One topic of great interest in inorganic chemistry is N₂ reduction to NH₃ ("nitrogen fixation"). This reaction is necessary for life as we know it, because nitrogen fixation is the major route through which nitrogen from the atmosphere becomes available for organisms (nitrogen is an integral component of DNA, RNA, and proteins, for example). Nitrogen reduction is catalyzed in nature by the "iron-molybdenum cofactor" of iron-molybdenum nitrogenase (illustration above; PDB; QuickTime movie). The iron-molybdenum cofactor has six iron atoms that have an unusual trigonal pyramidal geometry. We hypothesize that the low-coordinate iron is important to perform this difficult six-electron reduction.

We envision a mechanism for N₂ binding in which reduction of the cofactor weakens or breaks a bond between the central ion and a trigonal-pyramidal iron atom. The resulting three-coordinate iron may bind nitrogen, or may be protonated to form a hydride complex that subsequently reacts with nitrogenase substrates. In research that evaluates this idea, we have created soluble complexes containing low-coordinate iron atoms that reduce the N-N bond order of N₂ from three to two, while other known iron-N₂ complexes have N-N triple bonds. (Click here for animated GIF's of a neutral dinitrogen complex and a potassium-complexed dinitrogen complex.) Mössbauer and computational studies show that in neutral FeNNFe compounds, the N₂ unit is a triplet dianion N₂^2-, antiferromagnetically coupled to two high-spin Fe²⁺ ions. The low-coordinate iron (formally Fe^I) transfers one electron per iron to the ligand, in the N₂ complexes and in complexes of alkynes. The amount of electron transfer is highly metal-dependent, with more N₂ reduction in iron complexes, and less in cobalt and nickel complexes. Our work has been crucial in establishing these periodic trends. Current work is aimed at utilizing these insights to achieve reactivity of alkynes and N₂.

In other work, we have created synthetic complexes in which two reactive iron atoms are bridged by a sulfide; the Fe-S-Fe unit is reminiscent of the diiron units of the FeMoco shown above. We have found that the two iron atoms cooperatively break the N-N bond of certain hydrazines (see illustration below). Therefore, we are continuing to create other ligand environments that encourage cooperative reactions of multiple iron atoms.