Binding and Activation of Dioxygen by Copper Complexes

The first part of this work focused on the synthesis and characterization of a new class of peroxo copper(II) complexes. By reaction of the tridentate amino alcohol ligands 2,6-bis(pyrrolidinomethyl)-4-methylphenol (Hbpyrmp), 2,6-bis(piperidinomethyl)-4-methylphenol (Hbpipmp) and 2,6-bis(morpholinomethyl)-4-methylphenol (Hbmmp), respectively, with two equivalents of copper(II) perchlorate hexahydrate in methanolic solution, and treatment of the resulting solution with triethylamine and 3,5-di-tert-butylcatechol (3,5-DTBC), the copper coordination compounds

were obtained. Alternatively, the complexes may be synthesized using hydrogen peroxide instead of 3,5-DTBC. Complex 1 is one of a few examples of peroxo copper(II) complexes which could be characterized by single crystal X-ray diffraction analysis. The structures of only two dinuclear µ-peroxo copper(II) complexes could have been determined by X-ray crystallography yet. 1 and the analogous compounds 2 and 3 are tetranuclear complexes in which the peroxo ligand is bound to all four copper centers in an unusual µ4 coordination mode. The peroxo ligand is located above the rectangle defined by the four metals. Due to this coordination mode of peroxide the complexes exhibit remarkable stability. In contrast to most other peroxo copper(II) complexes they can be synthesized and isolated at room temperature. A further interesting structural motive is the bridging µ-ClO4- group. Compound 1 is the first crystallographically characterized complex in which perchlorate is found as a fourfold coordinating ligand.

The electronic spectra of complexes 1 - 3 in dichloromethane are dominated by a strong absorption at 390 nm with a shoulder at 420 - 440 nm. This structure can be interpreted as a superposition of a O22- to Cu(II) and a phenolate to Cu(II) charge transfer transition. A further band at about 580 nm is possibly caused by superposition of a second, less intense O22- to Cu(II) charge transfer transition with the d-d transitions.

As part of the investigation of vibrational spectroscopic properties 1 was also synthesized with 18O2 resulting in the isotopic labelled derivative 1-18O. In order to identify v(O-O) Raman and resonance Raman spectra of 1, 1-18O, 2 and 3 were recorded besides the routine IR spectra. In the resonance Raman spectrum of 1 bands at 878 and 353 cm-1 are observed shifting to 841 and 347 cm-1 in 1-18O. They are assigned to v(O-O) and v(Cu-O), respectively. The frequency shifts are not in accordance with the expected shifts for a harmonic diatomic oscillator. The shift of the band at 823 to 817 cm-1 also indicates coupling of vibrations within the central Cu4 unit. In the resonance Raman spectra of 2 and 3 the corresponding bands are found at 898 and 352 cm-1, and at 888 and 353 cm-1, respectively. The values of v(O-O) which were determined for complexes 1 - 3 are the highest ever observed for peroxo copper(II) complexes. This fact can therfore be denoted as a characteristic feature of µ4-peroxo copper(II) complexes.

As the precursor complex for the synthesis of the µ4-peroxo copper(II) complex 1 the tetranuclear µ4-oxo copper(II) complex

was identified. The structure of 4 is closely related to the structure of 1. Instead of the peroxo group it contains a O2- oxo anion in the center which is located in plane with the four copper centers. This structural motive is very unusual, only one other complex is known with an exactly planar Cu44-O) unit.

In the charge transfer region of the electronic spectrum of 4 in dichloromethane no distinct band is observed, only a broad shoulder which is attached to the inner ligand band in the range from 320 to 440 nm. The maximum of the d-d bands is located at 645 nm which is in agreement with square-planar copper(II) ions.

Magnetic susceptibilities of complexes 1 and 4 were measured by the Faraday-Curie method. The data were fitted to an isotropic Heisenberg model using three coupling constants J12, J13 and J14. In both complexes two of the exchange interactions are antiferromagnetic pre-dominating the remaining ferromagnetic interaction. The antiferromagnetic exchange interactions are strongly correlated with the ferromagnetic interaction preventing an exact determination of the parameters. The antiferromagnetic coupling constants are calculated to -2J = 250 +/- 50 cm-1 for 1 and -2J = 340 +/- 50 cm-1 for 4.

On treatment of a methanolic solution of complex 4 with sodium azide the complex

was obtained. Compound 5 crystallizes in a one dimensional polymeric structure, in which the dinuclear [Cu2(bpyrmp)(OMe)(N3)2] subunits are connected by antisymmetric bis(µ-1,3)-azide bridges. This compound is one of some rare examples of structurally characterized azido copper(II) complexes involving a dinucleating ligand in which the exogenously bridging solvent molecule is not substituted by azide.

The electronic spectrum of 5 shows a broad and strong absorption with a maximum at 360 nm. This is due to a superposition of a phenolate to Cu(II) and a N3- to Cu(II) charge transfer transition but which is clearly dominated by the latter. The position of the band is in line with consideration of terminal coordinated azide ions. The µ-1,3-bridges break at the weak axial azide-copper bond resulting in "free" dinuclear subunits. The maximum of the d-d transitions is found at 690 nm.

The reaction system forming after mixing of the tridentate dinucleating ligands with two equivalents of copper(II) perchlorate hexahydrate in methanol was investigated in more detail using UV/Vis spectroscopic methods. The Cu2bpyrmp system was studied examplarily. In basic methanolic solution two Cu2bpyrmp units dimerize to a tetranuclear µ4-oxo copper(II) complex with a structure analogous to the structurally characterized complex 4. This system forms structurally different complexes depending on the added ligands or substrates. Treatment of the µ4-oxo complex with hydrogen peroxide leads to a µ4-peroxo copper(II) complex showing a structure analogous to the structure of 1. This compound can also be generated by reaction of the µ4-oxo complex with 3,5-DTBC. Then peroxide is formed in situ by reduction of dioxygen. Treatment of the µ4-oxo complex with sodium azide leads to degradation of the tetranuclear structure. Dinuclear complexes with terminal azides and a structure analogous to the one observed in 5 are formed. A spectrophotometric titration shows that azide ions do not act as bridging ligands in the µ4-oxo complex and also do not substitute the bridging solvent molecule in the dinuclear product.

The peroxo complexes 1, 2 and 3 are stable in the solvents dichloromethane, acetone and, when an excess of H2O2 is used, in methanol. In acetonitrile the complexes decompose within several hours. Kinetic analysis reveals that the reaction is first-order in complex concentration. The rate constant was determined to k = 0.009 min­1 for 1.

The second part of this work focused on dinuclear copper(II) coordination compounds as functional and structural models for the type 3 copper enzyme catechol oxidase. The symmetric ligands 4-bromo-2,6-bis(4-methylpiperazin-1-ylmethyl)phenol (HL1), 4-bromo-2,6-bis((2-pyridylmethyl)aminomethyl)phenol (HL2), 4-bromo-2,6-bis((2-(2-pyridyl)ethyl)aminomethyl)phenol (HL3), 4-bromo-2,6-bis((2-(1-methyl-2-imidazolyl)ethyl)aminomethyl)phenol (HL4) and the unsymmetric ligands 4-bromo-2-(4-methylpiperazin-1-ylmethyl)-6-((2-pyridylmethyl)aminomethyl)phenol (HL5), 4-bromo-2-(4-methylpiperazin-1-ylmethyl)-6-((2-(2-pyridyl)ethyl)aminomethyl)phenol (HL6), 4-bromo-2-(4-methylpiperazin-1-ylmethyl)-6-((2-(1-methyl-2-imidazolyl)-ethyl)aminomethyl)phenol (HL7) were employed in the synthesis of the complexes

Compounds 6 - 12 are a series of structurally similar dinuclear copper(II) complexes. All of the complexes are µ-phenoxo bridged by the pentadentate dinucleating ligand and by a µ-hydroxo- or a µ-methanolato bridge. They differ by the terminal nitrogen donors of the ligand. For the symmetric complexes two piperazine units (6), two (2-pyridylmethyl)amine units (7), two (2-(2-pyridyl)ethyl)amine units (8) and two 2-(1-methyl-2-imidazolyl)ethyl)amine units were used. For the unsymmetric complexes a combination of one piperazine unit with one of the three other mentioned functional groups were employed (10, 11, 12). The structures of 6, 8, 9, 11 and 12 were determined by single crystal X-ray structure analysis. The complexes represent good structural models for the active sites of the met forms of catechol oxidases. They simulate the short Cu-Cu distance of about 3 Å as well as the N2O2 donorset of the enzyme.

All of the spectra contain an absorption in the charge transfer region, i.e. in the range from 300 to 500 nm. This is interpreted as a superposition of a PhO- to Cu(II) and a OH- to Cu(II) or MeO- to Cu(II) charge transfer transition. It cannot be distinguished whether the complexes isolated with a hydroxo bridge have eventually replaced this bridge by a methanolato bridge. The maxima of the d-d bands lie in the range from 613 to 646 nm. These positions are in accord with tetragonal coordination environments. The UV/Vis spectra of the symmetric complexes 6 - 9 reveal no clear trends. This comprises the position and intensity of the charge transfer bands as well as the positions of the d-d transitions. The UV/Vis spectra of the unsymmetric complexes 10 - 12 show a close relation to the spectra of the corresponding symmetric complexes.

The cyclic voltammograms of 6 - 12 reveal irreversible and ill defined cathodic reduction peaks in the range from -0.37 to -0.96 V. The reduction peaks are assigned to the one electron processes CuIICuII to CuIICuI and CuIICuI to CuICuI. The complexes exhibit a broad anodic oxidation peak in the range from +0.06 to +0.18 V. The breadth of these peaks suggests a superposition of two close one electron processes, which are assigned to the reoxidations to the CuIICuI and CuIICuII species, respectively.

The investigation of the catecholase activity of 6 - 12 revealed that only complexes 6, 10, 11 and 12 have significant catalytic activity with respect to the aerial oxidation of 3,5-DTBC to its corresponding o-quinone. Saturation kinetics were observed for the four compounds. A treatment on the basis of the Michaelis-Menten model was applied. Turnover numbers of 214 h-1 (6), 48 h-1 (11), 43 h-1 (12), 33 h-1 (10) and Michaelis constants KM of 2.410-4 mol/l (6), 3.110-4 mol/l (11), 1.410-3 mol/l (12) and 2.310-3 mol/l (10) were calculated. The differences in reactivity can be explained considering geometric parameters. Binding of catechol is only favoured for those complexes which contain the piperazine unit, for the other complexes the structural motive of a µ-hydroxo or µ-alkoxo bridge is more favoured. This was proven by UV/Vis spectroscopic investigations.


arrowBack to Dissertation