The L1 and L3 complexes in provide two very different phenolate ring orientations due to the different steric requirements of iso-the propyl and tert-butyl substituents on the pyrazole ligand. These lead to very different frontier molecular orbitals. L1 has the oop phenolate p-orbital in a π bonding interaction with the Cu dx2-y2 orbital, while L3 has the ip phenolate p-orbital in a pseudo-σ type interaction with the β LUMO (). The different LUMOs of the two complexes lead to their very different spectroscopic properties. L1 has a smaller splitting between the ip and the oop phenolate CT transitions with a low energy intense oop phenolate CT band. Alternatively, L3 has greater overlap of the ip phenolate π with the dx2-y2 orbital and a larger energy splitting between the ip and the oop phenolate CT bands, having the intense ip phenolate CT transition at a higher energy ().
These differences in orientation lead to significant differences in bonding. The spin density on these complexes shows that L3 has less radical character (9%) compared with L1 (14%), though the amount of radical character in both is small compared to a reference Zn phenoxyl analogue (74-77%, ). The O2 reaction with L1 CuIIphenolate forming a CuIsuperoxoquinone species is ~29 kcal/mol more favored than the L3 analogue () because of the presence of the lower energy phenolate LMCT transition in L1. Also the O2 reaction of CuIIphenolate forming a CuIIperoxoquinone species is ~26 kcal/mol more favorable for L1 than L3 (). Thus the L1 ring orientation is the preferred orientation for O2 attack.
Importantly, the reaction of O2 with L1 is not likely to proceed via a radical mechanism, since the amount of radical character in the ring is small. The energy for the radical reaction generating the superoxoquinone is high (51 kcal/mol), reflecting the energy of the associated LMCT transition. However, the energy of O2 attack for L1 CuIIphenolate to form a coordinated peroxoquinone is significantly lower (11.3 kcal/mol) than for the radical reaction, and in fact for this reaction on the corresponding ZnII phenolate complex or phenol (). The latter is due to the presence of an unoccupied dx2-y2 orbital in CuII but not in d10 ZnII. This orbital has good σ bonding interactions with the peroxy and O donor orbitals of the quinone, providing stabilization to form the bridging peroxoquinone.
The uncatalyzed O2
reaction of organic substrates, like phenols, are kinetically slow because of their spin forbidden nature (O2
, S= 1; phenol, S= 0). It has been shown that an oxidized metal center can act as a buffer to catalyze the spin forbidden O2
reaction with a coordinated substrate via a low energy LMCT transition.19
This model applied to L1 is depicted in , where the triplet O2
is antiferromagnetically aligned with the CuII
(S = ½) due to orbital overlap along the reaction coordinate. The two electron reduction of triplet O2
from singlet phenolate could proceed by transferring one electron (spin up) from the phenolate directly onto the O2
orbital which is σ bonded to the carbon, and another electron (spin up) from CuII
through a π interaction with the triplet O2
. The latter could be compensated by transferring an electron (spin down) from the phenolate ligand to CuII
via the experimentally observed low energy intense LMCT (14730 cm-1
band for L1) pathway. This results in a two electron oxidation of a phenolate by triplet O2
without affecting the oxidation state of the CuII
. This mechanism does not require radical character in the ring and simply has the metal playing the role of a buffer in transferring an electron from the substrate to O2
but of proper spin.
Schematic representation of electron transfer in the O2 reaction of L1CuIIphenolate forming the CuIIperoxoquinone.
The generally proposed mechanism for the cofactor biogenesis of TPQ from tyrosine in amine oxidases invokes O2
attack on a transient CuI
tyrosyl radical species.44,45
The present study on model systems indicates that the O2
attack on the CuII
phenolate yielding a radical CuI
superoxyquinone is in fact, energetically unfavorable (51 kcal/mol). Alternatively, the reaction free energy for O2
attack on CuII
phenolate generating a metallocycle CuII
peroxoquinone species is only 11.3 kcal/mol which is comparable to the experimental barrier of 15 kcal/mol for TPQ formation (78 M-1
at high pH).2
In contrast to the radical reaction, this is spin forbidden. However as described above, a LMCT process can overcome this, and indeed an absorbance band is observed along the reaction coordinate of O2
These and other possible mechanisms for O2
activation by a phenolate/tyrosine ring (e.g. H atom abstraction of the phenol proton by a CuII
site) need to be evaluated for the actual enzyme active site. We finally note that L1 has the more preferred ring orientation for O2
attack and its ring orientation reflects the orientation of the tyrosine ring in the preprocessed active site of amine oxidase ().