Copper(I) complexes supported by certain capping ligands react with molecular oxygen in a 2:1 ratio to afford dinuclear copper dioxygen (Cu2/O2) complexes. Such complexes can be regarded as structural and functional models of the reactive intermediates of tyrosinase and catechol oxidase.
Numerous review articles on the subject have been published so far.82–95
The first structurally characterized dinuclear copper dioxygen complex was reported by Karlin and co-workers in 1988.96Reaction of the copper(I) complex supported by the tetradentate ligand TPA (tris(2-pyridylmethyl)amine, Scheme 3) with dioxygen at 80C in an organic solvent afforded a intensely purple-colored dinuclear copper(II) complex (1) having atrans-–1,2-peroxo bridge (Figure 4).96The penta-coordinate copper(II) ions exhibit a distorted trigonal bipyramidal geometry, where the equatorial ligands are the three pyridine nitrogens and one of the peroxo oxygen atoms and the aliphatic amine occupy the axial positions.96 The OO bond length of 1.43 A˚ is typical for a peroxo ligand (Table 1),97and the two cupric ions are antiferromagnetically
N
N N
R R
R
N
N
N N
N N N
N N H
Me
Me
N N
N H
Me Me
N N
R2 R1
B
H N N
N N R2 R1
R2 R1
N N
N
N
TPA
Bn3TACN: R = –CH2Ph i-Pr4DTNE
N
N Me
Et
Me
Et N
N
Me Me Me Me N
N Me Me L1
HB(3-CF3-5-CH3pz)3–: R1 = CF3, R2 = CH3 HB(3,5-i-Pr2pz)3–: R1 = R2 = i-Pr
LME L2 LTMPD
i-Pr3TACD N
N N
Scheme 3
coupled through the end-on peroxo bridge, making the complex EPR-silent.96 Broken-symmetry SCF-X-SW calculations have suggested that the electronic structure is dominated by the inter- action of the peroxide*-orbital (one of the antibonding*-orbitals of the O22ligand residing in the metal-peroxide plane) with the half-occupied copper dz2-orbitals.98 Complex (1) exhibits
Me
Me
N N
N
N
Me2TPA
P N
N t-Bu
t-Bu
SiMe3
SiMe3 t-Bu2P(NSiMe3)2–
N
N
N
N N N
R1
R2
N R
N N
Py2R
N R2 R1
N
Py1R1,R2
N NEt2
X ArPyNEt2
R1XYLR2
Scheme 3 continued
resonance Raman bands at 832 cm1 (788 cm1 with 18O2) and 561 cm1 (535 cm1 with 18O2), which are assigned as the OO and CuO stretching vibrational modes, respectively.98Peroxo- to-copper(II) charge-transfer bands appear at 525 nm ("ẳ11,500 M1cm1) and 590 nm (7,600 M1cm1, shoulder), together with ad–d band at 1,035 nm (160 M1cm1) due to trigonal bipyramidal copper(II) ( Table 1).96,98,99 These spectroscopic and structural features are signifi- cantly different from those of the oxy forms of hemocyanin, tyrosinase, and catechol oxidase, due to the different binding mode of the peroxo ligand in (1) and the proteins. Nonetheless, the (TPA)CuI precursor can bind dioxygen reversibly as in the case of the enzymes.99Similar (trans- -1,2-peroxo)dicopper(II) complexes have been generated using a series of tripodal tetradentate amine ligands, and the kinetics and thermodynamics of the reversible dioxygen binding process have been investigated in detail in some instances.94,100–108 In general, the-peroxo dicopper(II) complex is formed through a mononuclear (superoxo)copper(II) intermediate, and room-tempera- ture stability of the peroxo complex is precluded by a strongly negative reaction entropy.104,109
Adoption of tridentate instead of tetradentate ligands resulted in a drastically changed structure of the (-peroxo)dicopper(II) product. In initial studies by Kitajima and co-workers, a (-2:2- peroxo)dicopper(II) complex (2) was obtained as a deep purple solid when the copper(I) complex of a tridentate hydrotris(pyrazolyl)borate ligand (HB(3,5-i-Pr2pz)3, Scheme 3) was treated with dioxygen at 78C in acetone.41,110 The crystal structure of (2) was determined by X-ray crystallographic analysis (Figure 5).41,42 The same compound was obtained by the reaction of a bis(-hydroxo)dicopper(II) complex with the same ligand and H2O2under similar reaction conditions.42
The OO bond length of (2) is 1.41 A˚, which is nearly the same as that of compound (1) (1.43 A˚) (Table 1), but the CuCudistance (3.56 A˚) is significantly shorter than that in (1) (4.36 A˚) du e to the side-on peroxo bridging mode instead of the end-on type in (1).41,42Complex (2) exhibits a strong absorption feature at 349 nm ("ẳ21,000 M1cm1) together with a weak one at 551 nm (790 M1cm1), which were assigned as peroxide-to-copper(II) charge-transfer transitions (Table 1).111
O
Cu O
Cu
Figure 4 Chem 3D representation of the crystal structure of complex (1).96
Table 1 The OO bond length and the UV–vis and resonance Raman data of representative (-peroxo)dicopper(II) complexes that have been characterized by X-ray crystallography.
Complex
d(OO) (A˚)
max(") (nm (M1cm1))
(16O) ( (18O)) (cm1)
(1)96 1.432(6) 435 (1,700), 524 (11,300) OO 832 (788)
615 (5,800), 1035 (180)
(2)41,42 1.412(12) 349 (21,000), 551 (790) OO 741 (698)
(3)122 1.485(8) 360 (24,700), 532 (1,530) OO 760 (719)
(4)119 1.374(5) 380 (22,000), 520 (2,300) OO 739 (696)
The resonance Raman spectrum obtained by excitation into the low-energy charge-transfer band shows an OO stretching vibration at 741 cm1that shifts to 698 cm1upon18O-substitution (Table 1).41,42 The lower O–O value of (2) indicates that the OO bond of the side-on peroxo complex is significantly weakened as compared to that in the end-on peroxide (832 cm1). Com- pound (2) is also ESR-silent due to a strong antiferromagnetic interaction between the two cupric ions through the side-on peroxo bridge.41,42The similarity of the spectroscopic features of (2) and the oxy forms of the proteins strongly suggested that the peroxo state of the enzymes also possessed the side-on type binding mode. Indeed, five years after Kitajima’s report (1989), the structure of oxy-hemocyanin was solved and it was concluded that oxy-hemocyanin has the same (-2:2- peroxo)dicopper(II) core as in (2).22The correct prediction of the active oxygen intermediate in the enzymatic system by the model studies represents one of the greatest successes in bioinorganic chemistry. Subsequently, (-2:2-peroxo)dicopper(II) complexes that exhibit similar spectroscopic features (UV–vis, resonance Raman, EPR) to those of (2) but which are supported by various capping ligands have been reported,112–129and some of these also have been structurally character- ized (Table 1).115,119,122,129
The unique spectroscopic features of the (-2:2-peroxo)dicopper(II) core have been inter- preted by Solomonet al. using broken-symmetry SCF-X-SW calculations.1,16,130,131The HOMO and LUMO for the (-2:2-peroxo)dicopper(II) complex are formed by the interaction of the peroxide valence orbitals with the half-occupieddx2y2orbitals of the two copper(II) ions (Figure 6).1,16 There is a strong -bonding interaction between the *-orbital and the symmetric combin- ation of dx2y2 orbitals (Cu(dx2y2)þCu(dx2y2)) to form the LUMO.1,16 Since, the peroxide ligand has four-donor orbitals to the two copper centers, the bonding/antibonding interaction becomes much larger than that in the end-on peroxo bridge, in which there are only two-donor interactions.1,16 The larger -donor interaction in the side-on peroxo complex leads to a high energy for the peroxide*to copper(II) charge-transfer transition (350 nm) and a high intensity (20,000 M1cm1) compared to those of the end-on counterparts (530 nm, 10,000 M1cm1).1,16On the other hand, the interaction between the antisymmetric combination of dx2y2 orbitals (Cu(dx2y2)Cu(dx2y2)) and the high-energy unoccupied *-orbital on the peroxide ligand comprise the HOMO, in which some back-bonding of electron density from the copper centers into the*-orbital of the peroxide occurs.1,16 Since the*-orbital is strongly anti-bonding with respect to the OO bond, the-electron back donation from the copper ions to the peroxide ligand weakens the OO bond of the (-2:2-peroxo)dicopper(II) complex.1,16This is reflected in the significantly lower OO bond stretching frequency (740 cm1) of the side-on peroxide complex as compared to that of the end-on peroxide congener (830 cm1).1,16 In addition, very strong-donation (destabilization of the LUMO) and the-electron back donation (stabilization of HOMO) leads to the large stabilization of the singlet ground state, which under- lies the strong antiferromagnetic coupling between the two cupric ions.1,16
Interestingly, the thermal stability of the peroxo complexes supported by a relatively rigid dinucleating ligand L1 (complex (3), Figure 7) and by the perfluorinated hydrotris (pyrazolyl)
O
Cu O
Cu
Figure 5 Chem 3D representation of the crystal structure of complex (2).41
borate ligand HB(3-CF3-5-CH3pz)3is significantly enhanced as compared to those of the others, so that they can be handled even at room temperature.122,123 Furthermore, reversible dioxygen binding is accomplished using complex (3).122The square pyramidal geometry of each copper ion and the Cu2O2core in (3) are relatively distorted as compared to those of complex (2), due to the steric restriction induced by the ligand strap of L1(Table 1). The conformation of the axial ligands in (3) is syn, whereas they are anti in (2). A perdeuterated macrocyclic ligand d21-i-Pr3 TADC (the isopropyl groups are deuterated, Scheme 3) also allowed them to isolate (-2:2-peroxo) dicopper(II) complex (4).119
8.15.4.2 Bis(m-oxo)dicopper(III) Complexes
Tolman and co-workers brought about another breakthrough in the Cu2/O2 chemistry in 1995.90,92,132 Oxygenation of the copper(I) complex supported by 1,4,7-triazacyclononane ligand (TACN) carrying a bulky substituent such as benzyl group (Bn3TACN, Scheme 3) at a low temperature (80C) gave an unprecedented complex (5),132 the X-ray structure of which was determined by adopting per-deuterated ligand d21-Bn3TACN.133,134 It is a bis(-oxo)dicopper (III) complex as shown in Figure 8, in which each copper ion exhibits a distorted square pyramidal geometry with the N3O2 donor set. There is no covalent bond between the two
πσ* O
Cu
O
Cu
O22–
Cu Cu
σ* O
Cu
O
Cu
O22–
LUMO
dx2 –y2 x2 –y2
LUMO
HOMO
HOMO d
Figure 6 Electronic structure of the (-2:2-peroxo)dicopper(II) complex.1,16
O Cu
O
Cu
Figure 7 Chem 3D representation of the crystal structure of complex (3).122
oxygen atoms in the Cu2O2core (OOẳ2.29 A˚), while the CuCudistance becomes significantly shorter (CuCuẳ2.79 A˚) than that of the (-2:2-peroxo)dicopper(II) core (CuCuẳ3.56 A˚) (Table 2).133,134The CuO (1.80 A˚) and Cu—Neq (1.99 A˚) bond lengths are also shorter than those of the corresponding bis(-hydroxo)dicopper(II) complexes (1.94 A˚, and2.06 A˚, respect- ively).133,134These data are consistent with the high oxidation level of the bis(-oxo)dicopper(III) complex.134A bond valence sum (BVS) analysis of the crystallographic data strongly supports the copper(III) oxidation state of the complex.134The axial ligands, on the other hand, weakly coord- inate to the copper ion with an anti-configuration (CuNaxẳ 2.30 A˚).134Compound (5) exhibits two intense absorption bands at 318 nm (12,000 M1cm1) and 430 nm (14,000 M1cm1), and excitation of the lower-energy band resulted in intense resonance enhancement of Raman bands at 602 cm1and 612 cm1(with16O2), which collapsed into one band at 583 cm1upon18O2substitution (Table 2).134
Since the initial report of (5), similar bis(-oxo)dicopper(III) complexes with a variety of ligands have been reported,115,116,118,119,121,128,135–145 some of which [(d28-i-Pr4DTNE)CuIII2(-O)2]2þ (6),135[(LME)2CuIII2(-O)2]2þ(7),137[(Me2TPA)2CuIII2(-O)2]2þ(8),141and [(t-Bu2P(NSiMe3)2)2CuIII2
(-O)2]2þ(9),143also have been characterized by X-ray crystallographic analysis (Table 2). In the absence of X-ray crystallographic information, the presence of the bis(-oxo)dicopper(III) core has also been deduced by X-ray absorption fine structure (EXAFS) spectroscopy.115,134,138
The copper(III) oxidation state of the bis(-oxo) core has been confirmed directly by a CuK-edge X-ray absorption spectroscopy (XAS) study on compound (7).146All complexes exhibit two intense
O
Cu O
Cu
Figure 8 Chem 3D representation of the crystal structure of complex (5).133,134
Table 2 The CuCudistance and the UV–vis and resonance Raman data of representative bis(-oxo) dicopper(III) complexes that have been characterized by X-ray crystallography.
Complex d(CuCu)
(A˚)
max(")(nm (M1cm1)) (16O) ( (18O)) (cm1) (5)134 2.794(2) 318 (12,000), 430 (14,000) 602, 612 (583)
(6)135 2.783(1) 316 (13,000), 414 (14,000) 600 (582)
(7)137 2.743(1) 306 (21,000), 401 (28,000) 610 (587)
(8)141 2.758(4) 258 (36,000), 378 (19,000) 590 (564)
(9)143 2.906(1) 444 (10,000) Not reported
oxo-to-copper charge transfer bands around 300 nm and 400 nm and an intense resonance Raman band at600 cm1, which has been assigned to the core breathing mode of the bis(-oxo)dicopper unit (Figure 9).147,148 The isotope shift of this band ( (16O2) (18O2)) is 20 – 30 cm1.92 In addition to the core breathing mode around 600 cm1, a weak Raman feature at 100 – 200 cm1 due to the core bending mode exists,147although little effort has so far been made to detect such a lower-energy frequency. In the case of an unsymmetric didentate ligand system,136,139a pairwise stretching mode around 600 cm1also has been detected. Furthermore, the bis(-oxo)dicopper(III) complexes are ESR-silent, consistent with the d8 electronic configuration of the tetragonal copper(III) ions. Computational analyses of the bis(-oxo)dicopper(III) core have been performed using different donor sets at various levels of theory.119,134,147,149–152Details about the structure, spectroscopic features, and reactivity of the bis(-oxo)dicopper(III) complexes are summarized in a recent review article.92
8.15.4.3 Formation Mechanisms
So far, we have learned that copper(I) complexes with certain ligands react with molecular oxygen to produce three types of Cu2/O2 complexes. Tetradentate ligands usually provide the (trans- –1,2-peroxo)dicopper(II) complex, whereas the tridentate and didentate ligands afford the (-2:2-peroxo)dicopper(II) and/or the bis(-oxo)dicopper(III) complexes (Scheme 4). The ligand denticity might be the major factor that controls the different binding mode (end-on vs. side-on) in the peroxo dicopper(II) complexes. Since the copper(II) ion generally favors a five-coordinate geometry (square pyramidal and trigonal bipyramidal), the tetradentate ligand can provide only one vacant site available for the adduct formation with O2, thus producing the end-on peroxo complex. On the other hand, the tridentate and didentate ligands can afford at least two vacant sites for the reaction with O2, making it possible to form the side-on adduct. In this respect, it is interesting to note that all the dicopper enzymes so far known have tri-coordinate dicopper units, producing the (-2:2-peroxo)dicopper(II) as the common intermediate.
LCuII O
O CuIIL + O2
LCuI
– O2
LCuII O O
LCuII O O
CuIIL LCuII
O O
+ O2 – O2
+ LCuI – LCuI
– LCuI + LCuI
LCuI II O O
CuII IL η1-superoxo
η2-superoxo à-η2:η2-peroxo
trans-à-1,2-peroxo
bis(à-oxo)
(d) (e)
(a)
(b) (c)
(f) (g)
Scheme 4
O Cu Cu
O N
N
N N O
Cu Cu O N
N
N N O
Cu Cu
O N
N
N N
(A) (B) (C)
Figure 9 Main vibrational (A) core breathing mode, (B) pairwise stretching mode, and (C) core bending mode of the bis(-oxo)dicopper(III) core.92
In the model reactions, mononuclear copper(II)-superoxo complexes both in the end-on (1-superoxo) and side-on (2-superoxo) binding modes have been characterized in the reactions of copper(I) complexes and O2.100,102,104,153 Such mononuclear copper(II)-superoxo complexes could be the intermediates for the formation of dicopper(II)-peroxo complexes as illustrated in Scheme 4(routes (a)–(b) and (d)–(e)), although direct evidence has yet to be obtained in order to confirm the postulated mechanism. Interconversions between the1- and2-superoxo frameworks (path (f )) and the (trans-–1,2-peroxo) and (-2:2-peroxo) bridging ligands (path (g)) are also possible depending on the ligand used. In fact, the pathway between the end-on and side-on peroxo ligand frameworks (path (g)) has been suggested to be involved in some instances.154,155 Depending on the ligand structure used, the kinetics of the process of formation of the peroxo intermediate may be first or second order with respect to the copper complexes.92,109,120,133This difference has been attributed to a different rate-determining step; the (superoxo)copper(II) complex formation (path (a)) vs. the reaction between the superoxo complex and copper(I) precursor (path (b)). Computational studies on the dioxygen binding process in the dinuclear copper(I) site of the enzymes have been performed to provide insights into energetics as well as an intersystem crossing mechanism from triplet O2to the singlet (-peroxo)dicopper(II) species.156–160 It has been well established that the (-2:2-peroxo)dicopper(II) complex and the bis (-oxo)dicopper(III) complex can interconvert with each other (Scheme 4, path (c)).90,133,161The equilibrium position between the two species in solution has been demonstrated to be significantly influenced by the ligand structure, solvents, and counter anions.90,128,133,161
Theoretical studies have indicated that the thermal stability of the two complexes are very close to each other and that the activation energy of the isomerization reaction is very small.149–152,158 Although solvent and counter anion effects are not clearly understood, the ligand effects can be interpreted by taking into account the steric effect of theN-alkyl substituents as well as the electron donating ability of the supporting ligands. Namely, less hindered substituents allow closer contact of the two copper ions, stabilizing the bis(-oxo)dicopper core (dCuCuẳ 2.8 A˚ vs. dCuCuẳ 3.5 A˚ for the (-2:2-peroxo)dicopper(II) complex), and strongly -donating alkylamine ligands stabilize the higher oxidation state (CuIII) in the bis(-oxo) complexes. It also has been proposed that the ligands that tend to give more tetragonal copper geometries tend to favor the bis(-oxo) complex, while ligands that favor more tetrahedrally distorted geometries favor the peroxo complex.119 Thus, didentate ligands and tridentate alkylamine ligands tend to provide bis(-oxo)dicopper(III) complexes,133,136–139,143–145 while tridentate aromatic ligands predominantly give (-2:2-peroxo) dicopper(II) complexes.41,112–114,120,122–124,126,127Recent studies have shown that subtle changes in the didentate ligands,121,162 tridentate alkylamine ligand,116,119 and tetradentate aromatic amine ligands141result in drastic changes in the structure of Cu2/O2complexes.