geometrically placed. All H atoms wererefined isotropically as a riding mode using the default SHELXTLparameters. The crystal data and structural refinements of 1 aresummarized in Table 1. CCDC reference number 682314.†

Table 1 Crystal data and structural refinements for 1

1

Empirical formula C20N18H92O78Cu10Ge2W18

Formula weight 5923.00Crystal system MonoclinicSpace group C2/ca/A 23.606(2)b/A 12.8670(9)c/A 32.583(3)b/A 107.593(5)V/A3 9434.1(14)Z 4Dc/g cm-3 4.170m/mm-1 24.784F(000) 10592Measured reflections 28237Independent reflections 8224Rint 0.0359Refinement method Full-matrix least-squares on F 2

Data/restraints/parameters 8224/8/620Goodness-of-fit on F 2 1.055Final R indices [I > 2s(I)] R1 = 0.0311, wR2 = 0.0677R indices (all data) R1 = 0.0364, wR2 = 0.0703

Results and discussion

Synthesis

Compound 1 was successfully made by the hydrothermal re-action of K8Na2 [A-a-GeW9O34]·25H2O, CuCl2·2H2O and detain water at 130 ◦C for 5 days under the guidance of ourrecent findings [{Cu6(m3-OH)3(en)3(H2O)3}(B-a-PW9O34)]·7H2O2a

and [Cu(H2O)2]H2[Cu8(dap)4(H2O)2(B-a-GeW9O34)2].2b This suc-cess reveals that the rational design and synthesis are possible to alimited extent within the selected family of compounds, althoughwe are far from the ultimate dream of “tailormaking” desiredproducts with specified structures and properties. In our case, itmay be possible to predict the existence of new phases withinthe similar structural types by analogy to the already knownphases. Additionally, in the formation of 1, two features should bepointed out: one is that the [A-a-GeW9O34]10- (A-a-GeW9) unitis converted to the [B-a-GeW9O34]10- (B-a-GeW9) unit, which hasbeen observed in previous studies,2b,5f,8 and is intimately related tothe case that the B-a-GeW9 unit can work as a heptadentate ligandto bond to the hexa-copper cluster formed in-situ; the other is thatthe en ligand in the product is from the decomposition of deta. Asimilar deaminization phenomenon has been observed previously.9

Crystal structure

Single-crystal X-ray structural analysis reveals that 1 crys-tallizes in the monoclinic space group C2/c and its struc-tural unit contains a hexa-supporting subunit [Cu(en)2]2[Cu(deta)-(H2O)]2[Cu6(en)2(H2O)2(B-a-GeW9O34)2] (1a) and six lattice watermolecules. As shown in Fig. 1a, 1a consists of a hexa-CuII

sandwiched [Cu6(en)2(H2O)2(B-a-GeW9O34)2]8- core, two pendant[Cu3(deta)(H2O)]2+ cations and two unique [Cu(en)2]2+ bridging

Fig. 1 (a) Polyhedral and ball-and-stick representation of the hexa-sup-porting subunit 1a. (b) The connection motif of the belt-like {Cu6} cluster.Atoms with “A or B” in their labels are symmetrically generated (A: -x,1 - y, 1 - z; B: 0.5 + x, -0.5 + y, z). (c) The 3-D framework viewed downthe b-axis. (d) The 65·8 CdSO4-type topology. The red balls represent the4-connected la subunits.

This journal is © The Royal Society of Chemistry 2009 Dalton Trans., 2009, 1300–1306 | 1301

Publ

ishe

d on

16

Janu

ary

2009

. Dow

nloa

ded

by U

nive

rsity

of

New

cast

le o

n 10

/08/

2013

11:

48:2

0.

View Article Online

cations of [Cu1(en)2]2+ and [Cu2(en)2]2+. Both Cu1 and Cu2 ionsare located on the special sites (0.00, 0.58, 0.25) and (-0.25,0.75, 0.50) and exhibit the elongated octahedral geometries thatconsist of four N atoms from two en ligands for the basalplane [Cu–Nen: 1.961(9)–2.022(9) A] and two terminal O atomsfrom adjacent two B-a-GeW9 units for the two axial positions[Cu–OPOM: 2.734(7)–2.767(9) A]. The pendant [Cu3(deta)(H2O)]2+

cation resides in a distorted square pyramidal geometry, wherethree N atoms from a tridentate deta ligand and one water ligandsit on the basal plane (Cu–Ndeta: 1.917(15)–1.975(11) A, and Cu–OW: 1.979(10) A), and one terminal O atom from B-a-GeW9 unitis situated on the axial site (Cu–OPOM: 2.328(7) A). The [Cu6(en)2

(H2O)2(B-a-GeW9O34)2]8- core is built by two trivacant KegginB-a-GeW9 units in a staggered fashion linked via a belt-like hexa-CuII{[Cu(en)]2Cu4O14(H2O)2}({Cu6}) cluster (Fig. 1b), resulting ina sandwich-type assembly. The {Cu6} cluster links two B-a-GeW9

moieties via fourteen bridging O atoms from lacunae of two B-a-GeW9 units (two m4-O from one GeO4 unit and twelve m2-Ofrom twelve WO6 groups). The Cu(4,4A), Cu(5,5A) and Cu(6,6A) cations form a {Cu6} cluster by the structure-directing role oflacunary sites of two B-a-GeW9 units. The Cu4 cation adopts adistorted square pyramidal geometry defined by two N atoms froman en ligand (Cu–N: 1.954(8)–2.009(8) A) and three O atoms (Cu–OPOM: 1.952(6)–2.406(6) A) from two B-a-GeW9 units. Both Cu5and Cu6 atoms form two distorted octahedral coordination con-figurations (Cu–OPOM: 1.936(7)–2.364(6) A, Cu–OW: 2.383(8) A).

The belt-like {Cu6} cluster in 1 is of particular interest.In comparison with the tetra-Cu sandwiched [Cu4(H2O)2(B-a-GeW9O34)2]12- unit,8 the {Cu6} cluster in 1 can be viewedas a derivative of {Cu4O14(H2O)2} in the [Cu4(H2O)2(B-a-GeW9O34)2]12- unit: two [Cu(en)]2+ groups are grafted onto twodiagonal corners of the rhombic {Cu4O14(H2O)2} unit (Fig. 2a,b)via six O atoms from lacunae of two B-a-GeW9 units. Incomparison with the octa-Cu sandwiched [Cu8(dap)4(H2O)2(B-a-GeW9O34)2]4- unit,2b the {Cu6} core in 1 can also be viewedas a derivative of the {[Cu(dap)]4Cu4O14(H2O)2} core in the[Cu8(dap)4(H2O)2(B-a-GeW9O34)2]4- unit: under the conditionof en replacing dap, two diagonal [Cu(dap)]2+ cations are

Fig. 2 Relationship of polyhedral representations of (a) hexa-Cu clusterin 1a, (b) tetra-Cu cluster in [Cu4(H2O)2(B-a-GeW9O34)2]12-, (c) oc-ta-Cu cluster in [Cu8(dap)4(H2O)2(B-a-GeW9O34)2]4-, (d) hexa-Cu clusterin [Cu(enMe)2]2{[Cu(enMe)2(H2O)]2[Cu6(enMe)2(B-a-SiW9O34)2]}·4H2O,(e) hexa-Cu cluster in [{Cu6(m3-OH)3(en)3(H2O)3}(B-a-PW9O34)]·7H2O,and (f) hexa-Cu cluster in [(CuCl)6(AsW9O33)2]12-. The carbon atoms oforganic ligands in a and c–e are omitted for clarity.

removed from the {[Cu(dap)]4Cu4O14(H2O)2} core (Fig. 2a,c).Notice that the {Cu6} cluster in 1 is somewhat differentfrom the hexa-Cu {[Cu(enMe)]2Cu4O14} cluster (Fig. 2d) indiscrete [Cu(enMe)2]2{[Cu(enMe)2(H2O)]2[Cu6(enMe)2(B-a-SiW9O34)2]}·4H2O reported by our lab.5d The {Cu6} cluster in1 is made up of two CuN2O3 square pyramids and four CuO6

octahedra through edge-sharing modes (Fig. 2a) whereas the hexa-Cu cluster in [Cu(enMe)2]2{[Cu(enMe)2(H2O)]2[Cu6(enMe)2(B-a-SiW9O34)2]}·4H2O is constructed by two CuN2O3 squarepyramids, two CuO5 square pyramids and two CuO6 octahedravia edge-sharing modes (Fig. 2d). In addition, the {Cu6}core in 1 is completely distinct from the two coplanar hexa-Cu units {Cu6(m3-OH)3(en)3(H2O)3} (Fig. 2e) in [{Cu6(m3-OH)3(en)3(H2O)3}(B-a-PW9O34)]·7H2O2a and {Cu6O12Cl6}(Fig. 2f) in (n-BuNH3)12[(CuCl)6(AsW9O33)2]·6H2O.10 In [{Cu6(m3-OH)3(en)3(H2O)3}(B-a-PW9O34)]·7H2O, the hexa-Cu clusterwith C3 symmetry exhibits a triangular alignment built bysix edge-sharing CuN2O4 and CuO6 octahedra,2a whereas in(n-BuNH3)12[Cu6Cl6(B-a-AsW9O33)2]·4H2O, the hexa-Cu clusterwith D3d symmetry shows a hexagonal alignment formed bysix edge-sharing CuO4Cl square pyramids.10 Furthermore, thegeometry of the hexa-Cu cluster in 1 is distinct from that of thehexa-Fe cluster in [Fe6(OH)3(A-a-GeW9O34(OH)3)2]11-,11 and inthe latter, six Fe3+ ions are aligned in a trigonal prismatic fashion.Notably, another discrete hexa-Yb sandwiched Dawson-basedphosphotungstate [{Yb6(m6-O)(m3-OH)6(H2O)6}(a-P2W12O56)2]14-

has also reported by Hill et al., in which six Yb3+ ions arearranged in a trigonal antiprismatic motif.12 In a word, althoughthe above-mentioned sandwich-type TMSPs consist of six metalions in the sandwich belts, the geometries of the hexa-metalclusters are somewhat different. It should be noted that 1 is anovel 3-D extended structure with 65·8 CdSO4 topology built byhexa-CuII sandwiched polyoxotungstates, which is very rare in thesandwich-type subfamily.2b–d,5b

The most intriguing feature of 1 is that each 1a as a completecluster connects four others through four [Cu(en)2]2+ bridges toform a novel 3-D framework (Fig. 1c). From the topological pointof view, the simplification of 3-D networks to node-and-spacerrepresentations that illustrate their connectivity has facilitated theanalysis and understanding of complicated topologies by crystalengineering.13 The circuit symbols and Schlafli (vertex) notationscan be used to describe topologies and facilitate comparisonof networks of different composition and metrics. The 3-Dframework of 1 is a 4-connected 3-D network, where each 1asubunit acts as a 4-connected node. A topological analysis ofthis net was performed with OLEX.14 The long topological(O’Keeffe) vertex symbol is (6·6·62·6·6·•) for the 1a node, whichgives the short vertex (Schafli) symbol of (65·8). Comparing thetopology of 1 with those of known minerals, 1 possesses the65·8 CdSO4 topology (Fig. 1d),13,15 being obviously different fromother 4-connected topologies of minerals, for example, diamond(62·62·62·62·62·62), NbO (62·62·62·62·82·82), PtS (4·4·82·82·82·82) andCrB4 (4·62·6·6·6·6) (Fig. S1).15a

IR spectrum

In the IR spectrum of 1 (Fig. S2†), four characteristic vibrationbands resulting from the Keggin polyoxoanion framework areobserved at 940 cm-1, 877 cm-1, 771 cm-1 and 695 cm-1, which

1302 | Dalton Trans., 2009, 1300–1306 This journal is © The Royal Society of Chemistry 2009

Publ

ishe

d on

16

Janu

ary

2009

. Dow

nloa

ded

by U

nive

rsity

of

New

cast

le o

n 10

/08/

2013

11:

48:2

0.

View Article Online

are ascribed to n(W–Ot), n(Ge–Oa), n(W–Ob) and n(W–Oc),respectively. The stretching bands of–OH and–NH2 groups areobserved at 3436 cm-1 and 3284–3255 cm-1 and two bands centeredat 1592 cm-1 and 1459 cm-1 are assigned to the bending vibrationpatterns of–NH2 and–CH2 groups, respectively. The appearance ofthese characteristic signals confirms the presence of organic aminegroups in 1, being consistent with the single-crystal structuralanalysis.

Magnetic properties

Since 1 contains a well-isolated {Cu6} cluster between twodiamagnetic trivacant Keggin polyoxoanion fragments with acentrosymmetrical topology, it is of interest to probe its magneticproperties. The magnetic susceptibility of 1 measured at a field of5 kOe in the temperature range 2–300 K is shown on Fig. 3. Thecm slightly rises from 0.012 emu mol-1 at 300 K to 0.159 emu mol-1

at 24 K, then exponentially to the maximum of 1.313 emu mol-1

at 2 K. At 300 K, the cmT is equal to 3.621 emu mol-1 K, which issomewhat lower than that expected for ten magnetically uncoupledCuII ions assuming g = 2 (3.750 emu mol-1 K). When the sampleis cooled from 300 K to 57 K, the cmT shows a slight decrease.Further cooling, the cmT gradually increases to reach a maximumof 3.924 emu mol-1 K at 13 K, before dropping to 2.622 emumol-1 K at 2 K. This behavior indicates the presence of domi-nant ferromagnetic exchange interactions (J1, J3–J5 > 0) withinthe {Cu6} cluster accompanying weak intracluster interactions(zJ¢ < 0) and antiferromagnetic interactions (J2 < 0), accountingfor the decrease in cmT at low temperatures. The inverse magneticsusceptibility data in the temperature range of 2–300 K are fittedto the Curie–Weiss equation with C = 3.58 emu mol-1 K and q =0.11 K (Fig. 4). The small positive Weiss constant suggests weakdominant ferromagnetic interactions.

Fig. 3 Temperature dependence of cm and cmT for 1. The solid linesrepresent the best fit to the model. The inset shows the exchange modelwithin the {Cu6} cluster.

In order to analyze magnetic coupling interactions withinthe {Cu6} cluster, we examined the structural parameters of 1.Because the exchange interactions within the {Cu6} cluster aremediated through the oxo-bridges, it is important to examinethe bond lengths and angles. Since the well-isolated {Cu6}cluster is centrosymmetrical (Fig. 1b), five types of Cu ◊ ◊ ◊ Cu

Fig. 4 The temperature dependence of the inverse magnetic susceptibilitycm

-1 for 1 between 2 and 300 K, and the red solid line is generated fromthe best fit by the Curie–Weiss expression in the range 2–300 K.

distances exist in the {Cu6} cluster: 3.17 A (Cu4 ◊ ◊ ◊ Cu5), 3.13 A(Cu4 ◊ ◊ ◊ Cu6A), 3.18 A (Cu5 ◊ ◊ ◊ Cu6), 3.20 A (Cu5 ◊ ◊ ◊ Cu6A) and3.08 A (Cu5 ◊ ◊ ◊ Cu5A); and eight types of Cu–O–Cu angles: 106.1◦

(Cu4–O30A–Cu5), 81.1◦ (Cu4–O18–Cu5), 104.8◦ (Cu4–O27A–Cu6A), 93.6◦ (Cu5–O33A–Cu6), 94.7◦ (Cu5–O34–Cu6), 90.8◦

(Cu5–O18–Cu6A), 94.8◦ (Cu5–O34A–Cu6A) and 101.4◦ (Cu5–O34–Cu5A). A classical correlation between the experimentalexchange constants and the Cu–O–Cu bond angles has beenconcluded from extensive experimental and theoretical studiesperformed on the spin exchange within [Cu2O2] units constitutedby hydroxide16a and alkoxide16b,c bridges. It is clear from this vastamount of research that the nature and strength of the exchangeare chiefly affected by the Cu–O–Cu bond angles (U). The classicalcorrelation between the experimental exchange constants and theCu–O–Cu bond angles indicates that the complexes are generallyantiferromagnetic for U > 98◦, while ferromagnetic for U <

98◦.16d–f furthermore, dominant ferromagnetic interactions areexpected because more Cu–O–Cu bond angles are less than 98◦ inthe {Cu6} cluster.

Accordingly, we have fitted the magnetic data to a simplemodel with the intercluster coupling (zJ¢) using the molecularfield approximation. The magnetic exchange model of the {Cu6}cluster in 1 is shown in Fig. 3, the vertices with numbers 1, 2, 3,4, 5 and 6 symbolize Cu4, Cu5, Cu6, Cu6A, Cu5A and Cu4A,respectively (Fig. 1b). The interactions between Cu4 ◊ ◊ ◊ Cu5 andCu4A ◊ ◊ ◊ Cu5A are defined as the coupling constant J1; theinteractions between Cu4 ◊ ◊ ◊ Cu6A and Cu4A ◊ ◊ ◊ Cu6 are definedas the coupling constant J2; the interactions between Cu5 ◊ ◊ ◊ Cu6and Cu5A ◊ ◊ ◊ Cu6A are defined as the coupling constant J3;the interactions between Cu5 ◊ ◊ ◊ Cu6A and Cu5A ◊ ◊ ◊ Cu6 aredefined as the coupling constant J4 and the interaction betweenCu5 ◊ ◊ ◊ Cu5A is defined as the coupling constant J5. Thus, theisostropic spin Hamiltonian for the {Cu6} cluster is described as:

H = -2J1(S1S2 + S5S6) - 2J2(S1S4 + S3S6) - 2J3(S2S3 + S4S5)- 2J4(S2S4 + S3S5) - 2J5S2S5

The [Cu1(en)2]2+, [Cu2(en)2]2+ bridges and [Cu3(deta)(H2O)]2+

pendant cations are included as paramagnetic ions in the fittingprocedure. Due to the size and lower symmetry of the molecules,a matrix diagonalization method to evaluate the various CuII

This journal is © The Royal Society of Chemistry 2009 Dalton Trans., 2009, 1300–1306 | 1303

Publ

ishe

d on

16

Janu

ary

2009

. Dow

nloa

ded

by U

nive

rsity

of

New

cast

le o

n 10

/08/

2013

11:

48:2

0.

View Article Online

pairwise exchange parameters within such a hexa-CuII clusteris not easy. Similarly, it is difficult to deduce the theoreticalexpression by applying the equivalent operator approach on thebasis of the Kambe vector coupling method.17 Therefore, we usedthe MAGPACK program package18 to fit the magnetic data of1. The best-fitting set of parameters are J1 = 0.40 cm-1, J2 =-4.42 cm-1, J3 = 6.24 cm-1, J4 = 6.23 cm-1, J5 = 7.64 cm-1,zJ¢ = -0.049 cm-1 and g = 2.1. The agreement factor R, definedas

∑[(cM)obs - (cM)cal]2/

∑(cM)obs

2, is equal to 1.99 ¥ 10-4. Thenegative zJ¢ and J2 values account for the very small intraclusterand antiferromagnetic interactions observed below 13 K. Notethat the observation that the cm is less field-dependent for appliedfields below 1000 Oe further identifies the presence of smallintermolecular interactions (Fig. S3†).19

As shown in Fig. 5, the field dependence of magnetizationreveals that the magnetization curve at 2 K increases with raising

Fig. 5 Field dependence of the magnetization of 1 at 2 K.

applied field but its value surprisingly stays smaller than thetheoretical value calculated from the Brillouin function for tenuncoupled CuII spins. Such behavior may suggest that ferromag-netic interactions (J1 > 0, J3 >0, J4 >0 and J5 > 0) coexistwith antiferromagnetic interactions (J2 < 0) in the structure.This explanation is ascertained by the fact that the maximumof the cMT value of 3.924 emu mol-1 K at 13 K is far smallerthan the theoretical expected value of 7.50 emu mol-1 K forthe ferromagnetic hexa-CuII cluster through oxygen bridges andfour paramagnetic CuII ions. This phenomenon of the coexistenceof competitive ferromagnetic and antiferromagnetic exchangesbetween CuII magnetic centers in one compound was observedpreviously.2b,5d,20 In addition, no divergence between the ZFCand FC curves (Fig. S4†) and no frequency peaks are foundfor 1. cac(T) gives a behavior analogous to that of cdc(T) andshows no evidence of frequency dependence and peaks (Fig. S5†).Obviously, even down to 2 K, no sharp phase transition indicativeof magnetic order is observed.

Since 1, (enMe)2]2{[Cu(enMe)2(H2O)]2[Cu6(enMe)2(B-a-SiW9-O34)2]}·4H2O,5d [{Cu6(m3-OH)3(en)3(H2O)3}(B-a-PW9O34)]·7H2O2a and (n-BuNH3)12[Cu6Cl6(B-a-AsW9O33)2]·4H2O10,21

all contain coplanar hexa-Cu clusters, it is meaningful tocompare their magnetic properties. Although both hexa-Cuclusters in 1 and (enMe)2]2{[Cu(enMe)2(H2O)]2[Cu6(enMe)2(B-a-SiW9O34)2]}·4H2O,5d adopt the same alignment mode (Fig. 6a,b)and exhibit the ferromagnetic behavior, the magnitudes oftheir magnetic coupling constants are somewhat different(the magnitude of the former is smaller than that of thelatter), which may be related to the influence that trivacantPOM fragments (dGe > dSi) impose their geometries ontothe hexa-Cu clusters. Moreover, according to the reportedrelationship that the complexes are generally anti-ferromagneticfor U > 98◦, while ferromagnetic for U < 98◦,16d–f becausemore Cu–O–Cu bond angles are less than 98◦ in the hexa-Cu

Fig. 6 The distribution of Cu–O–Cu (U) bond angles in hexa-Cu clusters in 1 (a), (enMe)2]2{[Cu(enMe)2(H2O)]2[Cu6(enMe)2(B-a-SiW9O34)2]}·4H2O(b), [{Cu6(m3-OH)3(en)3(H2O)3}(B-a-PW9O34)]·7H2O (c) and (n-BuNH3)12[Cu6Cl6(B-a-AsW9O33)2]·4H2O (d). Atoms with “A or B” in their labels aresymmetrically generated (A:-x, 1 - y, 1 - z; B: 1 - x, 1 - y, 1 - z; C, 3 - y, 2 + x - y, z; D: 1 - x + y, 3 - x, z; E: 2/3 + x - y, 1/3 + x, 4/3 - z; F: 1 - y,1 + x - y, z; G: 2/3 - x, 4/3 - y, 4/3 - z; H:-x + y, 1 - x, z; I:-1/3 + y, 1/3 - x + y, 4/3 - z).

1304 | Dalton Trans., 2009, 1300–1306 This journal is © The Royal Society of Chemistry 2009

Publ

ishe

d on

16

Janu

ary

2009

. Dow

nloa

ded

by U

nive

rsity

of

New

cast

le o

n 10

/08/

2013

11:

48:2

0.

View Article Online

clusters in 1 and (enMe)2]2{[Cu(enMe)2(H2O)]2[Cu6(enMe)2(B-a-SiW9O34)2]}·4H2O reported by our lab,5d their ferromagneticinteractions within hexa-Cu clusters are expected. On the otherhand, the magnitude differences of their magnetic couplingconstants can explain that the structures and symmetries of POMfragments can tune or control the strength of magnetic exchangeinteractions of TM clusters encapsulated in POM matrixes byimposing their geometries and symmetries onto TM clusters.The differences in magnetic behaviors for 1 (ferromagnetic),[{Cu6(m3-OH)3(en)3(H2O)3}(B-a-PW9O34)]·7H2O (anti-ferro-magnetic)2a and (n-BuNH3)12[Cu6Cl6(B-a-AsW9O33)2]·4H2O(ferromagnetic)10,21 are mainly controlled by the distinctalignment modes of the hexa-Cu clusters. According to thestructural parameters of the hexa-Cu cluster in [{Cu6(m3-OH)3(en)3(H2O)3}(B-a-PW9O34)]·7H2O (Fig. 6c), the exchangepathway between each pair of CuII ions are transmitted throughone m-oxo and one m-OH bridge or two m-OH bridges with Cu–O–Cu bond angles of 90.0–104.8◦. In such a competitive environment,since the numbers of bond angles less than 98◦ and bond anglesmore than 98◦ are comparable, the weak antiferromagneticinteractions (J1 = -3.31 cm-1 and J2 = -0.67 cm-1) can beunderstood in [{Cu6(m3-OH)3(en)3(H2O)3}(B-a-PW9O34)]·7H2O.2a

For (n-BuNH3)12[Cu6Cl6(B-a-AsW9O33)2]·4H2O, all Cu–O–Cubond angles are smaller than 98◦ (Fig. 6d) in the hexagonal copperclusters, as a result, the stronger ferromagnetic interactions can bepredicted, which are also confirmed by the results of theoreticalsimulation.10,21

Electron spin resonance (EPR)

The X-band EPR spectrum of the polycrystalline sample 1recorded at room temperature is depicted in Fig. 7. At roomtemperature, the EPR spectrum displays a broad anisotropicsignal at ca. 3400 G with the line width DHPP = 1000 G. Thebroad resonance having no hyperfine structure indicates that itmay originate from the magnetic exchange interactions withinthe hexa-CuII cluster and thermal perturbation,22 which is furtherconfirmed by magnetic susceptibility measurements. The principalvalue of the g tensor (g = 2.103) indicates the presence ofthe expected CuII ions residing in the octahedral or squarepyramidal stereochemistry with the unpaired electron on thedx2-y2 orbital.22,23 This result is in good agreement with X-raysingle-crystal diffraction analysis. Additionally, the principal value

Fig. 7 X-band powder EPR spectrum of 1 recorded at room temperature.

g = 2.103 derived from the EPR spectrum is also consistent withthat (g = 2.1) from the analysis of magnetic properties, whichsuggests the rationality of the isostropic spin Hamiltonian for thehexa-CuII cluster.

Conclusions

In summary, a novel hexa-CuII sandwiched polyoxotungstate hasbeen hydrothermally synthesized and structurally characterized,which exhibits the 65·8 CdSO4-type 3-D framework. Magneticstudy indicates that ferromagnetic interactions exist within thehexa-Cu cluster in 1, and magnetic properties of several dif-ferent hexa-Cu substituted POMs reported by us, Yamase andTsukerblat have been compared. Moreover, the successful prepa-rations of 1, [{Cu6(m3-OH)3(en)3(H2O)3}(B-a-PW9O34)]·7H2O2a

and [Cu(H2O)2]H2[Cu8(dap)4(H2O)2(B-a-GeW9O34)2]2b can pro-vide guidance for discovering novel TMSPs with novel topologyaccompanying unique properties. Further work on exploring newTMSPs with topologies of minerals is in progress.

References

1 (a) L. Cronin, C. Beugholt, E. Krickemeyer, M. Schmidtmann, H.Bogge, P. Kogerler, T. K. K. Luong and A. Muller, Angew. Chem., Int.Ed., 2002, 41, 2805; (b) W. Chen, Y. Li, Y. Wang, E. Wang and Z. Su,Dalton Trans., 2007, 4293.

2 (a) J.-W. Zhao, H.-P. Jia, J. Zhang, S.-T. Zheng and G.-Y. Yang, Chem.–Eur. J., 2007, 13, 10030; (b) J.-W. Zhao, J. Zhang, S.-T. Zheng and G.-Y.Yang, Chem. Commun., 2008, 570; (c) J.-W. Zhao, B. Li, S.-T. Zhengand G.-Y. Yang, Cryst. Growth Des., 2007, 7, 2658; (d) Z. Zhang, J.Liu, E. Wang, C. Qin, Y. Li, Y. Qi and X. Wang, Dalton Trans., 2008,463.

3 (a) T. J. R. Weakley, H. T. jun.Evans, J. S. Showell, G. F. Tourne andC. M. Tourne, J. Chem. Soc., Chem. Commun., 1973, 139; (b) N. Casan-Pastor, J. Bas-Serra, E. Coronado, G. Pourroy and L. C. W. Baker,J. Am. Chem. Soc., 1992, 114, 10380.

4 (a) P. Mialane, A. Dolbecq, E. Riviere, J. Marrot and F. Secheresse,Angew. Chem., Int. Ed., 2004, 43, 2274; (b) P. Mialane, C. Duboc, J.Marrot, E. Riviere, A. Dolbecq and F. Secheresse, Chem.–Eur. J., 2006,12, 1950; (c) S. Nellutla, J. van Tol, N. S. Dalal, L.-H. Bi, U. Kortz,B. Keita, L. Nadjo, G. A. Khitrov and A. G. Marshall, Inorg. Chem.,2005, 44, 9795.

5 (a) S.-T. Zheng, D.-Q. Yuan, J. Zhang, H.-P. Jia and G.-Y. Yang, Chem.Commun., 2007, 1858; (b) S.-T. Zheng, M.-H. Wang and G.-Y. Yang,Chem.–Asian J., 2007, 2, 1380; (c) J.-W. Zhao, S.-T. Zheng and G.-Y. Yang, J. Solid State Chem., 2007, 180, 3317; (d) S.-T. Zheng, D.-Q. Yuan, J. Zhang and G.-Y. Yang, Inorg. Chem., 2007, 46, 4569; (e)S.-T. Zheng, J. Zhang and G.-Y. Yang, Angew. Chem., Int. Ed., 2008, 47,3909; (f) J.-W. Zhao, J. Zhang, S.-T. Zheng and G.-Y. Yang, Eur. J. Inorg.Chem., 2008, 3809; (g) J.-W. Zhao, J. Zhang, S.-T. Zheng andG.-Y. Yang, Inorg. Chem., 2007, 46, 10944; (h) C. Pichon, A. Dolbecq,P. Mialane, J. Marrot, E. Riviere and F. Secheresse, Dalton Trans.,2008, 71; (i) C. Pichon, A. Dolbecq, P. Mialane, J. Marrot, E. Riviere,M. Goral, M. Zynek, T. McCormac, S. A. Borshch, E. Zueva andF. Secheresse, Chem.–Eur. J., 2008, 14, 3189; (j) A. Dolbecq, J.-D.Compain, P. Mialane, J. Marrot, E. Riviere and F. Secheresse, Inorg.Chem., 2008, 47, 3371.

6 L. Bi, U. Kortz, S. Nellutla, A. C. Stowe, J. van Tol, N. S. Dalal, B.Keita and L. Nadjo, Inorg. Chem., 2005, 44, 896.

7 (a) G. M. Sheldrick, SHELXS97, Program for Crystal Structure Solu-tion, University of Gottingen, Gottingen, Germany, 1997; (b) G. M.Sheldrick, SHELXL97, Program for Crystal Structure Refinement,University of Gottingen, Gottingen, Germany, 1997.

8 U. Kortz, S. Nellutla, A. C. Stowe, N. S. Dalal, U. Rauwald, W.Danquah and D. Ravot, Inorg. Chem., 2004, 43, 2308.

9 S.-T. Zheng, M.-H. Wang and G.-Y. Yang, Inorg. Chem., 2007, 46,9503.

10 T. Yamase, K. Fukaya, H. Nojiri and Y. Ohshima, Inorg. Chem., 2006,45, 7698.

This journal is © The Royal Society of Chemistry 2009 Dalton Trans., 2009, 1300–1306 | 1305

Publ

ishe

d on

16

Janu

ary

2009

. Dow

nloa

ded

by U

nive

rsity

of

New

cast

le o

n 10

/08/

2013

11:

48:2

0.

View Article Online

11 L.-H. Bi, U. Kortz, S. Nellutla, A. C. Stowe, J. van Tol, N. S. Dalal, B.Keita and L. Nadjo, Inorg. Chem., 2005, 44, 896.

12 X. Fang, T. M. Anderson, C. Benelli and C. L. Hill, Chem.–Eur. J.,2005, 11, 712.

13 B. Moulton, H. Abourahma, M. W. Bradner, J. Lu, G. J. McManus andM. J. Zaworotko, Chem. Commun., 2003, 1342.

14 O. V. Dolomanov, A. J. Blake, N. R. Champness and M. Schroder,J. Appl. Crystallogr., 2003, 36, 1283.

15 (a) M. O’Keeffe, M. Eddaoudi, H. Li, T. Reineke and O. M.Yaghi, J. Solid State Chem., 2000, 152, 3; (b) M.-L. Tong,X.-M. Chen and S. R. Batten, J. Am. Chem. Soc., 2003, 125,16170.

16 (a) V. H. Crawford, H. V. Richardson, J. R. Wason, D. J. Hodgson andW. E. Hatfield, Inorg. Chem., 1976, 15, 2107; (b) L. Merz and W. Haase,J. Chem. Soc., Dalton Trans., 1980, 875; (c) M. Handa, N. Koga and S.Kida, Bull. Chem. Soc. Jpn., 1988, 61, 3853; (d) E. Ruiz, P. Alemany, S.Alvarez and J. Cano, J. Am. Chem. Soc., 1997, 119, 1297; (e) G. Aromı,J. Ribas, P. Gamez, O. Roubeau, H. Kooijman, A. L. Spek, S. Teat, E.MacLean, H. Stoeckli-Evans and J. Reedijk, Chem.–Eur. J., 2004, 10,

6476; (f) M. P. Shores, B. M. Bartlett and D. G. Nocera, J. Am. Chem.Soc., 2005, 127, 17986.

17 K. Kambe, J. Phys. Soc. Jpn., 1950, 48, 15.18 (a) J. J. Borras-Almenar, J. M. Clemente-Juan, E. Coronado and B. S.

Tsukerblat, Inorg. Chem., 1999, 38, 6081; (b) J. J. Borras-Almenar, J. M.Clemente-Juan, E. Coronado and B. S. Tsukerblat, J. Comput. Chem.,2001, 22, 985.

19 C.-F. Wang, J.-L. Zuo, B. M. Bartlett, Y. Song, J. R. Long and X.-Z.You, J. Am. Chem. Soc., 2006, 128, 7162.

20 U. Kortz, S. Nellutla, A. C. Stowe, N. S. Dalal, J. van Tol and B. S.Bassil, Inorg. Chem., 2004, 43, 144.

21 N. Zamstein, A. Tarantul and B. Tsukerblat, Inorg. Chem., 2007, 46,8851.

22 H. So, H. Jung, J.-H. Choy and R. L. Belford, J. Phys. Chem. B, 2005,109, 3324.

23 (a) S. Reinoso, P. Vitoria, J. M. Gutierrez-Zorrilla, L. Lezama, L. S.Felices and J. I. Beitia, Inorg. Chem., 2005, 44, 9731; (b) J.-W. Zhao,S.-T. Zheng, W. Liu and G.-Y. Yang, J. Solid State Chem., 2008, 181,637.

1306 | Dalton Trans., 2009, 1300–1306 This journal is © The Royal Society of Chemistry 2009

Publ

ishe

d on

16

Janu

ary

2009

. Dow

nloa

ded

by U

nive

rsity

of

New

cast

le o

n 10

/08/

2013

11:

48:2

0.

View Article Online