Summary and conclusions

We have performed Car-Parrinello molecular dynamics to identify the oxidative species of the Fenton reagent in water. Starting from different initial conditions, we observed in two molecular dynamics simulations, the spontaneous formation of the contested ferryl ion, which confirms the model first proposed by Bray and Gorin, and agrees with the overall energetics obtained for the reactants in vacuo.

Starting from the pentaaqua iron(II) hydrogen peroxide complex in aqueous solution, the oxygen-oxygen bond of the H$_2$O$_2$ ligand cleaves almost immediately to form a pentaaqua iron(III) hydroxo complex and an OH. radical. The OH. radical immediately abstracts either directly, or via one or two solvent water molecules, a hydrogen of a water ligand to form tetraaqua dihydroxo iron(IV) and a water molecule. This is in agreement with the formation of the OH. radical being energetically unfavorable. So, the oxygen-oxygen dissociation is made energetically possible, because a fast transfer along an H-bond wire through the solvent to a low energy end product is found by the OH.. The dihydroxo iron(IV) ion was found to be a meta-stable complex which in our first simulation transformed into the ferryl ion, again in agreement with the relative energies in the gas phase.

Starting from artificially separated reactants (i.e. the hydrogen peroxide and the pentaaqua iron(II) complex with a vacant coordination site), we simulated another possible reactive pathway. We found the coordination process to be followed by spontaneous reaction to again the ferryl ion and a water molecule. In this pathway, the meta-stable intermediate of coordinated Fe $^{\mathrm{II}}$-H$_2$O$_2$ as proposed in the literature was not formed, but instead immediate dissociation of the oxygen-oxygen bond took place, as soon as the reactants were close enough to each other. In contrast with the two step mechanism found earlier via a dihydroxo iron(IV) intermediate, more direct formation of the iron(IV) oxo ion took place via hydrogen abstraction by OH. from Fe $^{\mathrm{III}}$-OH$^-$, soon after the oxygen-oxygen cleavage. In this mechanism the energy needed to form the OH. radical can be accounted for by the energy gain of the Fe $^{\mathrm{II}}$-H$_2$O$_2$ bond formation.

Our simulations disfavor but do not rule out completely the Haber and Weiss OH. radical mechanism (which is, especially in biochemistry, often taken as synonymous to Fenton chemistry). In the initial step of the iron catalyzed hydrogen peroxide dissociation, always first a very short-lived OH. radical appears, and the L-Fe $^{\mathrm{III}}$-OH$^-$ complex. However, this radical has no independent existence, it abstracts a hydrogen either immediately or in a short transfer via one or two solvent molecules from a water ligand to form a dihydroxo iron(IV) complex, or even directly from the OH ligand to form the ferryl ion; in both cases neutralizing itself to a water molecule. Also when other ligands than water molecules are used, such as chelating agents, the radical may scavenge these ligands. The degradation of chelating agents, limiting the number of catalytic cycles one complex can undergo, is a notorious phenomenon in Fenton chemistry.