Rates of water exchange on the [Fe4(OH)2(hpdta)2(H2O)4]0 molecule and its implications for geochemistry

The ammonium salt of [Fe(4)O(OH)(hpdta)(2)(H(2)O)(4)](-) is soluble and makes a monospecific solution of [Fe(4)(OH)(2)(hpdta)(2)(H(2)O)(4)](0)(aq) in acidic solutions (hpdta = 2-hydroxypropane-1,3-diamino-N,N,N',N'-tetraacetate). This tetramer is a diprotic acid with pK(a)(1) estimated at 5.7 ± 0.2 and pK(a)(2) = 8.8(5) ± 0.2. In the pH region below pK(a)(1), the molecule is stable in solution and (17)O NMR line widths can be interpreted using the Swift-Connick equations to acquire rates of ligand substitution at the four isolated bound water sites. Averaging five measurements at pH < 5, where contribution from the less-reactive conjugate base are minimal, we estimate: k(ex)(298) = 8.1 (±2.6) × 10(5) s(-1), ΔH(++) = 46 (±4.6) kJ mol(-1), ΔS(++) = 22 (±18) J mol(-1) K(-1), and ΔV(++) = +1.85 (±0.2) cm(3) mol(-1) for waters bound to the fully protonated, neutral molecule. Regressing the experimental rate coefficients versus 1/[H(+)] to account for the small pH variation in rate yields a similar value of k(ex)(298) = 8.3 (±0.8) × 10(5) s(-1). These rates are ∼10(4) times faster than those of the [Fe(OH(2))(6)](3+) ion (k(ex)(298) = 1.6 × 10(2) s(-1)) but are about an order of magnitude slower than other studied aminocarboxylate complexes, although these complexes have seven-coordinated Fe(III), not six as in the [Fe(4)(OH)(2)(hpdta)(2)(H(2)O)(4)](0)(aq) molecule. As pH approaches pK(a1), the rates decrease and a compensatory relation is evident between the experimental ΔH(++) and ΔS(++) values. Such variation cannot be caused by enthalpy from the deprotonation reaction and is not well understood. A correlation between bond lengths and the logarithm of k(ex)(298) is geochemically important because it could be used to estimate rate coefficients for geochemical materials for which only DFT calculations are possible. This molecule is the only neutral, oxo-bridged Fe(III) multimer for which rate data are available.