www.elsevier.com/locate/ica

Inorganica Chimica Acta 320 (2001) 12–21

A new dinuclear unsymmetric copper(II) complex as model for theactive site of catechol oxidase

Christiane Fernandes a, Ademir Neves a,*, Adailton J. Bortoluzzi a,Antonio S. Mangrich b, Eva Rentschler c, Bruno Szpoganicz a, Erineu Schwingel d

a Laboratorio de Bioinorganica e Cristalografia, Departamento de Quımica, LABINC CAMPUS-TRINIDADE,Uni�ersidade Federal de Santa Catarina, 88040-900 Florianopolis SC, Brazil

b Departamento de Quımica, Uni�ersidade Federal do Parana, 81531-970 Curitiba PR, Brazilc Max-Planck Institut fur Strahlenchemie Stiftstrasse 34-36, D-45470 Mulheim an der Ruhr, Germany

d Centro de Ciencias Exatas, Agrarias e das Engenharias, Uni�ersidade do Sul de Santa Catarina, UNISUL, 88130-000 Florianopolis SC, Brazil

Received 15 December 2000; accepted 14 March 2001

Abstract

The crystal structure, magnetic, redox and spectroscopic properties of a novel unsymmetrical dinuclear copper(II) complex,prepared by the reaction between copper(II) perchlorate, sodium acetate and the unsymmetrical, binucleating ligand HTPPNOL,where HTPPNOL is N,N,N �-tris-(2-pyridylmethyl)-1,3-diaminopropan-2-ol, is reported. HTPPNOL (1 equiv.) reacted with 1equiv. of copper(II) ion, in methanol, and produced the mononuclear copper complex [Cu(TPPNOL)](ClO4)(BPh4) (1). On theother hand, the reaction of 1 equiv. of HTPPNOL with 2 equiv. each of copper (II) ion and acetate, in methanol, produced thedinuclear complex [Cu2(TPPNOL)(OOCCH3)](ClO4)2 (2), whose structure has been determined by X-ray diffraction. In complex2, as a result of the inherent asymmetry of the ligand HTPPNOL, one copper ion is five-coordinated (distorted trigonal-bipyra-midal) while the other copper is four-coordinated (distorted square-planar). Then, as a result of the presence of distinct geometriesfor the metal centres, complex 2 exhibits a ferromagnetic coupling (J= +25.41 cm−1). Titration experiments carried out on thedinuclear complex suggest a pKa=8.0, which was related to the aquo/hydroxo equilibrium. Complex 2 is able to oxidise3,5-di-tert-butylcatechol to the respective o-quinone. The oxidation reaction was studied by following the appearance of thequinone spectrophotometrically, at pH 8.0 and 25 °C. © 2001 Elsevier Science B.V. All rights reserved.

Keywords: Crystal structures; Copper complexes; Unsymmetrical ligand complexes; Catecholase activity

1. Introduction

The importance of copper ions as catalysts in bothenzymic [1] and nonenzymic [2] chemical systems is wellknown. Many of the copper proteins activate molecularoxygen, and the functions of these proteins are varied;they serve as oxygen carries (hemocyanin) [3,4] or ascatalysts in oxygenation (oxygenases) or oxidation (oxi-dases) reactions [5,6].

Hemocyanin functions as an oxygen transporterprotein in arthropods and molluscs. X-ray diffraction

studies of the deoxy form have revealed that the twononinteracting Cu(I) ions are each coordinated by threehistidine N atoms. The Cu···Cu distance has been deter-mined to be 2.9–3.8 A� for the enzyme isolated from thespiny lobster Panuliris interruptus [7] and to be 4.6 A�for the enzyme isolated from the horseshoe crab Limu-lus polyphemus [8].

Tyrosinase catalyses the aerial oxidation ofmonoplenols to o-diphenols (cresolase activity) and theoxidation of o-diphenols (catechol) to o-quinones (cate-cholase activity). No X-ray crystal structure is availableuntil now, but spectroscopic investigations [9] and EX-AFS measurements [10] seem to reveal that the activesite is similar to that of hemocyanin.

* Corresponding author. Tel.: +55-2148-331-9466; fax: +55-2148-331-9711.

E-mail address: ademir@qmc.ufsc.br (A. Neves).

0020-1693/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.

PII: S 0 0 2 0 -1693 (01 )00470 -4

C. Fernandes et al. / Inorganica Chimica Acta 320 (2001) 12–21 13

Catechol oxidase is another copper enzyme, isolatedfrom phytogenic materials, that only catalyses the oxi-dation of catechol to quinone without acting on ty-rosi-nase. This reaction is of great importance in medicaldiagnosis for the determination of the hormonally ac-tive catecholamines adrenaline, noradrenaline anddopa [11]. The copper in the isolated catechol oxidaseswas found to be EPR-silent and was assigned to anantiferromagnetically spin-coupled Cu(II)–Cu(II) pair[12]. XAS investigations on the native met forms fromLycopus europaeus and Ipomoea batatas have revealedthat the active site consists of a dinuclear copper(II)centre, where the metals are coordinated by four N/Odonor ligands [13]. The short metal–metal distance of2.9 A� and the results of EPR investigations indicate a�-hydroxo bridged dicopper(II) active site in the metform of the protein [14]. Recently, Kablunde et al.reported the crystal structure of this enzyme fromsweet potato, for oxidised and reduced forms. In bothforms, each of the two copper atoms is coordinated bythree histidine residues. In the oxidised catechol ox-idase structure, two cupric ions are 2.9 A� apart, with ahydroxide ion as the fourth ligand, attributed to abridging solvent molecule. Upon reduction of the en-zyme the metal–metal separation increases to 4.4 A� ,while the histidine moves only slightly and no signifi-cant conformational change is observed for the otherresidues of the protein [15]. Notable advances in theunderstanding of the structural and chemical proper-ties of copper metalloenzymes have been achieved bymodel studies of synthetic analogues, where complexesderivated from dinucleating ligands providing symmet-rically arranged, three-coordinate copper centres withnitrogen donors have been used, because this arrange-ment simulates the coordination environment in hemo-cyanin [16], tyrosinase [17] and catechol oxidase [18].Furthermore, it is known that coordination numberasymmetry in metalloenzymes presents the possibilityof ‘open’ coordination sites for direct interaction ofthe metal centre with substrate; for this reason someof the biomimetic models reported recently use non-symmetric dinucleating ligands to produce unsymmet-ric dicopper complexes [19,20].

In fact, we are interested in the synthesis of dinu-clear copper(II) complexes as a structural and func-tional model for catechol oxidase using theunsymmetrical ligand HTPPNOL, which should beprovide one or more open coordination sites [21]. Inparticular, we report here comparative studies on thecatechol oxidase activity exhibited by the mononuclearcomplex [Cu(HTPPNOL)](BPh4)(ClO4) (1) and by thedinuclear complex [Cu2(TPPNOL)(OAc)](ClO4)2 (2), inan attempt to model the activity of the copper con-taining enzyme catechol oxidase.

2. Experimental

2.1. General

All elemental analyses were performed on a Perkin–Elmer 2400 analyser. The IR spectra were recordedwith a Perkin–Elmer Model 781 spectrometer (KBrdisk or in film) and the electronic absorption spectra(200–1000 nm range) were recorded in CH3CN with aPerkin–Elmer L19 spectrophotometer. Electrochemicalexperiments were performed with a Princeton AppliedResearch (PARC) model 273 potentiostat/galvanostat,equipped with an IBM/AT-386 microcomputer. Cyclicvoltammograms (CVs) were acquired in acetonitrilecontaining 0.1 M [TBA][(PF)6] (tetrabutylammoniumhexafluorphosphate) as supporting electrolyte, at 150mV s−1, under argon atmosphere at room temperature(r.t.). The electrochemical cell employed was of a stan-dard three-electrode configuration: a platinum workingelectrode, a platinum-wire auxiliary electrode, and anSCE reference electrode constructed in our laboratory.The performance of the reference electrode was moni-tored by measuring the Fc+/Fc couple of ferrocene.Magnetic susceptibility measurements were performedon a polycrystalline sample of 2 in the temperaturerange of 2–296.2 K in an applied field of 1 T with aQuantum Design MPMS SQUID magnetometer. Cor-rections for diamagnetic contributions were applied byusing Pascal’s constants [22]. The sample holder dia-magnetism was measured and subtracted from the rawdata. The potentiometric studies were performed with acorning pH/ion analyser model 350 fitted with blue-glass and calomel reference electrodes, designated asp[H]. Double distilled water in presence of KMnO4 wasused to prepare the water solutions. The electrode wascalibrated using the data obtained from a potentiomet-ric titration of known volume of a standard 0.100 MHCl with a standard 0.100 M KOH. The ionic strengthof the HCl solution was maintained at 0.100 M by theaddition of KCl. The temperature was 25.00�0.05 °Cand the experimental solutions, adjusted to 0.100 M inionic strength by addition of KCl, were titrated with0.100 M standard CO2-free KOH. Equilibrium mea-surements were made on solutions containing 0.05mmol of the dinuclear copper(II) complex. The experi-ments were carried out on 50.00 ml of experimentalsolution in a thermostated cell, purged with argoncleaned by two 0.100 M KOH solutions. The additionof 0.02 ml of CO2-free KOH was carried out with anSchott Gerade automatic burette, with repeating addi-tions of base once a constant pH value had beenreached. Computations were all carried out with theBEST program and species diagrams were obtained fromthe SPE and SPEPLOT programs [23]. Kinetic experi-ments were carried out spectrophotometrically on aDiode Array spectrophotometer 8452 A by measuring

C. Fernandes et al. / Inorganica Chimica Acta 320 (2001) 12–2114

the quinone formation using the 400 nm peak, at pH8.0 and 25 °C. The auto-oxidation of the 3,5-dtbcsubstrate was measured and subtracted from the totalin order to obtain the extent of the oxidation reactioncatalysed by the copper(II) complex.

2.2. Syntheses

2.2.1. Ligand HTTPPNOL:(N,N,N �-tris-(2-pyridylmethyl)-1,3-diaminpropan-2-ol)

The ligand was prepared by a condensation reactionbetween N,N-bis-(2-pyridylmethyl)-1,3-diaminpropan-2-ol and 2-pyridinecarboxaldehyde, which has been re-ported previously [21].

2.2.2. [Cu(TPPNOL)](ClO4)(BPh4) (1)Caution: the complex isolated as the perchlorate salt

should be handled as a potentially explosive compound.The mononuclear complex was synthesised by the

reaction between [Cu(OH2)6](ClO4)2 and the ligandHTPPNOL, in methanol, with the addition of NaBPh4

[24].

2.2.3. [Cu2(TPPNOL)(OAc)](ClO4)2 (2)A methanolic solution (20 ml) of HTPPNOL (190

mg, 0.5 mmol) and NaOAc·3H2O (140 mg, 1 mmol)was added to a methanolic solution of [Cu(OH2)6]-(ClO4)2 (370 mg, 1 mmol), resulting in a deep bluesolution. The reaction mixture was allowed to stand at

10°C for 1 week. Dark-blue single crystals were formedand were filtered off. Yield: 83%. Anal. Found: C,36.69; H, 3.65; N, 9.34. Calc. for Cu2C23H27O11N5Cl2:C, 36.95; H, 3.64; N, 9.37%.

2.3. X-ray crystallography

Geometric parameters and intensity data for complex2 were collected on a CAD4 diffractometer, at r.t., withgraphite-monochromated Mo K� radiation (�=0.71073 A� ). The collected reflections were corrected forLorentz and polarisation effects. The structures weresolved by direct methods with SHELXS-97 [25] andrefined by full-matrix least-squares procedure based onF2 with SHELXL-97 [26]. A drawing of the molecularstructure presented in this article was carried out withthe ZORTEP program [27]. Further relevant crystallo-graphic information for complex 2 are listed in Table 1.

A dark-blue irregular shaped crystal of 2, with di-mensions 0.50×0.33×0.10 mm, was isolated andmounted at the goniometer head for X-ray measure-ments. The unit cell parameters were defined on thesetting angles of 25 centred reflections in the range9.82���14.00°. The intensities of 6730 reflectionswere collected from �/2� scans within the range 2.24���27.00°. Since the ratio of reflections/parameters isgreater than 10 and the crystals of 2 showed a lowscattering capacity, the reflections with ��25° havebeen omitted in the structure refinement. Of the 5539reflections that are associated to a � angle up to 25 °,5359 reflections were unique (Rint=0.0523) and 2551reflections were considered as observed with I�2�(I).Intensity control was carried out using three standardsreflections, which were measured at regular intervalsand no significant loss of intensity was observed duringthe data collection. An empirical absorption correctionwas applied on the basis of azimuthal scans of sevenappropriate reflections (Tmin=0.494; Tmax=0.880).Hydrogen atoms were placed in calculated positionsand included in the structure factor calculation using ariding model. The perchlorate groups were found to bedisordered. For the first perchlorate (C11) three oxygenatoms were found to be disordered over two positionsabout the threefold axis whereas for the second per-chlorate ion (C12) two alternative positions were foundfor all oxygen atoms. The site occupancies for thedisordered atoms were fixed at 0.55 and 0.45 in the firstand 0.6 and 0.4 in the second perchlorate groups. Therefinement of 411 parameters, positional and an-isotropic thermal parameters for all non-H atoms, ex-cept for the disordered oxygen atoms of the secondperchlorate, which were refined isotropically, convergedat R=0.0722 for observed reflections and Rw=0.2016for all unique reflections. The highest peak found in thefinal difference map was 1.27 e A� −3 and it was locatednear an oxygen atom (O24�) of the second perchlorategroup.

Table 1Selected crystallographic data for [Cu2(TPPNOL)(OOCCH3)](ClO4)2

(2)

Empirical formula C23H27Cl2Cu2N5O11

747.48Formula weight (g mol−1)Crystal system monoclinic

C2/cSpace groupUnit cell dimensions

26.357(5)a (A� )b (A� ) 12.764(3)c (A� ) 18.848(4)� (°) 90� (°) 106.84(3)� (°) 90

V (A� 3) 6069(2)Z 8Dcalc (g cm−3) 1.636�(Mo) (mm−1) 1.642F(000) 3040Reflections collected 6730Reflections observed 2551Independent reflections 5359 [Rint=0.0523]

411ParametersFinal R indices [I�2�(I)] a R=0.0722R indices (all data) b Rw=0.2016

1.015Goodness-of-fit on F2

a R1=��Fo�−�Fc�/��Fo�.b wR2={�[w(Fo

2−Fc2)2]/�[w(Fo

2)2]}1/2;w=[�2(Fo2)+(aP)2+bP ]−1,

where P= (Fo2+2Fc

2)/3.

C. Fernandes et al. / Inorganica Chimica Acta 320 (2001) 12–21 15

Fig. 1. View of the binuclear cation 2 with labelling scheme. Theellipsoids are shown at the 40% probability level. Hydrogen atomshave been omitted for clarity.

analysis, spectroscopic and electrochemical analysis.The dinuclear complex 2 was also characterised bysingle-crystal X-ray crystallography and magnetic mea-surement. The ligand HTPPNOL shows a hydroxyfunction, which serves as a bridging alkoxo group in adinuclear complex; one half of the ligand providesnitrogen groups from one tertiary amine and two pyri-dine nitrogens, while the other half provides one nitro-gen from tertiary amine and one pyridine nitrogen. Theligand HTPPNOL results in a dinuclear complex withan unsymmetric coordination environment similar tothat found for the active site in catechol oxidase, andrepresents the possibility of ‘open’ coordination site fordirect interaction of one metal centre with the substrate,as 3,5-di-tert-butylcatechol. The monunuclear complex[Cu(TPPNOL)](ClO4)(BPh4) (1) was obtained by thereaction of 1 equiv. of HTPPNOL and 1 equiv. of[Cu(OH2)6](ClO4)2, and 2 equiv. of Na(BPh4), resultingin a blue solution. The IR spectrum of 1 shows(BPh4

−) and (ClO4−) stretches at 736 and 1094

cm−1, respectively. The dinuclear complex[Cu2(TPPNOL)(OAc)](ClO4)2 (2) shows sym and ass

stretching at 1440 and 1560 cm−1, attributed to theacetate bridge and a band at 1086 cm−1, attributed tothe (ClO4

−) counter-ion.

3.2. Crystal structure [Cu2(TPPNOL)(OAc)](ClO4)2 (2)

The X-ray crystal structure of (2) reveals that eachcrystallographic asymmetric unit consists of a dinuclearcopper cation and two perchlorate ions. Both copperatoms are linked by the ligand-derived alkoxide and byan exogenous acetate ion. A perspective view of thecation is displayed in Fig. 1 and the selected bondlengths and angles are presented in Table 2. The asym-metry of the ligand leads to two distinct copper centres:Cu1 is five-coordinated (distorted trigonal-bipyrami-dal), which is coordinated to two pyridine nitrogensand a tertiary amine nitrogen and also to an oxygendonor from the alkoxo and acetate groups, while theCu2 (distorted square-planar) is four-coordinated, withtwo cis oxygen atoms from alkoxo and acetate groups,and two cis nitrogen donor atoms from pyridine andamine groups. The distinct arrangement for Cu1 andCu2 becomes clear by analysing the angles listed inTable 2. For the Cu2 centre, with distorted trigonal-bipyramidal geometry, the tertiary amine N1 and theacetate oxygen O2 occupy axial positions(N1�Cu1�O2=178 °), with bond lengths of 2.034(6)and 1.934(6) A� , respectively. The pyridine nitrogenatoms N11, N31 and alkoxo O1 donors occupy theequatorial positions, the respective bond lengths being2.120(8), 2.042(7) and 1.888(6) A� . The distorted trigo-nal-bipyramid geometry around Cu2 can be confirmedfrom the structural index parameter =0.73 (thisparameter is 1 for a perfect trigonal-bipyramidal ge-

Table 2Selected bond lengths (A� ) and bond angles (°) for [Cu2(TPPNOL)-(OOCCH3)](ClO4)2 (2)

Cu1�O1 1.888(6)1.934(6)Cu1�O2

Cu1�N1 2.034(6)Cu1�N31 2.042(7)Cu1�N11 2.120(8)Cu2�O1 1.852(6)Cu2�N21 1.937(7)Cu2�O3 1.945(5)Cu2�N2 1.987(7)

97.4(2)O1�Cu1�O2O1�Cu1�N1 84.5(3)O2�Cu1�N1 178.1(3)

134.1(3)O1�Cu1�N31O2�Cu1�N31 96.8(3)N1�Cu1�N31 82.1(3)

109.7(3)O1�Cu1�N1197.3(3)O2�Cu1�N11

N1�Cu1�N11 81.7(3)N31�Cu1�N11 111.4(3)

165.6(3)O1�Cu2�N21O1�Cu2�O3 98.0(2)

95.3(3)N21�Cu2�O3O1�Cu2�N2 83.6(3)N21�Cu2�N2 83.8(3)O3�Cu2�N2 173.5(3)

128.9(3)Cu2�O1�Cu1

3. Results and discussion

3.1. Syntheses

The unsymmetrical ligand HTPPNOL reacts inmethanol with copper(II) perchlorate hexahydrated toform green or blue compounds identified as mononu-clear and dinuclear complexes, which can be isolated inhigh yields as perchlorate or tetraphenylborate salts.Blue or green solutions were formed rapidly, resultingin compounds which were characterised by elemental

C. Fernandes et al. / Inorganica Chimica Acta 320 (2001) 12–2116

Fig. 2. Schematic representation of the orbital orientations for com-plex 2.

compounds that these differences in bond length for thetwo metals to the oxygen atoms of the syn–syn biden-tate carboxylate bridge are usually not larger than 0.10A� [31].

3.3. Magnetic susceptibility

For the dinuclear complex 2, the crystal structurereveals that Cu2 is square-planar but Cu1 is trigonal-bipyramidal. For square-planar Cu2, the magnetic or-bital is dx 2−y 2 with lobes directed towards O1 and O3.For trigonal-bipyramidal Cu1, however, the magneticorbital is dz 2, with unpaired spin density along theN1�Cu1�O2 axis. In addition, the different geometriesmean that the magnetic orbitals will have differentenergies. Thus, the alkoxide can interact with the un-paired electron in the dx 2−y 2 orbital in Cu2, but it doesnot overlap with the orbital having the unpaired elec-tron on Cu1 (namely dz 2), which is oriented towards theacetate group, as shown in Fig. 2. Similar magneticresults were observed for the binuclear copper complex[Cu2(OH)(H2O)(O2CCH3)(bpy)2](ClO4)2, where bpy is2,2�-bipyridine, synthesised by Christou and co-workers[32].

For complex 2, variable-temperature susceptibilitymeasurements in the range of 2–290 K show the exis-tence of ferromagnetic coupling. At room temperaturethe magnetic moment was found close to that expectedfor two isolated d9 copper sites. On decreasing tempera-ture the effective magnetic moment increases from 2.63�B per molecule at 290 K to 2.83 �B at 20 K (Fig. 3).Using the Hamiltonian

H= −2JS1·S2

ometry) [28]. Cu1 is four-coordinated square-planar.The Cu�O distances are 1.945(5) and 1.852(6) for theacetate and alkoxide, and the Cu�N distances for thepyridine nitrogen and the amine nitrogen are 1.937(7)and 1.987(7) A� , respectively. These angles and bondlengths are similar to those found in[Cu2L�Et(OAc)](ClO4) [29] and [Cu2(BMDP)-(OAc)](ClO4)2·2H2O [19], where L is the symmetricanalogue of the unsymmetric ligand HBMDP (N,N,N �-tris((N-methyl-2-benzimidazolyl)methyl)-N �-methyl-1,3-diamino-2-propanol). The Cu1�Cu2 distance of 3.363 A�and Cu1�O1�Cu2 bridging angle of 128.9(8)° fall in therange found for similarly coordinated �-alkoxo cop-per(II) dimers [30]. The acetate group is syn–syn biden-tate bridging, which is a very common bridging modeon dinuclear �-carboxylate transition metal complexes.The Cu1�O2 distance (1.934(6) A� ) is very similar to theCu2�O3 distance (1.945(5) A� ). It is observed for similar

Fig. 3. A plot of �M vs. T and �eff vs. T for complex 2. Solid lines show the best fit obtained.

C. Fernandes et al. / Inorganica Chimica Acta 320 (2001) 12–21 17

Fig. 4. EPR spectrum for the compounds under study, at 77K: (a)[Cu2(TPPNOL)(OAc)2](ClO4)2 (2) and (b) [Cu(HTPPNOL)]-(BPh4)(ClO4) 1. (I) Frozen acetonitrile solution spectrum for 1; (II)simulated spectrum of 1 using the WINEPR SIMFONIA program.

trigonal-bipyramidal complexes have a low-energy ab-sorption band with a high-energy shoulder in the visibleregion [33]. As the mononuclear complex shows a bandat 758 nm and a well defined shoulder at 843 nm, wepropose that the copper ion in 1 is coordinated to fivenitrogen donor atoms in a distorted square-pyramidalgeometry, and the hydroxo oxygen of the ligand doesnot participate in the metal coordination.

The dinuclear complex 2 too shows two well definedbands at 667 (�=210 M−1 cm−1) and 889 nm (�=150M−1 cm−1) in CH3CN solution and most probably is areflection of the two different geometries around thecopper centres. Interestingly, the UV–Vis spectrum forthe closely related [Cu(BMDP)(OAc)](ClO4)2·2H2Ocomplex, reported by Droege and co-workers [19], inacetonitrile, displays only a single band at 692 nm(�=103 M−1 cm−1), and in the symmetric tetrakis-(benzimidazole) analogue [Cu2�L�Et(OAc)](ClO4) alsoa single band is observed at 776 nm (�=61 M−1 cm−1)[34].

The X-band EPR spectrum recorded in a frozenacetonitrile solution of 1 shows four lines, typical ofCu(II) ions (I=3/2), as a result of the hyperfine inter-action between the unpaired electron and the metalcentre (Fig. 4(b)). The values for A�=180×10−4 G,g�=2.1885, A�=25×10−4 G and g�=2.0500 wereobtained from the simulated spectrum. As g��g�, asquare-pyramidal geometry can be proposed for 1, inagreement with the geometry proposed from the elec-tronic spectroscopy results. For complex 2, the spec-trum can be interpreted as a triplet state, originatingfrom exchange-coupled pairs of copper(II) ions. Fromthis spectrum the �ms= �1 (g1=2.098; g2=2.049)and the selection-rule forbidden �ms= �2 (g=4.3)transitions are clearly recognised (see Fig. 4(a)).

3.5. Electrochemical

The electrochemical behaviour of the compounds hasbeen studied using cyclic voltammetry. Cyclic voltam-mograms were recorded in acetonitrile with[TBA][(PF)6] as the supporting electrolyte in the poten-tial range −2.0 to 1.2V versus SCE (Fig. 5(a)). Forcomplex 1 a reversible redox process occurs at −0.19V versus NHE, at 150 mV s−1. Repetitive CV scans of1 show that the redox couple is maintained with nosignificant decrease in either cathodic or anodic waves.This process was attributed to a Cu(II)/Cu(I) redoxcouple. The process at 1.3 V versus NHE was at-tributed to the BPh4

− counter-ion. Cyclic voltammetryof 2 (Fig. 5(b)) shows two redox processes at −0.38and −0.56 V versus NHE, and these irreversible pro-cesses were tentatively attributed to 2Cu2+ +1e−�Cu2+Cu1+ and Cu2+Cu1+ +1e−�2Cu1+.

the data could be simulated with g values of 2.08 and acoupling constant of J= +25.41 cm−1. On furtherdecreasing the temperature �eff decreases to reach 2.46�B at 2 K. This behaviour was be taken into accountwith an additional theta-Weiss constant, = −0.92 K,for antiferromagnetic intermolecular interactions.

3.4. Electronic and EPR spectroscopies

The electronic spectrum of 1, in acetonitrile, displaysa band at 758 nm (�=241 M−1 cm−1) and a shoulderat 843 nm (�=211 M−1 cm−1), attributed to d–dtransitions. The electronic absorption spectra for five-coordinated copper(II) complexes normally fall intotwo general categories. Square-pyramidal complexestypically show a high-energy absorption band in thevisible region with a low-energy shoulder. In contrast,

C. Fernandes et al. / Inorganica Chimica Acta 320 (2001) 12–2118

Fig. 5. Cyclic voltammograms (scan rate 150 mV s−1) of 1×10−3 Msolutions of (a) complex 1 and (b) complex 2, in CH3CN (0.1 M[TBA][(PF)6], at a platinum working electrode for 2 and vitreouscarbon for 1, with the internal standard ferrocene).

3.6. Titration experiments

Complex 2 was investigated in aqueous solution, bypotentiometric titration, to determine the presence ofwater molecules coordinated to the metal centres. Asindicated in Fig. 1, in solid state, this compound didnot show any coordinated solvent molecule. However,in aqueous solution it was possible to determine onewater molecule coordinated, which we believe to becoordinated to the four-coordinated metal centre. Thetitration curve is shown in Fig. 6 for the[Cu2(TPPNOL)(OAc)(H2O)](ClO4)2 and [Cu2-(TPPNOL)(OAc)(OH)](ClO4)2 species. The pKa valueof 8.0 was determined, which is associated with theaquo/hydroxo equilibrium (see Scheme 1). Complex 2can therefore provide hydroxide ions at pH�8.0,which can act as a nucleophile.

3.7. Catalytic acti�ity for the oxidation of3,5-di-tert-butylcatechol

The two-electron oxidation of 3,5-dtbc to quinonewas investigated since this is the reaction that thecopper-containing enzyme catechol oxidase catalyses.There are a number of factors that need to be consid-ered, such as electrochemical properties, the geometryimposed by the ligands on the metal ion, the nature ofany exogenous donors, the basicity of the donor atomand the steric features of the ligands [35]. In order tocompare the difference between mononuclear and dinu-clear complexes, we have studied our copper complexesfor catecholase activity, using the changes in the elec-tronic spectrum after the addition of the 3,5-dtbcsubstrate.

Quinone has a �max=400 nm, and we monitored theincrease in absorbance versus time at this wavelength.For complex 2, an increasing in the quinone band withtime was observed, as shown in Fig. 7. The mononu-clear complex 1 was inactive in this oxidation reaction.

Kida and co-workers have studied a number ofCu(II) complexes and have shown that dinuclear com-plexes catalyse the reaction while the mononuclear spe-cies are either not as efficient or inactive depending onthe steric demands of the ligands [36]. This behaviourfor 1 can be explained by the fact that two coppercentres must be nearby to link two phenolic oxygenatoms of catechol to mediate the redox reaction.

In a preliminary experiment, to show the catecholoxidase activity on pH, 100 �l of the substrate 3,5-dtbc

Fig. 6. Plot of species distribution [Cu2(TPPNOL)(OAc)(H2O)]-(ClO4)2 and [Cu2(TPPNOL)(OAc)(OH)](ClO4)2, with pH.

Scheme 1. Aquo/hydroxo equilibrium for the dinuclear complex [Cu2(TPPNOL)(OAc)](ClO4)2 (2), in water, at pH 8.0.

C. Fernandes et al. / Inorganica Chimica Acta 320 (2001) 12–21 19

Fig. 7. Increase of the quinone band after addition of 3,5-dtbc (100 �lof 0.165 M solution) to a solution of 2 (8×10−4 M in 3 ml ofoxygenate methanol). The spectra were recorded every 1 min.

(0.165 M) was added to a solution containing 100 �l ofthe complex (2.4×10−3 M), 100 �l of the buffer solu-tion (5×10−4 M; pH 6.0, 6.57, 7.0, 7.5, 8.0, 8.5), in 3ml of oxygenated methanol, in a spectrophotometriccell at 25 °C. In all cases, catecholase activity wasnoted. An increase in the formation of the o-quinone atpH 8.05 was observed (see Fig. 8). This result is inagreement with the titration experiments, which suggesta pKa=7.9 for the aquo/hydroxo equilibrium for com-plex 2 (see Scheme 1). The effect of pH on the oxida-tion reaction indicates that the deprotonation of thecoordinated water molecule results in increasing qui-none formation, and it is in accordance with the pKa

value determined by potentiometric titration. At pH8.0, the hydroxide ion coordinated to one copper centredeprotonates the 3,5-dtbc substrate which coordinatesas a bridging ligand to the copper centres, prior to theintramolecular electron transfer reaction. The electrontransfer reaction than takes place, increasing the qui-none formation as the determining step of the reaction,and the Cu2

I species in the presence of molecular oxy-gen, is reoxidised back to the original copper(II) com-plex (see proposed mechanism in Scheme 2). Applyingthe Michaelis–Menten approach it is possible to char-acterise the kinetic behaviour using the parameters: kcat

3.12×10−3 s−1, KM 1.19×10−3 M and kcat/KM 1.62M−1 s−1 (Fig. 9).

The electrochemical properties of 1 and 2 have beenstudied to determine whether a correlation exists be-tween their redox properties and the reactivity toward3,5-dtbc (see Fig. 5). Initially, there was no obviouscorrelation between the redox potential and the reactiv-ity, since mononuclear complex 1 has a redox potentialanodically shifted compared to 2, which however, isable to oxidise 3,5-dtbc to quinone, while complex 1 is

Fig. 8. Dependence of the reaction rates on pH for the oxidation of3,5-dtbc catalysed by complex 2.

Scheme 2. Proposed mechanism for the interaction between the dinuclear complex and the 3,5-dtbc substrate.

C. Fernandes et al. / Inorganica Chimica Acta 320 (2001) 12–2120

Fig. 9. Dependence of the reaction rates on the 3,5-dtbc concentra-tions for the oxidation catalysed by complex 2.

References

[1] P.A. Vigato, S. Tamburini, D.E. Fenton, Coord. Chem. Rev.106 (1990) 25.

[2] V. McKee, M. Zvagulis, V. Dagdigian, M.C. Path, C.A. Reed,J. Am. Chem. Soc. 106 (1984) 4765.

[3] K.A. Magnus, H. Ton-That, J.E. Carpenter, Chem. Rev. 94(1994) 727.

[4] E.I. Solomon, M.J. Baldwin, M.D. Lowery, Chem. Rev. 92(1992) 521.

[5] O. Hayaishi, Oxygenases, Academic Press, New York, 1962.[6] H. Sigel, Metal Ions in Proteins, Marcel Dekker, New York,

1981.[7] A. Volbeda, W.G. Hol, J. Mol. Biol. 206 (1989) 531.[8] A. Volbeda, W.G. Hol, J. Mol. Biol. 209 (1989) 249.[9] E.I. Solomon, K.W. Penfield, D.E. Wilcox, Struct. Bonding

(Berlin) 53 (1993) 1.[10] G.L. Woolery, L. Powers, M. Winkler, E.I. Solomon, K. Lerck,

T.G. Spiro, Biochim. Biophys. Acta 788 (1984) 155 (see alsopp. 23 and 501).

[11] R. Than, A.A. Feldman, B. Krebs, Coord. Chem. Rev. 182(1999) 211.

[12] W. Kaim, B. Schwedereski, Bioinorganic Chemistry, Wiley,Chichester, 1994.

[13] A. Rompel, H. Fischer, K. Buldt-Karentzopoulos, D. Meiwes,F. Zippel, H.-F. Nolting, C. Hermes, B. Krebs, H. Witzel, J.Inorg. Biochem. 59 (1995) 715.

[14] F. Zippel, F. Ahlers, R. Werner, W. Haase, H.-F. Nolting, B.Krebs, Inorg. Chem. 35 (1996) 3409.

[15] T. Kablunde, C. Eicken, J.C. Sacchettini, B. Krebs, NatureAmerica 5 (1998) 1084.

[16] T.N. Sorrell, D.L. Jameson, C.J. O’Connor, Inorg. Chem. 23(1984) 190.

[17] A. Neves, L.M. Rossi, A. Horn, I. Vencato, A.J. Bortoluzzi, C.Zucco, A.S. Mangrich, Inorg. Chem. Commun. (1999) 334.

[18] F. Zippel, F. Ahlers, R. Werner, W. Haase, N. Hans-Friedrich,B. Krebs, Inorg. Chem. 35 (1996) 3409.

[19] J.H. Satcher, M.W. Droege, T.J.R. Weakley, R.T. Taylor, In-org. Chem. 34 (1995) 3317.

[20] F. Tuna, L. Patron, Y. Journaux, M. Andruch, W. Plass, J.-C.Trombe, J. Chem. Soc., Dalton Trans. (1999) 539.

[21] C. Fernandes, A. Neves, I. Vencato, A.J. Bortoluzzi, V.Drago, T. Weyhermuller, E. Rentschler, Chem. Lett. (2000)540.

[22] O. Kahn, Molecular Magnetism, VCH, New York, 1993.[23] A.E. Martell, R.J. Motekaitis, Determination and Use of Sta-

bility Constants, 2nd ed., VHC, New York, 1992.[24] C. Fernandes, A. Neves, A.J. Bortoluzzi, B. Spoganicz, E.

Schwingel, Inorg. Chem. Commun., in press.[25] G.M. Sheldrick, SHELXS-97, Program for the Solution of Crys-

tal Structures, University of Gottingen, Germany, 1990.[26] G.M. Sheldrick, SHELXL-97, Program for the Refinement

of Crystal Structures, University of Gottingen, Germany,1997.

[27] I. Zsolnai, ZORTEP, An Interactive ORTEP Program, Universityof Heidelberg, Germany, 1997.

[28] A.W. Addison, T.N. Rao, J. Reedijk, J. Van Rijn, G.C. Ver-schoor, J. Chem. Soc., Dalton Trans. (1984) 1349.

[29] V. McKee, M. Zvagulis, J.V. Dagdigian, M.G. Patch, C.A.Rud, J. Am. Chem. Soc. 106 (1984) 4765.

[30] N.N. Murthy, K.D. Karlin, C. Berini, J. Luchinat, J. Am.Chem. Soc. 119 (1997) 2156.

[31] Y. Nishida, T. Tokii, Y. Mori, J. Chem. Soc., Chem. Commun.(1988) 675.

inactive. It seems, as observed by Malachowski andco-workers, that the electrochemical potential displaysa secondary role in determining the reactivity propertiesof copper complexes [37].

4. Conclusions

A dinuclear copper complex 2, in which the metalcentres show different coordination numbers, was pro-duced by the direct reaction of copper perchlorate andsodium acetate, with the unsymmetrical ligand HTPP-NOL. The mononuclear complex 1 was formed byreaction of copper perchlorate with the ligand. Com-plex 2 described and characterised herein provides someimportant features for an effective catechol oxidasemodel. The metal coordination spheres, the Cu�Cudistance and the reactivity with the 3,5-dtbc substrateindicate that 2 is a potential model for the dinuclearcopper site in catechol oxidase. Although the mononu-clear complex 1 exhibits a higher redox potential than2, it was observed that 1 is ineffective in the oxidationof 3,5-dtbc, since two metallic centres must be closetogether in order to link the substrate and to promotethe electronic transfer. For 2, at pH 8.0, an increase inthe catecholase activity was noted, since in this pH thecomplex shows a hydroxide coordinated, as indicatedby titration experiments. It can be therefore concludedthat the hydroxide ion deprotonates the 3,5-dtbc sub-strate, and results in increasing o-quinone formation.

Acknowledgements

This work was supported by grants from the CNPqand PRONEX (Brazil) and MPI (Germany).

C. Fernandes et al. / Inorganica Chimica Acta 320 (2001) 12–21 21

[32] G. Christou, S.P. Perlepes, E. Libby, K. Folting, J.C. Huffman,R.J. Webb, D.N. Hendrickson, Inorg. Chem. 29 (1990) 3657.

[33] U. Rajendran, R. Viswanathan, M. Palaniadavas, N. Laskimi-naraya, J. Chem. Soc., Dalton Trans. (1994) 1219.

[34] Y. Nakao, Y. Oohata, R. Fujiwara, M. Itadani, T. Sakurai, A.Ichimura, W. Mori, Bull. Chem. Soc. Jpn. 66 (1993) 2112.

[35] M.R. Malachowski, H.B. Huynh, L.J. Tomlinson, R.S. Kelly,J.W. Furbeee, J. Chem. Soc., Dalton Trans. (1995) 31.

[36] N. Oishi, Y. Nishida, K. Ida, S. Kida, Bull. Chem. Soc. Jpn. 53(1980) 2847.

[37] M.R. Malachowski, B. Dorsey, J.G. Sackett, R.S. Kelly, A.L.Ferko, R.N. Hardin, Inorg. Chim. Acta 249 (1996) 85.

.