# 7.10: Multisite Redox Enzymes (Part 6)

## N2 and Related Complexes

The triple bond of N2 has one $$\sigma$$ and two $$\pi$$components. Each nitrogen atom has a lone pair oriented along the N-N direction. The two lone pairs allow N2 to bind in an end-on fashion in either a terminal or a bridging mode. Both modes of binding are illustrated in the binuclear zirconium complex333 shown in Figure 7.34. In this and in many other N2 complexes, the N-N bond is not significantly lengthened and is therefore presumed to be insignificantly weakened in the complex. Interestingly, the complex in Figure 7.34, despite not having long N-N distances, forms hydrazine quantitatively upon protonation. Only one of the three N2 molecules is reduced, and all four electrons required come from the two Zr(III) by presumed internal electron transfer. The related $$\mu])-N2 complex [W(\(\eta^{5}$$-C5Me5)Me2(SC6H2Me3)]2($$\mu$$-N2) is one of the few dinitrogen complexes to contain an S donor ligand.333a

In addition to the N lone pairs, the $$\pi$$ components of the N$$\equiv$$N triple bond can serve as donor-acceptor orbitals in the Dewar-Chatt-Duncanson (olefin binding) manner. This less-common mode of N2 binding is illustrated by the structure of the Ti complex334 shown in Figure 7.35. Here, as in the few other known side-on bound N2 complexes,335 the N-N bond is significantly lengthened. The lengthened bond at 1.30 Å is presumed to be sufficiently weakened [v(N-N) = 1280 cm-1] that it is susceptible to further lengthening and reduction. As the N-N distance lengthens, it is more appropriate to consider the ligand as a deprotonated diimide or hydrazine.

Complexes that have proven particularly useful are bis(dinitrogen)phosphines of Mo(0) and W(0) such as M(N2)2(Ph2PCH2CH2PPh2)2 and M(N2)2(PPh2Me)4. As shown in Figure 7.36, treatment of the complexes336-337a with acid leads to the formation of the diazenido(-H) and hydrazido(-2H) complexes, and sometimes to the production of ammonia. The finding of a bound N2H22- species is consistent with the proposed presence of similar bound species in nitrogenase. The complexes of reduced dinitrogen intermediates are stabilized by multiple M-N binding. Further protonation of these intermediates or treatment of the original complex with strong acid leads to the formation of NH3 from the bound nitrogen. Here the Mo(0) starting complex has enough electrons [six from the Mo(0) → Mo(VI) conversion] to reduce one N2 molecule in conjunction with its protonation from the external solution.

In a general sense this reaction may be telling us something about nitrogenase. The enzyme may be able to deliver six reducing equivalents to N2, and protonation, perhaps carefully orchestrated by neighboring amino-acid or homocitrate groupings, may facilitate the process. However, it is virtually certain that the Mo in nitrogenase is not able to change its oxidation state by six units. In the enzyme the multimetal, multisulfur FeMoco site may serve the equivalent function, by providing multiple sites at which reduced intermediates can simultaneously bind.

Only a few of the known N2 complexes contain S-donor ligands. One of these, Mo(N2)2(S(CH2C(CH3)2CH2S)3), shown in Figure 7.37, has four thioether S-donor atoms bound to Mo(0). This Mo(0) complex shows reactivity reminiscent of the related phosphine complexes.337a A remarkable complex (Figure 7.38) has been isolated338 in which two lone pairs of trans-diimide bind to two Fe, concomitantly with H-binding of the two diimide hydrogen atoms to coordinated sulfur atoms. The ability of an Fe-S system to stabilize the very reactive trans-N2H2 grouping adds support to the notion that similar metal-sulfide sites of nitrogenase may stabilize related intermediates along the N2 → 2NH3 reaction path.

Most of the model systems involving N2 do not lead to NH3 formation. Moreover, many systems that do form NH3 are not catalytic. However, certain V-based and Mo-based systems can catalytically reduce N2 to N2H4 or NH3 using strong reducing agents.339 Although kinetic studies indicate the possibility of intermediates, little structural information is available at present on these interesting systems.

## Insights from Relevant Inorganic Reactivity

Certain studies on inorganic systems that do not model the nitrogen-fixation process can nevertheless potentially give insight into nitrogenase action. Two categories of relevant chemistry are acetylene binding/reactivity and dihydrogen binding/activation. Modes of dihydrogen activation on sulfide systems have previously been discussed in the section on hydrogenase.

Acetylene has long been known to bind to metal centers using its $$\pi$$and $$\pi$$* orbitals as, respectively, $$\sigma$$-donor and $$\pi)-acceptor orbitals. Even when the metal is predominantly sulfur-coordinated,340,341 such side-on bonding of RC2R is well known340,341 as in MoO(S2CNR2)2(RC\(\equiv$$CR) and Mo(S2CNR2)2(RC$$\equiv$$CR)2. The direct interaction of acetylene with the metal center must be considered as a potential binding mode for nitrogenase substrates.

A totally different, sulfur-based mode of acetylene binding is now also well established. For example, (Cp')2Mo2S4 reacts with acetylene342,225 to produce

$$\tag{7.18}$$

containing a bridging ethylene-1,2-dithiolate (dithiolene). The acetylene binds directly to the sulfur atoms by forming S—C bonds. Acetylenes or substituted (activated) acetylenes are able to displace ethylene from bridging or terminal 1,2-dithiolate ligands225,341 to produce the 1,2-dithiolenes. In these reactions the sulfur rather than the metal sites of the cluster are reactive toward these small unsaturated molecules. Clearly, for nitrogenase, where we do not know the mode of binding, sulfur coordination might be a viable possibility. The (Cp')2Mo2S4 systems that bind H2 and C2H2, wherein bound C2H2 can be reduced to C2H4 and displaced by C2H2, are potential models for substrate reduction by nitrogenase.225,342

The versatility of transition-metal sulfur systems is further illustrated by the observation that activated acetylene can insert into a metal-sulfur bond in Mo2O2S2(S2)22-, forming a vinyl-disulfide-chelating ligand

It has recently been suggested343 that the presence of a dihydrogen complex is required for H2 to be displaced by N2 to form a dinitrogen complex. This reaction would explain the required stoichiometry of N2 reduction and H2 evolution. Such an explanation had been suggested previously with dihydride complexes acting as the N2-binding and N2-displacing site.182 Clearly, this new suggestion is an interesting embellishment of potential N2/H2 relationships.

At present, the activation process that is at work in the enzyme is unknown. We need greater structural definition of the active site, which should be forthcoming through the continued application of sophisticated diffraction and spectroscopic probes. Diffraction alone, however, will be incapable of locating protons and possibly other low-molecular-weight ligands. Therefore, spectroscopic probes such as ENDOR10 and ESEEM,277-Z79,344 which are based on EPR spectroscopy, and x-ray-based techniques, such as EXAFS and XANES, will remain crucial in elucidating mechanistically significant structural details.