Based on a rational ligand design for stabilizing high-valent {Fe(μ-O)2Fe} cores, a new family of dinucleating bis(tetradentate) ligands with varying terminal donor functions has been developed: redox-inert biomimetic carboxylates in H4julia, pyridines in susan, and phenolates in H4hilde(Me2). Based on a retrosynthetic analysis, the ligands were synthesized and used for the preparation of their diferric complexes [(julia){Fe(OH2)(μ-O)Fe(OH2)}]·6H2O, [(julia){Fe(OH2)(μ-O)Fe(OH2)}]·7H2O, [(julia){Fe(DMSO)(μ-O)Fe(DMSO)}]·3DMSO, [(hilde(Me2)){Fe(μ-O)Fe}]·CH2Cl2, [(hilde(Me2)){FeCl}2]·2CH2Cl2, [(susan){FeCl(μ-O)FeCl}]Cl2·2H2O, [(susan){FeCl(μ-O)FeCl0.75(OCH3)0.25}](ClO4)2·0.5MeOH, and [(susan){FeCl(μ-O)FeCl}](ClO4)2·0.5EtOH, which were characterized by single-crystal X-ray diffraction, FTIR, UV-Vis-NIR, Mössbauer, magnetic, and electrochemical measurements. The strongly electron-donating phenolates afford five-coordination, while the carboxylates and pyridines lead to six-coordination. The analysis of the ligand conformations demonstrates a strong flexibility of the ligand backbone in the complexes. The different hydrogen-bonding in the secondary coordination sphere of [(julia){Fe(OH2)(μ-O)Fe(OH2)}] influences the C-O, C[double bond, length as m-dash]O, and Fe-O bond lengths and is reflected in the FTIR spectra. The physical properties of the central {Fe(μ-O)Fe} core (d-d, μ-oxo → Fe(III) CT, νas(Fe-O-Fe), J) are governed by the differences in terminal ligands - Fe(III) bonds: strongly covalent π-donation with phenolates, less covalent π-donation with carboxylates, and π-acceptation with pyridines. Thus, [(susan){FeCl(μ-O)FeCl}](2+) is oxidized at 1.48 V vs. Fc(+)/Fc, which is shifted to 1.14 V vs. Fc(+)/Fc by methanolate substitution, while [(julia){Fe(OH2)(μ-O)Fe(OH2)}] is oxidized ≤1 V vs. Fc(+)/Fc. [(hilde(Me2)){Fe(μ-O)Fe}] is oxidized at 0.36 V vs. Fc(+)/Fc to a phenoxyl radical. The catalytic oxidation of cyclohexane with TONs up to 39.5 and 27.0 for [(susan){FeCl(μ-O)FeCl}](2+) and [(hilde(Me2)){Fe(μ-O)Fe}], respectively, indicates the potential to form oxidizing intermediates.

The reaction of the new dinucleating ligand susan with Fe(BF)·6HO results in formation of the homovalent FeFe complex [(susan){Fe(μ-F)Fe}] and the mixed-valence FeFe complex [(susan){FeF(μ-F)FeF}] depending on the absence or presence of dioxygen, respectively. Complex [(susan){FeF(μ-F)FeF}] is the first molecular mixed-valence complex with a fluorido bridge. The short Fe-μ-F bond of 1.87 Å causes a large reorganization energy, resulting in a localized class II system with an intervalence charge-transfer band of high energy at 10000 cm.

It is widely recognized that enhancement of protein stability is an important biotechnological goal. However, some applications at least, could actually benefit from stability being strongly dependent on a suitable environment variable, in such a way that enhanced stability or decreased stability could be realized as required. In therapeutic applications, for instance, a long shelf-life under storage conditions may be convenient, but a sufficiently fast degradation of the protein after it has performed the planned molecular task in vivo may avoid side effects and toxicity. Undesirable effects associated to high stability are also likely to occur in food-industry applications. Clearly, one fundamental factor involved here is the kinetic stability of the protein, which relates to the time-scale of the irreversible denaturation processes and which is determined to some significant extent by the free-energy barrier for unfolding (the barrier that "separates" the native state from the highly-susceptible-to-irreversible-alterations nonnative states). With an appropriate experimental model, we show that strong environment-dependencies of the thermodynamic and kinetic stabilities can be achieved using robust protein engineering. We use sequence-alignment analysis and simple computational electrostatics to design stabilizing and destabilizing mutations, the latter introducing interactions between like charges which are screened out at high salt. Our design procedures lead naturally to mutating regions which are mostly unstructured in the transition state for unfolding. As a result, the large salt effect on the thermodynamic stability of our consensus plus charge-reversal variant translates into dramatic changes in the time-scale associated to the unfolding barrier: from the order of years at high salt to the order of days at low salt. Certainly, large changes in salt concentration are not expected to occur in biological systems in vivo. Hence, proteins with strong salt-dependencies of the thermodynamic and kinetic stabilities are more likely to be of use in those cases in which high-stability is required only under storage conditions. A plausible scenario is that inclusion of high salt in liquid formulations will contribute to a long protein shelf-life, while the lower salt concentration under the conditions of the application will help prevent the side effects associated with high-stability which may potentially arise in some therapeutic and food-industry applications. From a more general viewpoint, this work shows that consensus engineering and electrostatic engineering can be readily combined and clarifies relevant aspects of the relation between thermodynamic stability and kinetic stability in proteins.

The dinucleating bis(tetradentate) ligand susan [= 4,7-dimethyl-1,