Back to Table of Contents

47051-G3
New Chemistry of Iron in its Highest Oxidation State

John Ferguson Berry, University of Wisconsin (Madison)

Normal 0 false false false MicrosoftInternetExplorer4

Introduction

The Berry Group is investigating the synthesis, characterization, electronic structure, and chemistry of high-valent transition metal complexes that represent key reactive intermediates in the catalytic oxidation of organic substrates.  One synthetic method that we are using in order to stabilize highly reactive metal complexes in high oxidation states (for example, Fe(VI)) is to bind the metal atom with a polydentate, polyanionic ligand.  The high negative charge on these ligands should serve to stabilize a high positive charge on the metal center, allowing for high-valent species to be observed.  Tris-amidoamine ligands ([N(CH₂CH₂NR)₃]^3–, abbreviated [^RN₃N]^3–, R = alkyl, aryl) constitute a general class of polyanionic ligands that we have investigated.  Inspired by the synthesis of an Fe(IV)-cyano complex using an [^R3SiN₃N]^3– ligand by Schrock (Cummins, C. C.; Schrock, R. R., Inorg. Chem. 1994, 33, 395), we decided to investigate high-valent iron chemistry of [^RN₃N]^3– ligands with R = neopentyl (NP) or xylyl (XY).  A particularly intriguing initial synthetic target was an Fe(VI)-nitrido ^RN₃N complex, since only one Fe(VI)-nitrido species had ever been synthesized before (Berry, J. F.; Bill, E.; Bothe, E.; George, S. D.; Mienert, B.; Neese, F.; Wieghardt, K., Science 2006, 312, 1937).  We quickly found, much to our surprise, that the ^RN₃N ligands are redox non-innocent, regardless of whether NP or XY substituents are used.  Therefore, in addition to the redox capabilities of the iron center, an Fe(^RN₃N) complex can exhibit additional redox activity and additional redox states derived from the oxidation of the ^RN₃N ligand.  This newly discovered feature of the Fe(^RN₃N) system will allow us to use these compounds to promote multi-electron transfer reactions at a single Fe center.

Results & Discussion

1. Synthesis

The ligands H₃^NPN₃N (Kisanga, P. B.; Verkade, J. G. Tetrahedron, 2001, 57, 467) and H₃^XYN₃N (Greco, G. E.; Schrock, R. R. Inorg. Chem. 2001, 40, 3850) were synthesized according to previously reported methods.  The lithiated ligand Li₃(^XYN₃N) was added to FeCl₃·THF producing an air-sensitive deep blue solution, from which a deep blue crystalline compound can be isolated.  Mass spectrometry has shown the composition of this compound to be Fe(^XYN₃N).  We were initially puzzled by the intense blue color of this compound because we expected a high-spin Fe(III) species to be colorless due to a ⁶A ground state.  We therefore synthesized the analogous Ga(^XYN₃N) complex to make sure that the color that we observed was not a property of the ligand itself.  Lithiated [^XYN₃N]^3– was added to GaI₃, yielding a deliquescent, colorless crystalline compound that was identified by mass spectrometry as Ga(^XYN₃N). 

2. Spectroscopy & Electronic Structure

While the Ga complex is colorless, the “Fe(III)” complex Fe(^XYN₃N) shows an extremely intense absorption in the visible region centered at ~ 650 nm.  We propose that this absorption at 650 nm is due to an internal electron tranfer in the Fe(^XYN₃N) complex that signifies the presence of a coordinated aminyl radical in the compound.  We suggest therefore that, instead of the nominal [(Fe)³⁺(^XYN₃N)^3–] oxidation state assignment, that one of the amido arms of the N₃N ligand is oxidized to an aminyl radical and that the complex should consequently be assigned as [(Fe)²⁺(^XYN₃N)^•,2–].  Since all three amido arms of the ligand may equally accommodate a radical, the ligand must exist in an {N^–, N’^–, N”^•} mixed-valence form.  Absorption of light at 650 nm likely involves an intra-ligand intervalence charge transfer (IVCT) that may be mediated by the Fe(II) center.

The electronic structure assigned above is consistent with electrochemical measurements that we have made on the free ligands H₃^XYN₃N as well as H₃^NPN₃N.  Both ligands are irreversibly oxidized by multiple electrons at potentials > 450 mV vs ferrocene/ferricinium.  We suggest that the ^XYN₃N ligand is oxidized upon its addition to FeCl₃·THF (a rather strong one-electron oxidant!), forming the above-mentioned Fe(II)-radical complex.  This proposal is supported by the preparation of the corresponding Ga(III) complex that is formed via substitution of the I^– ligands from GaI₃; no redox chemistry is possible with Ga(III), so the colorless Ga(^XYN₃N) species is observed as the product of the reaction.  We are currently investigating direct metallation of ^XYN₃N with an Fe(II) source.

Summary.

We have discovered that N₃N ligands are redox non-innocent, and that their complexes with transition metals such as iron may undergo multiple electron transfer reactions that are either metal-based or ligand-based. 

Back to top