PAPER www.rsc.org/dalton | Dalton Transactions

Dinuclear nickel complexes with a Ni2O2 core: a structural and magnetic study

Matilde Fondo,*a Ana M. Garcıa-Deibe,a Noelia Ocampo,a Jesus Sanmartın,b Manuel R. Bermejob andAntonio L. Llamas-Saizc

Received 5th May 2006, Accepted 21st June 2006First published as an Advance Article on the web 4th July 2006DOI: 10.1039/b606414k

The acetylacetonate complexes [Ni2L1(acac)(MeOH)]·H2O, 1·H2O and [Ni2L3(acac)(MeOH)]·1.5H2O,2·1.5H2O (H3L1 = (2-(2-hydroxyphenyl)-1,3-bis[4-(2-hydroxyphenyl)-3-azabut-3-enyl]-1,3-imidazolidine and H3L3 = (2-(5-bromo-2-hydroxyphenyl)-1,3-bis[4-(5-bromo-2-hydroxyphenyl)-3-azabut-3-enyl]-1,3-imidazolidine) were prepared and fully characterised. Their crystal structures showthat they are dinuclear complexes, extended into chains by hydrogen bond interactions. Thesecompounds were used as starting materials for the isolation of the corresponding[Ni2HLx(o-O2CC6H4CO2)(H2O)]·nMeOH and [Ni2HLx(O2CCH2CO2)(H2O)]·nH2O dicarboxylatecomplexes (x = 1, 3; n = 1–3). The crystal structures of [Ni2HL1(o-O2CC6H4CO2)(H2O)]·MeOH,3·MeOH, [Ni2HL3(o-O2CC6H4CO2)(H2O)]·3MeOH, 4·3MeOH and [Ni2HL1(O2CCH2CO2)(H2O)]·2.5H2O·0.25MeOH·MeCN, 5·2.5H2O·0.25MeOH·MeCN, were solved. Complexes 3–5 show dinuclear[Ni2HLx(dicarboxylate)(H2O)] units, expanded through hydrogen bonds that involve carboxylate andwater ligands, as well as solvate molecules. The variable temperature magnetic susceptibilities of all thecomplexes show an intramolecular ferromagnetic coupling between the Ni(II) ions, which is attemptedto be rationalized by comparison with previous results and in the light of molecular orbital treatment.Magnetisation measurements are in accord with a S = 2 ground state in all cases.

Introduction

Current research concerning the structural and magnetic proper-ties of polynuclear 3d metal complexes is aimed at understandingthe structural and chemical features governing electronic exchangecoupling through bridging ligands. The greatest effort in thissense has been devoted to the study of dinuclear copper(II)complexes and, for symmetrically bridged Cu(II) compounds witha Cu2O2 core, a good degree of success has been accomplished incorrelating structure and magnetic properties.1–6 For diphenoxo-bridged dinuclear nickel(II) complexes some correlations were alsotried.7,8 In this sense, Nanda and co-workers proposed a linearcorrelation between the J value and the Ni–O–Ni angle (h),7 theangle crossover from antiferromagnetic to ferromagnetic couplingoccurring close to 97◦ for a series of macrocyclic dinuclearcomplexes. Bu et al.8 also attempted to correlate the bridgingNi–O–Ni angle with the J parameter in dinuclear symmetricallyO-bridged complexes, concluding that the relation is not linearand that ferromagnetic coupling is possible for angles smallerthan 93.5◦. In spite of these helpful correlations, a large numberof other structural parameters, such as the Ni–O distance, theangle between the Ni–O–Ni plane and the remaining coordinationplane, the apical distances, etc., seem to greatly influence the super-exchange mechanism in these systems. Therefore, a more accuratecorrelation between J and different structural parameters would

aDepartamento de Quımica Inorganica, Facultade de Ciencias, Universidadede Santiago de Compostela, E-27002, Lugo, SpainbDepartamento de Quımica Inorganica, Facultade de Quımica, Universidadede Santiago de Compostela, E-15782, Santiago de Compostela, SpaincUnidade de Raios X. RIAIDT. Vicerrectorado de Investigacion, Universi-dade de Santiago de Compostela, E-15782, Santiago de Compostela, Spain

be expected if more structural data on bis-oxo-bridged dinuclearNi(II) complexes were available.

We are interested in the coordination chemistry of a compart-mental Schiff base, which seems to yield dinuclear complexeswith an inherent ferromagnetic coupling.9–11 We have recentlyreported that the strength of this coupling in dinuclear copper(II)complexes with a Cu2O2 core is highly dependent on the anglebetween the basal coordination planes.11 Following with this work,we decided to investigated the magnetic coupling in dinuclearnickel(II) complexes with this type of ligand. Accordingly, asystematic method to isolate Ni(II) compounds, with a Ni2O2

metallacycle and similar coordination environments, was drawn.Their structural and magnetic features are reported herein.

Results and discussion

Synthesis

The acetylacetonate complexes [Ni2L1(acac)(MeOH)]·H2O,1·H2O and [Ni2L3(acac)(MeOH)]·1.5H2O, 2·1.5H2O (H3L1 andH3L3 in Scheme 1), were prepared by direct interaction betweennickel acetylacetonate and the corresponding Schiff base in 2 :1 molar ratio in methanol–acetonitrile (Scheme 2(a)). Thesecomplexes were used as precursors in the synthesis of dicarboxylateones. Therefore, 1·H2O and 2·1.5H2O were treated with malonicor o-phthalic acid, in 2 : 1 and 1 : 1 molar ratios (Scheme 2(b)).In all cases, the acetylacetonate precursors undergo a ligandexchange reaction. Thus, the acetylacetonate ligand is displaced bythe carboxylate ligand, the dinuclear nature of the coordinationcompounds being maintained independently of the precursor :acid molar ratio used. Accordingly, this seems a suitable syntheticapproach to isolate dinuclear nickel complexes with similarstructural features.

4260 | Dalton Trans., 2006, 4260–4270 This journal is © The Royal Society of Chemistry 2006

Scheme 1

Scheme 2

All the complexes were fully characterised by IR spectro-scopy, electrospray mass spectrometry and 1·H2O, 2·1.5H2O,[Ni2HL1(o-O2CC6H4CO2)(H2O)]·MeOH, 3·MeOH, [Ni2HL3-(o-O2CC6H4CO2H)(H2O)]·3MeOH, 4·3MeOH, and [Ni2HL1-(O2CCH2CO2)(H2O)]·2.5H2O·0.25MeOH·MeCN,5·2.5H2O·0.25MeOH·MeCN, by single-crystal X-ray diffractionstudies. The elemental analyses were recorded on dried crystalsand agree with the formulations 1·H2O, 2·1.5H2O, 3·MeOH,4·3MeOH, 5·2.5H2O and [Ni2HL3(O2CCH2CO2)(H2O)]·3H2O,6·3H2O, showing that 5·2.5H2O·0.25MeOH·MeCN loses the

most volatile solvents on drying. The IR spectra of the complexesshow a strong band in the range 1635–1645 cm−1, assigned tom(CNiminic), which suggest the coordination of the ligand to themetal ions through the imine nitrogen atoms. In addition, abroad band at ca. 3400 cm−1 agrees with the hydration of thecompounds. Besides, a shoulder at ca. 3270 cm−1 is observedfor 3–6, in accordance with the non-deprotonation of one ofthe phenol ligand groups. The mass spectra of all the complexesshow a peak of 100% intensity at ca. 571 or 809 m/z, due to thefragments [Ni2L1]+ or [Ni2L3]+, respectively, corroborating thedinuclear nature of 1–6 in solution. Peaks of lower intensitiesassigned to [Ni2Lx(acac)Na]+ or [Ni2HLx(dicarboxylate)Na]+

fragments can be also observed, indicating that the externalligand remains coordinated to the metal atoms in solution.

Crystallographic characterisation

[Ni2L1(acac)(MeOH)]·H2O and [Ni2L3(acac)(MeOH)]·1.5H2O.An ORTEP view of 1 and 2 is shown in Fig. 1. Experimental details

Fig. 1 An ORTEP view of the crystal structures of 1 and 2: R = H for 1and R = Br for 2. Hydrogen atoms have been omitted for clarity. Ellipsoidsare drawn at 40% probability.

Table 1 Crystal data and structure refinement for 1·H2O to 5·2.5H2O·0.25MeOH·MeCN

1·H2O 2·1.5H2O 3·MeOH 4·3MeOH 5·2.5H2O·0.25MeOH·MeCN

Empirical formula C33H40N4O7Ni2 C33H38Br3N4O7.5Ni2 C36H38N4O9Ni2 C38H43Br3N4O11Ni2 C32.25H41N5O10.75Ni2

Mr 722.11 967.82 788.12 1088.91 788.12T/K 120(2) 120(2) 120(2) 120(2) 293(2)k/A 0.71073 0.71073 0.71073 0.71073 0.71073Crystal system Monoclinic Orthorhombic Monoclinic Monoclinic MonoclinicSpace group P21/n Pbca P21/c P21/n C2/ca/A 11.569(3) 20.515(5) 15.899(2) 16.6463(16) 44.353(5)b/A 20.834(5) 14.726(5) 11.4993(17) 10.6223(10) 8.850(5)c/A 14.508(3) 24.311(5) 19.048(3) 23.150(2) 19.517(5)b/◦ 112.486(3) 90 108.36(1) 100.273(2) 112.433(5)Z 4 8 4 4 8l/mm−1 1.219 4.342 1.204 3.976 1.128Crystal size/mm 0.40 × 0.16 × 0.06 0.25 × 0.10 × 0.03 0.33 × 0.26 × 0.15 0.43 × 0.06 × 0.04 0.30 × 0.25 × 0.25Reflections collected 52423 131465 31327 6184 35842Independent reflections (Rint) 6682 (0.0503) 6491 (0.0848) 7984 (0.0327) 5811 (0.0579) 6581 (0.0976)Data/restraints/parameters 6682/0/429 6491/0/491 7984/0/477 5811/0/525 6581/1/466Final R indices [I > 2r(I)] R1 = 0.0401 R1 = 0.0767 R1 = 0.0337 R1 = 0.0504 R1 = 0.0563

wR2 = 0.0904 wR2 = 0.2122 wR2 = 0.0666 wR2 = 0.0983 wR2 = 0.1338R indices (all data) R1 = 0.0668 R1 = 0.0809 R1 = 0.0599 R1 = 0.1052 R1 = 0.0921

wR2 = 0.0988 wR2 = 0.2173 wR2 = 0.0770 wR2 = 0.1125 wR2 = 0.1520

This journal is © The Royal Society of Chemistry 2006 Dalton Trans., 2006, 4260–4270 | 4261

are given in Table 1 and main distances and angles are listed inTables 2 and 3.

Table 2 Main distances (A) and angles (◦) for 1·H2O

Ni(11)–O(101) 1.997(2) Ni(12)–O(102) 1.995(2)Ni(11)–N(101) 2.004(2) Ni(12)–N(102) 1.998(3)Ni(11)–N(103) 2.157(2) Ni(12)–N(104) 2.175(2)Ni(11)–O(11) 2.064(2) Ni(12)–O(12S) 2.146(2)Ni(11)–O(103) 2.078(2) Ni(12)–O(103) 2.0618(19)Ni(11)–O(10) 2.029(2) Ni(12)–O(10) 2.086(2)Ni(11) · · · Ni(12) 3.0825(10)

O(101)–Ni(11)–N(101) 89.38(9) O(102)–Ni(12)–N(102) 91.60(9)O(101)–Ni(11)–N(103) 172.72(9) O(102)–Ni(12)–N(104) 174.19(9)O(101)–Ni(11)–O(11) 87.79(8) O(102)–Ni(12)–O(12S) 89.98(8)O(101)–Ni(11)–O(103) 93.21(8) O(102)–Ni(12)–O(103) 92.83(8)O(101)–Ni(11)–O(10) 91.50(8) O(102)–Ni(12)–O(10) 94.03(8)N(101)–Ni(11)–O(11) 87.72(9) N(102)–Ni(12)–O(12S) 89.84(9)N(101)–Ni(11)–N(103) 83.43(9) N(102)–Ni(12)–N(104) 82.60(9)N(101)–Ni(11)–O(10) 175.95(9) N(102)–Ni(12)–O(10) 174.28(9)N(101)–Ni(11)–O(103) 103.05(9) N(102)–Ni(12)–O(103) 98.81(9)O(10)–Ni(11)–O(11) 88.36(8) O(10)–Ni(12)–O(12S) 91.17(8)O(10)–Ni(11)–N(103) 95.60(8) O(10)–Ni(12)–N(104) 91.78(8)O(10)–Ni(11)–O(103) 80.86(8) O(10)–Ni(12)–O(103) 79.92(8)O(11)–Ni(11)–N(103) 90.78(8) O(12S)–Ni(12)–N(104) 89.93(8)O(103)–Ni(11)–O(11) 169.19(8) O(103)–Ni(12)–O(12S) 170.82(8)O(103)–Ni(11)–N(103) 89.53(8) O(103)–Ni(12)–N(104) 88.16(8)Ni(11)–O(103)–Ni(12) 96.24(8) Ni(11)–O(10)–Ni(12) 97.00(8)

Table 3 Main distances (A) and angles (◦) for 2·1.5H2O

Ni(11)–O(101) 2.014(4) Ni(12)−O(102) 2.014(4)Ni(11)–N(101) 2.019(5) Ni(12)−N(102) 2.014(5)Ni(11)–N(103) 2.164(5) Ni(12)−N(104) 2.198(4)Ni(11)–O(11) 2.047(5) Ni(12)−O(12S) 2.128(4)Ni(11)–O(103) 2.093(4) Ni(12)−O(103) 2.079(4)Ni(11)–O(10) 2.032(4) Ni(12)−O(10) 2.095(4)Ni(11) · · · Ni(12) 3.0934(18)

O(101)–Ni(11)–N(101) 89.89(19) O(102)–Ni(12)–N(102) 90.66(18)O(101)–Ni(11)–N(103) 173.40(18) O(102)–Ni(12)–N(104) 173.31(17)O(101)–Ni(11)–O(11) 88.37(17) O(102)–Ni(12)–O(12S) 91.12(17)O(101)–Ni(11)–O(103) 94.22(16) O(102)–Ni(12)–O(103) 92.96(16)O(101)–Ni(11)–O(10) 90.14(16) O(102)–Ni(12)–O(10) 94.91(16)N(101)–Ni(11)–O(11) 87.7(2) N(102)–Ni(12)–O(12S) 90.32(18)N(101)–Ni(11)–N(103) 84.0(2) N(102)–Ni(12)–N(104) 82.68(19)N(101)–Ni(11)–O(10) 176.8(2) N(102)–Ni(12)–O(10) 174.35(18)N(101)–Ni(11)–O(103) 102.1(2) N(102)–Ni(12)–O(103) 98.77(18)O(10)–Ni(11)–O(11) 89.10(18) O(10)–Ni(12)–O(12S) 90.50(17)O(10)–Ni(11)–N(103) 95.87(18) O(10)–Ni(12)–N(104) 91.73(17)O(10)–Ni(11)–O(103) 81.17(16) O(10)–Ni(12)–O(103) 80.04(15)O(11)–Ni(11)–N(103) 89.0(2) O(12S)–Ni(12)–N(104) 89.62(17)O(103)–Ni(11)–O(11) 169.93(18) O(103)–Ni(12)–O(12S) 169.99(18)O(103)–Ni(11)–N(103) 89.42(18) O(103)–Ni(12)–N(104) 87.40(17)Ni(11)–O(103)–Ni(12) 95.71(16) Ni(11)–O(10)–Ni(12) 97.08(16)

Both complexes are very similar and they will be discussedtogether. The main difference between them is that the centralaromatic ring, as well as the C107–C108 carbon chain, is disor-dered over two sites in 2·1.5H2O, while this does not occur in1·H2O. The crystal structures of 1·H2O and 2·1.5H2O show thatthey consist of dinuclear [Ni2Lx(acac)(MeOH)] units (Fig. 1) withwater as solvate. The heptadentate (Lx)3− Schiff base (x = 1 or 3)accommodates two nickel ions into its two N2O compartments.Each one of them is formed by contiguous imine (N101 or N102)and imidazolidine nitrogen atoms (N103 or N104) and a terminalphenol O atom (O101 or O102), with the imidazolidinic NCNgroup (N103C120N104) spanning both nickel ions. In addition,the central phenol oxygen atom (O103) bridges both metal atoms.The coordination spheres of the nickel centres are completed bya methanol molecule and an exogenous acetylacetonate ligand.The latter ligand is acting as bridging bidentate in a l2-g2:g1-O,Omode. One carbonyl oxygen atom (O10) is bridging both metalions, while the other one (O11) is only coordinated to one nickelatom (Ni11). The sixth coordination position on Ni12 is occupiedby a methanol molecule. Thus, the nickel centres are triple bridgedby the NCN group, one endogenous and one exogenous oxygenatoms. The double O-bridge gives rise to a Ni2O2 metallacycle,with intramolecular Ni11 · · · Ni12 separations of ca. 3.08 A and aNi–Oph–Ni angle slightly lower than the Ni–Ocarbonyl–Ni one.

As a result of the structural features described, the complexesare asymmetric, with both nickel ions N2O4 hexacoordinated.Distances and angles around the metal centres (Tables 2 and3) agree with distorted octahedral environments, with bothpolyhedra sharing an edge.

Finally, it could be of note that the coordinated methanolmolecule and the water solvate are implicated in hydrogenbonds, which contribute to expand the initial dinuclear nodes.Accordingly, the water solvate gives rise to three interactions:with both terminal phenol oxygen atoms of the dinuclear unit[distances O101 · · · O1w and O102 · · · O1w of ca. 2.7 A] and withthe coordinated methanol molecule of a different unit [distanceO1w · · · O1s# of 2.634(3) A]. These interactions through the nodesexpand the dinuclear blocks into chains (Fig. 2).

[Ni2HL1(o-O2CC6H4CO2)(H2O)]·MeOH, 3·MeOH, [Ni2HL3-(o-O2CC6H4CO2)(H2O)]·3MeOH, 4·3MeOH and [Ni2HL1(O2-CCH2CO2)(H2O)]·2.5H2O·0.25MeOH·MeCN, 5·2.5H2O·0.25MeOH·MeCN. The crystal structures of the threedicarboxylate complexes are very similar and they will bediscussed together.

Fig. 2 A view of the packing of 1, showing the expansion of the blocks into a chain. H atoms have been omitted for clarity.

4262 | Dalton Trans., 2006, 4260–4270 This journal is © The Royal Society of Chemistry 2006

ORTEP views of 1, 2 and 3 are shown in Fig. 3 and 4.Experimental details are given in Table 1 and main distances andangles are listed in Tables 4–6.

Fig. 3 An ORTEP view of the crystal structure of 3 and 4: R = H for 3and R = Br for 4 attached to C103, C116 and C123. Hydrogen atoms havebeen omitted for clarity. Ellipsoids are drawn at 40% probability.

Fig. 4 An ORTEP view of the crystal structure of 5. Hydrogen atomshave been omitted for clarity. Ellipsoids are drawn at 50% probability.

The crystal structures of all the complexes show that theasymmetric unit contains dinuclear molecules of the type[Ni2HLx(O2CCnHmCO2)(H2O)] (x = 1 or 3; n = 6, m = 4 or n= 1, m = 2), in addition to methanol, water and/or acetonitrileas solvates. In all the cases, the Schiff base acts in a similar way tothat observed in 1 and 2. Thus, the compartmental heptadentateligand provides an N2O2 donor set around each nickel atom.Curiously in these derivatives, the two pockets of the Schiff baseare not equivalent, as only one of the terminal phenol groups(O101) and the central one (O103) seem to be deprotonated, whilstterminal O102 remains protonated. Thus, these ligands show theuncommon feature of behaving as dianionic since, as far as weknow, this type of ligand had been hitherto reported as trianionicin all the complexes,9–12 including 1 and 2, although this behaviouris not unfamiliar for other multidentate O,N donor ligands.13 Thishypothesis is corroborated not only by the Ph–O and Ph–OH

Table 4 Main distances (A) and angles (◦) for 3·MeOH

Ni(11)–O(101) 1.9976(16) Ni(12)–O(102) 2.0534(16)Ni(11)–N(101) 1.9983(19) Ni(12)–N(102) 2.0076(19)Ni(11)–O(1W) 2.1154(17) Ni(12)–O(12) 2.0568(16)Ni(11)–N(103) 2.1525(18) Ni(12)–N(104) 2.1599(19)Ni(11)–O(103) 2.0468(15) Ni(12)–O(103) 2.0383(15)Ni(11)–O(10) 2.1011(15) Ni(12)–O(10) 2.0620(15)Ni(11) · · · Ni(12) 3.0894(6)

O(101)–Ni(11)–N(101) 90.95(7) O(102)–Ni(12)–N(102) 87.92(7)O(101)–Ni(11)–N(103) 174.98(7) O(102)–Ni(12)–N(104) 171.30(7)O(101)–Ni(11)–O(103) 91.02(6) O(102)–Ni(12)–O(103) 88.40(6)O(101)–Ni(11)–O(10) 90.91(6) O(102)–Ni(12)–O(10) 92.28(6)O(101)–Ni(11)–O(1W) 86.70(7) O(102)–Ni(12)–O(12) 97.36(7)N(101)–Ni(11)–O(103) 102.84(7) N(102)–Ni(12)–O(103) 100.41(7)N(101)–Ni(11)–O(1W) 88.27(7) N(102)–Ni(12)–O(12) 89.55(7)N(101)–Ni(11)–O(10) 177.37(7) N(102)–Ni(12)–O(10) 179.45(7)O(103)–Ni(11)–O(10) 78.99(6) O(103)–Ni(12)–O(10) 80.10(6)O(103)–Ni(11)–O(1W) 168.70(7) O(103)–Ni(12)–O(12) 168.70(6)O(103)–Ni(11)–N(103) 88.99(6) O(103)–Ni(12)–N(104) 88.83(7)O(10)–Ni(11)–O(1W) 89.97(6) O(10)–Ni(12)–O(12) 89.92(6)O(10)–Ni(11)–N(103) 94.02(6) O(10)–Ni(12)–N(104) 95.37(6)N(101)–Ni(11)–N(103) 84.16(7) N(102)–Ni(12)–N(104) 84.47(7)O(1W)–Ni(11)–N(103) 94.24(7) O(12)–Ni(12)–N(104) 86.81(7)Ni(11)–O(103)–Ni(12) 98.27(6) Ni(11)–O(10)–Ni(12) 95.81(6)

Table 5 Main distances (A) and angles (◦) for 4·3MeOH

Ni(11)–O(101) 1.987(5) Ni(12)–O(102) 2.082(5)Ni(11)–N(101) 1.994(6) Ni(12)–N(102) 2.017(6)Ni(11)–N(103) 2.162(6) Ni(12)–N(104) 2.154(6)Ni(11)–O(1W) 2.111(5) Ni(12)–O(12) 2.042(5)Ni(11)–O(103) 2.049(5) Ni(12)–O(103) 2.033(5)Ni(11)–O(10) 2.089(5) Ni(12)–O(10) 2.059(5)Ni(11) · · · Ni(12) 3.0680(14)

O(101)–Ni(11)–N(101) 90.6(2) O(102)–Ni(12)–N(102) 83.8(2)O(101)–Ni(11)–N(103) 173.2(2) O(102)–Ni(12)–N(104) 165.2(2)O(101)–Ni(11)–O(103) 89.0(2) O(102)–Ni(12)–O(103) 85.55(19)O(101)–Ni(11)–O(10) 88.5(2) O(102)–Ni(12)–O(10) 97.2(2)O(101)–Ni(11)–O(1W) 88.0(2) O(102)–Ni(12)–O(12) 99.1(2)N(101)–Ni(11)–O(103) 103.2(2) N(102)–Ni(12)–O(103) 105.0(2)N(101)–Ni(11)–O(1W) 88.9(2) N(102)–Ni(12)–O(12) 88.1(2)N(101)–Ni(11)–O(10) 177.5(2) N(102)–Ni(12)–O(10) 174.7(2)N(101)–Ni(11)–N(103) 83.7(2) N(102)–Ni(12)–N(104) 84.3(2)O(103)–Ni(11)–O(1W) 167.6(2) O(103)–Ni(12)–O(12) 166.5(2)O(103)–Ni(11)–O(10) 79.1(2) O(103)–Ni(12)–O(10) 80.2(2)O(103)–Ni(11)–N(103) 88.7(2) O(103)–Ni(12)–N(104) 89.1(2)O(10)–Ni(11)–O(1W) 88.73(19) O(10)–Ni(12)–O(12) 86.6(2)O(10)–Ni(11)–N(103) 97.4(2) O(10)–Ni(12)–N(104) 95.4(2)O(1W)–Ni(11)–N(103) 95.7(2) O(12)–Ni(12)–N(104) 89.2(2)Ni(11)–O(103)–Ni(12) 97.5(2) Ni(11)–O(10)–Ni(12) 95.4(2)

distances, which average 1.310 and 1.357 A, respectively, and in ourcase are d(C100–O101) ca. 1.31 A and d(C119–O102) ca. 1.35 A,but also by the Ni–O102Hterminal distance, which is significantlylonger than the Ni–O101terminal one in all cases (Tables 4–6),both being in the typical ranges described for protonated anddeprotonated Ni–OPh bonds, respectively.13

The coordination spheres of the nickel ions are completedby a water molecule (O1w), linked as terminal ligand, and byan exogenous bridging dicarboxylate ligand. The dianion ofthe dicarboxylic acid is also acting as a l2-g2:g1-O,O donor,a scarcely described binding mode for this type of ligand.14

In these complexes, one oxygen atom of one carboxylic group(O10) is bridging both metal centres, while an oxygen atom ofthe other carboxylic group (O12) is bound to only one of thenickel ions. The other two carbonyl oxygen atoms of the groups

This journal is © The Royal Society of Chemistry 2006 Dalton Trans., 2006, 4260–4270 | 4263

Table 6 Main distances (A) and angles (◦) for 5·2.5H2O·0.25MeOH·MeCN

Ni(11)–N(101) 1.994(4) Ni(12)–N(102) 2.016(4)Ni(11)–O(101) 2.024(3) Ni(12)–O(102) 2.061(4)Ni(11)–N(103) 2.236(3) Ni(12)–N(104) 2.130(4)Ni(11)–O(12) 2.083(3) Ni(12)–O(12W) 2.067(4)Ni(11)–O(103) 2.086(3) Ni(12)–O(103) 2.028(3)Ni(11)–O(10) 2.044(3) Ni(12)–O(10) 2.063(3)Ni(11) · · · Ni(12) 3.0989(19)

O(101)–Ni(11)–N(101) 91.17(143) O(102)–Ni(12)–N(102) 85.52(16)O(101)–Ni(11)–N(103) 173.09(13) O(102)–Ni(12)–N(104) 168.84(15)O(101)–Ni(11)–O(12) 95.4813) O(102)–Ni(12)–O(12W) 90.59(17)O(101)–Ni(11)–O(10) 93.59(12) O(102)–Ni(12)–O(10) 93.80(14)O(101)–Ni(11)–O(103) 91.84(13) O(102)–Ni(12)–O(103) 87.36(14)N(101)–Ni(11)–O(12) 94.44(13) N(102)–Ni(12)–O(12W) 88.70(17)N(101)–Ni(11)–O(103) 98.97(13) N(102)–Ni(12)–O(103) 104.40(13)N(101)–Ni(11)–O(10) 174.58(13) N(102)–Ni(12)–O(10) 176.29(13)N(101)–Ni(11)–N(103) 82.56(14) N(102)–Ni(12)–N(104) 84.00(15)O(10)–Ni(11)–O(12) 87.69(12) O(10)–Ni(12)–O(12W) 87.65(15)O(10)–Ni(11)–O(103) 78.30(11) O(10)–Ni(12)–O(103) 79.20(11)O(10)–Ni(11)–N(103) 92.56(12) O(10)–Ni(12)–N(104) 96.90(13)O(12)–Ni(11)–O(103) 164.58(12) O(12W)–Ni(12)–O(103) 166.52(15)O(12)–Ni(11)–N(103) 87.93(13) O(12W)–Ni(12)–N(104) 92.98(16)O(103)–Ni(11)–N(103) 86.30(12) O(103)–Ni(12)–N(104) 91.54(12)Ni(12)–O(103)–Ni(11) 97.72(12) Ni(11)–O(10)–Ni(12) 97.96(12)

(O11 and O13) remain uncoordinated, but participating in intra-and intermolecular H-bonds.

Therefore, the two metal centres are hexacoordinated inN2O4 octahedral environments, and triple bridged by theN103C120N104 Schiff base moiety, the endogenous O103 atomand the exogenous carboxylate ligand. The Ni–O and Ni–Ndistances, as well as the angles around the metal atoms, areindicative of quite regular octahedra, with a Ni–Ophenolate–Ni angleslightly more open than that of the Ni–Ocarboxylate–Ni bridge for 3and 4. This double O-bridge gives rise to a nearly planar Ni2O2

metallacycle, with an intramolecular Ni · · · Ni distance in the range3.07–3.1 A.

The main difference between the three dicarboxylate complexesis maybe found in the crystal packing. In all cases, an intricatehydrogen bond scheme contributes to expand the dinuclear blocks,which mainly results from the intermolecular interaction betweenthe coordinated water molecule O1w and the outer carboxy O13atom of an adjacent dinuclear block. However, the expansionof the blocks is different for the dissimilar complexes. Thus, theinteraction of one free oxygen atom of the carboxylate ligand withthe coordinated water molecule of a neighbouring block in 3, andvice versa, leads to a pseudo-tetranuclear complex, by interaction

of two dinuclear nodes. The growth of the blocks is higher for4. Thus, in this case, apart from the interaction between one freeoxygen atom of the carboxylate ligand with the coordinated watermolecule of a neighbouring block, the carboxylate ligand of thesecond unit also interacts with the water molecule of a third unitand so on, expanding the dinuclear nodes into a helical array,positioned parallel to b (Fig. 5).

In the case of 5, the network is also expanded throughinteractions between free oxygen carboxylate atoms and watermolecules. In this case, a water solvate is hydrogen bonded to theterminal phenol oxygen atoms of a [Ni2HL1(O2CCH2CO2)(H2O)]block of the same asymmetric unit and to an uncoordinatedcarboxylate oxygen atom of a neighbouring dinuclear unit. Thisinteraction is repeated in the space to generate a chain. Inaddition, two of these chains are linked through hydrogen bondsbetween the coordinated water of one dinuclear unit and/or theremaining uncoordinated carboxylate oxygen atom and watersolvates (Fig. 6), generating a double chain. Consequently, theseare further examples of the role of hydrogen bond interactions forconnecting building blocks into extended networks.

Magnetic properties

The magnetic properties of 1·H2O to 6·3H2O have been inves-tigated in the 2–300 K temperature range and some plots ofvMT vs. T are shown in Fig. 7–9, as examples. In all cases, thevMT value at 300 K (varying from 2.15 to 2.72 cm3 mol−1 K)increases upon cooling, to reach a maximum at low temperature,and then decreases with decreasing temperature for all complexes.The observed behaviour is indicative of a ferromagnetic interactionbetween the Ni(II) ions; the decreasing in vMT at low temperaturesbeing attributed to either interdinuclear antiferromagnetic inter-actions and/or the effect of the zero field splitting (ZFS) of the S= 2 ground state.

The experimental data were fitted with expression (1), derivedfrom the Van Vleck formula for two S = 1 ions,15 based on thespin Hamiltonian H = −JS1S2, and modified for intermolecularinteractions, in agreement with the crystal structures:

vMT = Ng2b2Tf (x)kT − J ′f (x)

(1)

with

f (x) = 2 exp(J/kT) + 10 exp(3J/kT)1 + 3 exp(J/kT) + 5 exp(3J/kT)

where the symbols have their usual meaning.

Fig. 5 A view (omitting H atoms for clarity) of a fragment of the H-bonded helical array of 4, almost along the helical axis parallel to b.

4264 | Dalton Trans., 2006, 4260–4270 This journal is © The Royal Society of Chemistry 2006

Fig. 6 A view (omitting H atoms for clarity) of a fragment of the 1D array of 5.

Fig. 7 A plot of vMT vs. T for 1·H2O: scattered points: experimentalresults; dashed line: best fitting.

Fig. 8 A plot of vMT vs. T for 3·MeOH: scattered points: experimentalresults; dashed line: best fitting.

The best least-squares fit gave the parameters shown in Table 7,the J values being consistent with the similar strength of the

Fig. 9 A plot of vMT vs. T for 5·2.5H2O: scattered points: experimentalresults; dashed line: best fitting.

resultant ferromagnetic interaction in all cases. This ferromagneticcoupling is confirmed by the field dependence of the magnetisationM = f (H) at 2 K. The reduced magnetisation M/Nb varies linearlywith low applied fields up to ca. 1 T and then progressively tendsto 4.0, which points to an S = 2 ground state for 1·H2O to 6·3H2O.

Since all these complexes are ferromagnetic, some attempts weremade to rationalize the found J values. First of all, it should benoted that the crystal structures of 1–5 show that the nickel atomsare tripled bridged: by the endogenous Ophenol atom, the exogenousOacac or Ocarboxylate and by the N103C120N104 moiety of the Schiffbase ligand. Thus, these complexes contain a Ni2O2 core, in ad-dition to an NCN bridge. To the best of our knowledge, just threeprevious papers try to correlate the Ni–O–Ni angle (h) with the Jvalue in dinuclear complexes with a Ni2O2 core,7,8,16 all of them arebased on symmetrically diphenoxo-bridged complexes, one beinga quantum-chemical ab initio study.16 The first correlation makesuse of macrocyclic ligands and the accidental orthogonalityoccurs close to 97◦.7 The second correlation8 is not linear andthe authors propose that ferromagnetic coupling is possible forNi–O–Ni angles smaller than 93.5◦, and the theoretical study

This journal is © The Royal Society of Chemistry 2006 Dalton Trans., 2006, 4260–4270 | 4265

Table 7 Magnetic data for 1–6

Complex J/cm−1 J ′/cm−1 g Ra

[Ni2L1(acac)(MeOH)]·H2O, 1·H2O 5.53 −0.50 2.22 9.6 × 10−4

[Ni2L3(acac)(MeOH)]·1.5H2O, 2·1.5H2O 5.23 −0.62 2.31 3.6 × 10−4

[Ni2HL1(phthal)(H2O)]·MeOH, 3·MeOH 4.88 −0.41 2.18 7.0 × 10−5

[Ni2HL3(phthal)(H2O)]·3MeOH, 4·3MeOH 6.60 −0.35 2.25 8.0 × 10−5

[Ni2HL1(mal)(H2O)]·2.5H2O, 5·2.5H2O 5.48 −0.35 2.06 2.3 × 10−4

[Ni2HL3(mal)(H2O)]·3H2O, 6·3H2O 6.91 −0.26 2.23 4.4 × 10−4

a R = ∑(vMT exp − vMT calc)2/

∑(vMT exp)2.

claims that ferromagnetic coupling should occurs between 85 and102◦.

The three studies point to dependence of the J parameter withthe Ni–O–Ni angle but it seems that there is not agreement in theangle at which accidental orthogonality occurs. For this reason,a revision of all the dinuclear nickel(II) complexes crystallograph-ically and magnetically characterised, just containing a single-atom double O-bridge, was carried out (Table 8),7–8,17–28 in orderto see if the experimental results previously described shed somelight about this question. When the J value is represented as afunction of the average Ni–O–Ni angle (Fig. 10), it seems clearthat the antiferromagnetic contribution dramatically increases forangles higher than 102◦, but it also becomes apparent that h is notthe only factor that affects the superexchange pathway. Hence, asimple relationship for the J value and the bridging angle doesnot seem to be applicable, even in those cases where the exchangecoupling occurs only through the equatorial plane.17–23,25–28 In fact,among these complexes, only three of them show ferromagneticcoupling.17,24 This ferromagnetic interaction can be easily justifiedfor one of the complexes, which shows an apical–basal/basal–apical distribution of the O-bridges,24 but it does not seem tohave an evident explanation for the other two cases, as othercomplexes with apparently similar structural features render a netantiferromagnetic coupling.18

In our case study, in addition to the oxygen bridges, thereis an NCN link, as mentioned above, overall providing two

Fig. 10 A plot of J vs. Ni–O–Ni angle for related dinuclear nickelcomplexes with a Ni2O2 core (Table 8).

different pathways (axial and equatorial) for exchange coupling.Consequently, the situation is more complicated in the complexesunder study and they are not comparable from the magneticpoint of view with those previously indicated.7–8,17–28 Accordingly,a search of dinuclear nickel complexes with a l-g1:g1-O,O bridge,in addition to the Ni2O2 core, was made, as the mechanism

Table 8 Comparison of some magnetostructural parameters for complexes containing a Ni2O2 corea

Complex Ni–Oph/A Ni–Oph–Ni/◦ Type of ligand Geometry Jb/cm−1 Ref.

[Ni2(ips)2(NO3)2(EtOH)2] 1.983(1), 2.053(1) 96.4(1) Acyclic Oh +26.4 17[Ni2(H2L1)(H2O)4](ClO4)2 2.017(1), 2.021(1) 98.12(6) Acyclic Oh −34.6 18[Ni(ps)(NO3)2(2-pic)]2 2.002(3), 2.060(3) 98.7(1) Acyclic Oh +18 17[Ni2(Hhsae)2(OOCCH3)2(MeOH)2] 2.0160(9), 2.0530(9) 98.95(4) Acyclic Oh −13.8 19[Ni(ips)(NO3)2(DMF)]2 1.990(2), 2.065(1) 99.0(1) Acyclic Oh −18.4 17[Ni2L(NCS)2(H2O)2]2 2.047(2), 2.032(3) 99.2(1) Macrocyclic Oh −42.6 20[Ni2L(H2O)4][ClO4]2 2.031 99.5 Macrocyclic Oh −34 20,21[Ni2(L1)2(OH2)2]ClO4 2.029(7), 2.017(5) 99.5(3) Acyclic Oh −62 22[Ni2(L − H)2]ClO4 1.988(4), 1.993(4)/ 99.75(16) Acyclic SPY −21.84 23

1.993(4), 1.977(4) 99.86(15)[Ni2(HL)(MeOH)2](ClO4)2 2.019(3) 99.9(1) Acyclic Oh

c +14.2 24[Ni2(HL1)2Cl2(MeOH)2] 2.017(2), 2.024(2) 100.75(8) Acyclic Oh −9.2 25[Ni2L(CH3OH)2(ClO4)2] 2.028(4), 2027(3) 101.3(1) Macrocyclic Oh −59 20,26[Ni2(H2L1)(H2O)2Cl2] 2.014(5) 101.5(3) Macrocyclic Oh −54 27[Ni2L1(N3)]2 2.004(3), 2.028(3) 101.57(13) Acyclic SPY −45.8 8[Ni2L(imidazole)2](ClO4)2 2.022(6), 2.015(6) 104.1(3) Macrocyclic SPY −99.6 7[Ni2L(pyridine)2](ClO4)2 2.011 105.7 Macrocyclic SPY −134.2 20,28

a All complexes show just two bridging O-atoms in a Ni(l-OPh)2Ni core; for identity of the ligands refer to the appropriate publication. b J values referredto Hamiltonian H = −JS1S2. c Basal–apical/apical–basal distribution of the Ni(l-OPh)2Ni core.

4266 | Dalton Trans., 2006, 4260–4270 This journal is © The Royal Society of Chemistry 2006

Table 9 Comparison of some magnetostructural parameters for complexes containing a Ni2O2 core and a l-g1:g1-O,O carboxylate bridge

Complex Ni–O/A Ni–O–Ni/◦ Type of ligand Geometry Ja/cm−1 Ref.

[Ni2L(O2CCH3)(H2O)]ClO4 2.059(4), 2.064(4) 95.1(2) Macrocyclic Oh +5.8 292.158(4), 2.124(5) 90.6(3)

[LNi2(NH2CONH2)(OOCCH3)(CH3OH)2] 1.971(3), 1.986(3) 97.1(2) Acyclic Oh −7 302.080(3), 2.065(3) 91.4(1)

pipH[Ni2(PhSal)4(CH3COO)] 2.033(4), 2.045(5) 97.3(2) Acyclic Oh −9.32 31[Ni2L3(OAc)(H2O)2]b 2.070(4), 2.071(4) 99.53(18), 97.29(18) Acyclic Oh +3.6 32

2.105(4), 2.106(5)[Ni2(bpepO)(CH3COO)(H2O)2] 2.030(3), 2.036(3) 102.53(13) Acyclic Oh −10.6 33

2.176(3), 2.394(3) 87.76(10)

a J values referred to Hamiltonian H = −JS1S2. b Additional l-g1:g1-N,N instead of l-g1:g1-O,O bridge.

of magnetic exchange should be more similar (Table 9).29–33 Asexpected, there appears to be no correlation of the J parameterwith the Ni–O–Ni angle and/or the Ni–Oph distances too, and, aspreviously stated, many other parameters, like the angle betweenthe Ni–O–Ni plane and the remaining coordination plane, theapical distances, etc., can cooperatively work to affect the exchangemechanism.8 In spite of this, it should be noted that the percentageof crystallographically solved ferromagnetic complexes is higherin this case.

Therefore, the comparison of 1–5 with related compounds doesnot allow explaining their magnetic behaviour. For this reason, andin order to throw some light about the systematic ferromagneticcoupling for 1 to 6, extended Huckel calculations were carriedout with the CACAO program,34 using the crystallographiccoordinates for the five solved complexes (1–5), and typical frontierMOs are shown in Fig. 11. The most interesting result of thesecomputations is that they demonstrate that the bridging Oacac orOcarboxylate atom is in the equatorial plane while the bridging Ophenol

atom is a common apical site of both octahedrons for all thecomplexes.

Fig. 11 Frontier molecular orbitals for 1, exemplifying the situationfound for all the complexes and showing the basal–apical disposition ofthe O-bridges.

From these calculations, it emerges that in this type ofcomplex the Ni2O2 core does not lay on the xy plane.To the best of our knowledge, apart from the complex[Ni2L3(OAc)(H2O)2]·3MeCN·2H2O (7) reported by us,32 there isonly one previous example of this type of geometric distribution33

and, in the related case, one of the apical Ni–O distances (2.39A) seems too long for being considered as a true coordinationbond and, hence, to mediate the magnetic coupling. In fact, eventhe authors suggest that the polyhedron about one of the metalatoms may be understood as a highly distorted octahedron or as asquare pyramid. Consequently, and as far as we know, this is thefirst series of dinuclear Ni complexes that unequivocally comprisesthis feature.

Thus, these complexes contain an oxygen bridge and an NCNlink in the basal plane, in addition to the apical Ophenol bridge(Fig. 12). Therefore, we could tentatively ascribe the ferromagneticcoupling to the counter-complentarity effect of both basal bridges,as has been demonstrated for copper complexes.10 However, inthis case there is an additional apical superexchange pathway andperhaps this counter-complentarity effect is not the only reasonfor the parallel electron alignment.

Fig. 12 Geometric disposition of the Ni environments in 1–5.

Hence, as these complexes seem to be unique in their class,we have tried to see if there are some specific factors thatespecially affect the J value. So, multiple structural parameterswere analysed for 1–5 and the related complex 7, some of themsummarised in Table 10. Unfortunately, in spite of the efforts madein an attempt to rationalize the trend in the J value, no simpletendency seems to emerge and that further examples of this typeof compound are needed to try to find any structural/magneticcoupling trend. Nevertheless, it could be of note that the highest

This journal is © The Royal Society of Chemistry 2006 Dalton Trans., 2006, 4260–4270 | 4267

Tab

le10

Som

est

ruct

ural

and

mag

neti

cpa

ram

eter

sfo

rco

mpl

exes

1–5

and

7a

Com

plex

Ni–

OP

h–N

i(◦ )/

Ni–

Oac

–Ni/

◦A

v.N

i–O

–Ni/

◦a b

asal

plan

es/◦

a NiO

2pl

anes

/◦N

i···

Ni/

AA

v.N

i–O

Ph/A

Av.

Ni–

Oac

/AJ/

cm−1

[Ni 2

L1(a

cac)

(MeO

H)]·H

2O

,1·H

2O

96.2

/97.

096

.60

75.2

215

5.74

3.08

22.

070

2.05

75.

53[N

i 2L

3(a

cac)

(MeO

H)]·1.

5H2O

,2·1.

5H2O

95.7

/97.

196

.40

75.0

415

5.62

3.09

32.

086

2.06

35.

23[N

i 2H

L1(p

htha

l)(H

2O

)]·M

eOH

,3·M

eOH

98.3

/95.

897

.05

76.9

915

4.21

3.08

92.

042

2.08

14.

88[N

i 2H

L3(p

htha

l)(H

2O

)]·3M

eOH

,4·3M

eOH

97.5

/95.

496

.45

75.2

815

2.30

3.06

82.

041

2.07

46.

60[N

i 2H

L1(m

al)(

H2O

)]·2.

5H2O

,5·2.

5H2O

97.7

/98.

097

.85

75.7

415

4.42

3.09

92.

057

2.05

35.

48[N

i 2L

3(O

Ac)

(H2O

) 2]·3

MeC

N·2H

2O

,799

.5/9

7.3

98.4

078

.72

147.

693.

161

2.07

02.

105

3.60

aO

ph=

O10

3ph

enol

ic,O

ac=

brid

ging

Oac

acor

Oca

rbox

ylat

e.

J value corresponds to the complex with the most acute Ni–Oph–Niangle and the shortest Ni · · · Ni distance (4), whereas the smallestJ parameter is associated with the highest Ni–O–Ni average angleand the longest Ni · · · Ni distance (7).

Conclusions

A synthetic approach to systematically isolate a new seriesof asymmetrically bridged l-phenoxo–l-OR (OR = acac ordicarboxylate) dinuclear nickel complexes, with similar structuralfeatures, is described. The crystal structures of five of the sixisolated complexes were solved and all of them show dinuclearbuilding blocks, expanded in the space through hydrogen bondinteractions. The nickel atoms of the dinuclear blocks are inoctahedral environments, tripled bridged, with a Ni2O2 core inaddition to an NCN endogenous link. These complexes presentsome unusual features, as the rarely observed l2-g2:g1-O,O coordi-nation mode for the carboxylate ligand and the bisdeprotonationof the Schiff base for those carboxylate derivatives. The magneticcharacterisation of the complexes shows that all of them areferromagnetic. Accordingly, this work presents, as far as weknow, the most numerous series of ferromagnetically coupleddinuclear compounds containing a Ni2O2 core, which has beencrystallographically characterised. Molecular orbital treatmentreveals that the superexchange pathway in these complexes isuncommon for compounds with a Ni2O2 centre, as both octahedrashare a base-to-apical edge. To the best of our knowledge, thisstructural feature has not been analysed from the magnetic pointof view until now. Although no structural/magnetic correlationcould be drawn from the multiple studies carried out, it becomesapparent that the variation of the magnitude of J with the Ni–O–Ni angle for Ni2O2 dinuclear complexes is not simple and thata profound analysis of related and new examples is necessaryin order to rationalize the different geometric features whichinfluence the magnetic coupling.

Experimental

Chemicals

All solvents, salicylaldehyde, 5-bromosalicylaldehyde, tri-ethylenetetramine, nickel acetylacetonate, malonic acid and ph-thalic acid are commercially available and were used withoutfurther purification.

Physical measurements

Elemental analyses of C, H and N were performed on a Carlo ErbaEA 1108 analyzer. Infrared spectra were recorded as KBr pelletson a Bio-Rad FTS 135 spectrophotometer in the range 4000–600 cm−1. Electrospray mass spectra were obtained on a Hewlett-Packard LC/MS spectrometer, with methanol as the solvent.These measurements were performed in the in-house services,while magnetic susceptibility measurements for crushed crystallinesamples were carried out at the Servei de Magnetoquimica of theUniversitat de Barcelona with a Quantum Design SQUID MPMS-XL susceptometer, working in the range 2–300 K under magneticfields of 400 G (2–30 K) and 5000 G (30–300 K) and at 2 Kfrom 0 to 50000 G. Diamagnetic corrections were estimated from

4268 | Dalton Trans., 2006, 4260–4270 This journal is © The Royal Society of Chemistry 2006

Pascal Tables. The fit was performed minimizing the function R =∑

(vMT exp − vMT calc)2/∑

(vMT exp)2.

Synthesis of the ligands

H3Lx (x = 1, 3) were synthesized following a method previouslydescribed9,11 and satisfactorily characterized by elemental analy-ses, mass spectrometry and IR and 1H NMR spectroscopies.

Synthesis of the acetylacetonate complexes

[Ni2L1(acac)(MeOH)]·H2O, 1·H2O. H3L1 (0.34 g, 0.75 mmol)was added to a methanol–acetonitrile solution (40 ml) of Ni(acac)2

(0.38 g, 0.15 mmol). The resultant green solution was stirred inair for 6 h and a pale green solid precipitated. The solid wasfiltered off and washed with diethyl ether. Elemental analysis ofthe bulk sample is in agreement with the proposed stoichiometry(0.45 g, 85%), mp = 255–256 ◦C. Found: C 54.20; H 5.43; N 7.91.C33H40N4O7Ni2 requires: C 54.83; H 5.54; N 7.76%. MS (ES): m/z571.2 [Ni2L1]+, 695.2 [Ni2L1(acac)Na]+; IR (KBr, m/cm−1): 1637(C=N); 3348 (OH).

Single crystals of 1·H2O were obtained by slow evaporation ofa diluted methanol solution of the powdery sample.

[Ni2L3(acac)(MeOH)]·1.5H2O, 2·1.5H2O. H3L3 (0.38 g,0.55 mmol) was added to a methanol–acetonitrile solution (40 ml)of Ni(acac)2 (0.28 g, 1.10 mmol) in a 100 ml round bottom flaskequipped with an overhead stirrer and a water-cooled condenser.The mixture was heated to reflux with stirring for 4 h; then,it was immediately filtered and was left to slowly evaporateuntil a green solid precipitated. The solid was filtered off anddried in air to collect the sample. Its elemental analysis is inagreement with the proposed stoichiometry (0.37 g, 70%), mp>300 ◦C. Found: C 40.90; H 3.75; N 5.79. C33H38Br3N4O7.5Ni2

requires: C 40.92; H 3.93; N 5.79%. MS (ES): m/z 808.8 [Ni2L3]+,930.8 [Ni2L3(acac)Na]+; IR (KBr, m/cm−1): 1644 (C=N); 3393(OH).

Single crystals of 2·1.5H2O were obtained by slow evaporationof a diluted methanol solution of the powdery sample.

Synthesis of complexes derived from dicarboxylate ligands

The procedure to synthesize the four complexes is the same and,therefore, we only describe the synthesis of one of the complexesas an example.

[Ni2HL1(o-O2CC6H4CO2)(H2O)]·MeOH, 3·MeOH. To amethanol–acetonitrile (30 ml/30 ml) solution of [Ni2L1-(acac)(MeOH)]·H2O (0.10 g, 0.14 mmol) phthalic acid (0.024 g,0.14 mmol) was added. The mixture was refluxed for 4 h. Theobtained green solution was left to slowly evaporate and singlecrystals of [Ni2HL1(o-O2CC6H4CO2)(H2O)]·MeOH were isolated(0.07 g, 64%), mp >300 ◦C. Found: C 54.08; H 5.05; N 7.27.C36H38N4O9Ni2 requires: C 54.85; H 4.82; N 7.11%. MS (ES):m/z 571.2 [Ni2L1]+, 759.1 [Ni2HL1(O2CC6H4CO2)Na]+. IR (KBr,m/cm−1): 1644 (C=N), 3270, 3400 (OH).

The same complex is obtained if [Ni2L1(acac)(MeOH)]·H2O andphthalic acid are mixed in 2 : 1 molar ratios.

[Ni2HL3(o-O2CC6H5CO2)(H2O)]·3MeOH, 4·3MeOH. Yield:0.06 g, 56%, mp >300 ◦C. Found: C 41.21; H 3.70; N 5.16.C38H43Br3N4O11Ni2 requires: C 41.87; H 3.95; N 5.14%. MS (ES):

m/z 808.8 [Ni2L3]+, 996.7 [Ni2HL3(O2CC6H5CO2)Na]+. IR (KBr,m/cm−1): 1643 (C=N), 3270, 3394 (OH).

[Ni2HL1(O2CCH2CO2)(H2O)]·2.5H2O·0.25MeOH·MeCN, 5·2.5H2O·0.25MeOH·MeCN. Single crystals of [Ni2HL1(O2-CCH2COO)(H2O)]·2.5H2O·MeCN·0.25MeOH, 5·2.5H2O·0.25MeOH·MeCN, suitable for X-ray diffraction studies yielded[Ni2HL1(O2CCH2CO2)(H2O)]·2.5H2O on drying (0.17 g, 48%),indicating that the complex loses the most volatile solventupon standing, mp >300 ◦C. Found: C 47.38; H 5.39; N 7.52.C30H37N4O10.5Ni2 requires: C 48.75; H 5.01; N 7.58%. MS (ES).m/z 571.2 [Ni2L1]+, 697.0 [Ni2HL1(O2CCH2CO2)Na]+. IR (KBr,m/cm−1): 1644 (C=N), 3278, 3390 (OH).

[Ni2HL3(O2CCH2CO2)(H2O)]·3H2O, 6·3H2O. Yield: 0.08 g,38%, mp >300 ◦C. Found: C 36.58; H 3.52; N 5.46.C30H35Br3N4O11Ni2 requires: C 36.58; H 3.55; N 5.69%. MS (ES):m/z 808.8 [Ni2L3]+, 936.8 [Ni2HL3(O2CCH2CO2)Na]+. IR (KBr,m/cm−1): 1645 (C=N), 3274, 3409 (OH).

Crystallographic measurements

Crystal data and details of refinement are given in Table 1.Single crystals of 1·H2O to 5·2.5H2O·MeCN·0.25MeOH, suit-able for single-crystal X-ray studies, were obtained as detailedabove. Diffraction data were collected at 120 K for 1·H2Oto 4·3MeOH and at 293 K for 5·2.5H2O·0.25MeOH·MeCN,using either a Bruker SMART CCD-1000 (1·H2O, 3·MeOHto 5·2.5H2O·0.25MeOH·MeCN) or a Bruker-Nonius KappaCCD2000 (2·1.5H2O) diffractometer, employing graphite-monochromated Mo-Ka radiation (k = 0.71073 A) in all thecases. No significant decays were observed, and data were cor-rected for Lorentz and polarisation effects. Multi-scan absorptioncorrections were applied using SADABS.35 The structures weresolved by standard direct methods, employing SIR-92,36 andrefined by Fourier techniques based on F 2, using SHELXL-97.37 Non-hydrogen atoms were anisotropically refined, exceptthose corresponding to some disordered atoms of 2·1.5H2O(occupation sites of ca. 0.65 and 0.35) or solvated molecules withpartial occupation sites of 5·2.5H2O·0.25MeOH·MeCN, whichwere isotropically treated. Unfortunately, 4·3MeOH forms verythin needle-shaped crystals, which only allowed rather weakdiffraction data sets, so consequently, some geometric data are notvery accurate, but of enough quality to be discussed. Hydrogenatoms of organic groups were added theoretically and treatedas riding on concerned atoms. Methyl groups of acac ligandsfor 1·H2O and 2·1.5H2O appear disordered on two differentpositions with occupation sites of ca. 0.13 and 0.87. ThoseH atoms bonded to water molecules (1·H2O, 3·MeOH and5·2.5H2O·0.25MeOH·MeCN) and to terminal phenol groups ofthe ligands (3·MeOH to 5·2.5H2O·0.25MeOH·MeCN), could belocated on Fourier maps and treated either isotropically (1·H2O,3·MeOH and 5·2.5H2O·0.25MeOH·MeCN) or their coordinatesand thermal parameters fixed (2·1.5H2O and 4·3MeOH).

CCDC reference numbers 291778 and 606600–606603.For crystallographic data in CIF or other electronic format see

DOI: 10.1039/b606414k

This journal is © The Royal Society of Chemistry 2006 Dalton Trans., 2006, 4260–4270 | 4269

Acknowledgements

The authors thank Xunta de Galicia (PGIDT03XIB20901PR) forfinancial support.

References

1 L. Merz and W. Haase, J. Chem. Soc., Dalton Trans., 1980, 875.2 M. Handa, N. Koga and S. Kida, Bull. Chem. Soc. Jpn., 1988, 61, 3853.3 (a) V. H. Crawford, H. W. Richardson, J. R. Wasson, D. J. Hodgson

and W. E. Hatfield, Inorg. Chem., 1976, 15, 2107; (b) D. J. Hodgson,Prog. Inorg. Chem., 1975, 19, 173.

4 L. K. Thompson, S. K. Mandal, S. S. Tandon, J. N. Brisdon and M. K.Park, Inorg. Chem., 1996, 35, 3117.

5 (a) E. Ruiz, P. Alemany, S. Alvarez and J. Cano, J. Am. Chem. Soc.,1997, 119, 1297; (b) E. Ruiz, P. Alemany, S. Alvarez and J. Cano, Inorg.Chem., 1997, 36, 3683.

6 H. Hu, D. Zhang, Z. Chen and C. Liu, Chem. Phys. Lett., 2000, 329,255.

7 K. K. Nanda, L. K. Thompson, J. N. Bridson and K. Nag, J. Chem.Soc., Chem. Commun., 1994, 1337.

8 X.-H. Bu, M. Du, L. Zhang, D.-Z. Liao, J.-K. Tang, R.-H. Zhang andM. Shionoya, J. Chem. Soc., Dalton Trans., 2001, 593.

9 (a) M. Fondo, A. M. Garcıa-Deibe, J. Sanmartın, M. R. Bermejo, L.Lezama and T. Rojo, Eur. J. Inorg. Chem., 2003, 3703; (b) M. Fondo,A. M. Garcıa-Deibe, M. Corbella, J. Ribas, A. Llamas-Saiz, M. R.Bermejo and J. Sanmartın, Dalton Trans., 2004, 3503.

10 M. Fondo, A. M. Garcıa-Deibe, M. Corbella, E. Ruiz, J. Tercero, J.Sanmartın and M. R. Bermejo, Inorg. Chem., 2005, 44, 5011.

11 M. Fondo, N. Ocampo, A. M. Garcıa-Deibe, M. Corbella, M. R.Bermejo and J. Sanmartın, Dalton Trans., 2005, 3785.

12 See, for example: (a) P. K. Nanda, G. Aromı and D. Ray, Inorg. Chem.,2006, 3143; (b) R. C. Howell, K. V. N. Spence, I. A. Kahwa and D. J.Williams, J. Chem. Soc., Dalton Trans., 1998, 2727; (c) L.-W. Yang,S. Liu, E. Wong, S. J. Rettig and C. Orvig, Inorg. Chem., 1995, 34,2164; (d) B. Chiari, O. Piovesana, T. Tarantelli and P. F. Zanazzi, Inorg.Chem., 1982, 21, 2444.

13 See, for example: (a) M. Zimmer, G. Schulte, X.-L. Luo and R. H.Crabtree, Angew. Chem., Int. Ed. Engl., 1991, 30, 193; (b) H. Ohta, K.Harada, K. Irie, S. Kashino, T. Kambe, G. Sakane, T. Shibahara, S.Takamizawa, W. Mori, M. Nonoyama, M. Hirotsu and M. Kojima,Chem. Lett., 2001, 842; (c) N. Okabe and Y. Muranishi, ActaCrystallogr., Sect. C, 2002, 58, m475; (d) X. Cui, M. J. Calhorda, P. J.Costa, R. Delgado, M. G. B. Drew and V. Felix, Helv. Chim. Acta, 2004,87, 2613; (e) Z.-L. You, H.-L. Zhu and W.-S. Liu, Acta Crystallogr.,

Sect. E, 2004, 60, m805; (f) H. Adams, D. E. Fenton and P. E. McHugh,Inorg. Chim. Acta, 2004, 357, 3641; (g) J. B. Fontecha, S. Goetz and V.McKee, Dalton Trans., 2005, 923.

14 B. Prelesnik, R. Herak, D. Stojakovic and D. Poleti, Monatsh. Chem.,1986, 117, 47.

15 O. Kahn, Molecular Magnetism, VCH Publishers, New York, 1993.16 C. Wang, K. Fink and V. Staemmler, Chem. Phys., 1995, 192, 25.17 R. J. Butcher, C. J. O’Connor and E. Sinn, Inorg. Chem., 1981, 20, 3486.18 D. Black, A. J. Blake, K. P. Dancey, A. Harrison, M. McPartlin, S.

Parsons, P. A. Tasker, G. Whittaker and M. Schroder, J. Chem. Soc.,Dalton Trans., 1998, 3953.

19 N. Hoshino, T. Ito, M. Nihei and H. Oshio, Chem. Lett., 2002, 844.20 K. K. Nanda, R. Das, L. K. Thompson, K. Venkatsubramanian, P.

Paul and K. Nag, Inorg. Chem., 1994, 33, 1188.21 K. K. Nanda, R. Das, K. Venkatsubramanian, P. Paul and K. Nag,

J. Chem. Soc., Dalton Trans., 1993, 2515.22 H. Adams, D. E. Fenton, S. R. Haque, S. L. Heath, M. Ohba, H. Okawa

and S. E. Spey, J. Chem. Soc., Dalton Trans., 2000, 1849.23 E. Berti, A. Caneschi, C. Daiguebonne, P. Dapporto, M. Formica, V.

Fusi, L. Giorgi, A. Guerri, M. Micheloni, P. Paoli, R. Pontellini and P.Rossi, Inorg. Chem., 2003, 42, 348.

24 M. Mimura, T. Matsuo, N. Matsumoto, S. Takamizawa, W. Mori andN. Re, Bull. Chem. Soc. Jpn., 1998, 71, 1831.

25 S. Koizumi, M. Nihci and H. Oshio, Chem. Lett., 2003, 32, 812.26 R. Das and K. Nag, Inorg. Chem., 1991, 30, 2831.27 A. J. Downard, V. McKee and S. S. Tandon, Inorg. Chim. Acta, 1990,

173, 181.28 K. K. Nanda, R. Das, M. J. Newlands, R. Hynes, E. J. Gabe and K.

Nag, J. Chem. Soc., Dalton Trans., 1992, 897.29 K. K. Nanda, R. Das, L. K. Thompson, K. Venkatsubramanian and

K. Nag, Inorg. Chem., 1994, 33, 5934.30 S. Mukherjee, T. Weyhermuller, E. Bothe, K. Wieghardt and P.

Chaudhuri, Eur. J. Inorg. Chem., 2003, 863.31 R. J. Butcher, C. J. O’Connor and E. Sinn, Inorg. Chem., 1982, 21, 616.32 M. Fondo, N. Ocampo, A. M. Garcıa-Deibe, R. Vicente, M. Corbella,

M. R. Bermejo and J. Sanmartın, Inorg. Chem., 2006, 45, 255.33 D. Volkmer, B. Hommerich, K. Griesar, W. Haase and B. Krebs, Inorg.

Chem., 1996, 35, 3792.34 C. Mealli and D. M. Proserpio, J. Chem. Educ., 1990, 67, 399.35 SADABS, Area-Detector Absorption Correction, Siemens Industrial

Automation Inc., Madison, WI, 1996.36 SIR92 A program for crystal structure solution: A. Altomare, G.

Cascarano, C. Giacovazzo and A. Guagliardi, J. Appl. Crystallogr.,1993, 26, 343.

37 G. M. Sheldrick, SHELX97 Programs for Crystal Structure Analysis,Institut fur Anorganische Chemie der Universitat, Gottingen, Ger-many, 1998.

4270 | Dalton Trans., 2006, 4260–4270 This journal is © The Royal Society of Chemistry 2006