ARTICLE pubs.acs.org/est
Heterogeneous Reduction of PuO2 with Fe(II): Importance of the Fe(III) Reaction Product Andrew R. Felmy,†,* Dean A. Moore,† Kevin M. Rosso,† Odeta Qafoku,† Dhanpat Rai,‡ Edgar C. Buck,† and Eugene S. Ilton† † ‡
Pacific Northwest National Laboratory, Richland, Washington, United States Rai Enviro-Chem, LLC, Yachats, Oregon, United States
bS Supporting Information ABSTRACT: Heterogeneous reduction of actinides in higher, more soluble oxidation states to lower, more insoluble oxidation states by reductants such as Fe(II) has been the subject of intensive study for more than two decades. However, Fe(II)-induced reduction of sparingly soluble Pu(IV) to the more soluble lower oxidation state Pu(III) has been much less studied, even though such reactions can potentially increase the mobility of Pu in the subsurface. Thermodynamic calculations are presented that show how differences in the free energy of various possible solid-phase Fe(III) reaction products can greatly influence aqueous Pu(III) concentrations resulting from reduction of PuO2(am) by Fe(II). We present the first experimental evidence that reduction of PuO2(am) to Pu(III) by Fe(II) was enhanced when the Fe(III) mineral goethite was spiked into the reaction. The effect of goethite on reduction of Pu(IV) was demonstrated by measuring the time dependence of total aqueous Pu concentration, its oxidation state, and system pe/pH. We also re-evaluated established protocols for determining Pu(III) {[Pu(III) þ Pu(IV)] Pu(IV)} by using thenoyltrifluoroacetone (TTA) in toluene extractions; the study showed that it is important to eliminate dissolved oxygen from the TTA solutions for accurate determinations. More broadly, this study highlights the importance of the Fe(III) reaction product in actinide reduction rate and extent by Fe(II).
’ INTRODUCTION The heterogeneous reduction of redox active actinides from their soluble and potentially environmentally mobile oxidized forms [i.e., U(VI), Np(V), and Pu(V)] to their reduced and sparingly soluble tetravalent state by different chemical forms of Fe(II) has been the subject of intensive research for almost two decades.121 In particular, it has been recognized that apparent adsorption or structural incorporation of Fe(II) facilitates reduction of actinides and pertechnetate compared to aqueous Fe(II) alone.5,8,10,15,17,2225 For example, Liger et al.15 demonstrated that aqueous Fe(II) was ineffective as a reductant for U(VI) but that introduction of the iron(III) oxide hematite resulted in rapid reduction of U(VI) to sparingly soluble U(IV). Interpretations of such results typically have emphasized the strong effect of Fe(II) adsorption on its apparent reduction potential. Tacitly implied yet not often stated or recognized in such a conceptual model is the assumption that μaqFe(II) 6¼ μadFe(II), where μ represents the total chemical potential. Here we present another perspective, one that hypothesizes that Fe(II) association with iron(III) oxide surfaces facilitates the formation of specific iron oxide reaction products that intrinsically have a much lower free energy than those that would form in the absence of the initial iron(III) oxide. For example, Peretyazhko et al.22 found that the reduction of Tc(VII) by Fe(II) adsorbed onto iron oxides was orders of r 2011 American Chemical Society
magnitude faster than for Fe(II) adsorbed onto different phyllosilicate surfaces. This observation suggests that coordination of Fe(II) to the iron oxide surface creates reaction pathways that are not readily accessible in homogeneous systems or in heterogeneous reactions involving non-iron oxide surfaces. The opening up of mechanistic pathways to form these lower free energy Fe(III) reaction products could be a key factor in determining the extent and kinetics of redox reactions involving Fe(II). In the present study, we evaluate this hypothesis for the reduction of sparingly soluble Pu(IV) to soluble Pu(III) by Fe(II), a heterogeneous interfacial electron transfer system involving in principle Fe(II) contact with solid-phase Pu(IV). Indeed, thermodynamic calculations presented herein show that the chemical form of the resulting Fe(III) reaction product should be critically important in determining the reaction extent, monitored here as the time-dependent aqueous concentration of Pu(III). For example, given reduction of PuO2(am) by Fe(II)aq to form two-line ferrihydrite (2LF) or various common FeOOH polymorphs, Received: December 16, 2010 Accepted: March 28, 2011 Revised: March 11, 2011 Published: April 06, 2011 3952
dx.doi.org/10.1021/es104212g | Environ. Sci. Technol. 2011, 45, 3952–3958
Environmental Science & Technology
ARTICLE
Figure 1. Calculated Pu(III) concentrations (—) in equilibrium with PuO2(am) and with different iron oxide reaction products. Initial aqueous Fe(II) = 1 mM.
PuO2 ðamÞ þ Fe2þ þ Hþ þ H2 O T Pu3þ þ FeðOHÞ3 ð2LFÞ ð1Þ PuO2 ðamÞ þ Fe2þ þ Hþ T Pu3þ þ FeOOH
ð2Þ
the predicted concentrations of Pu(III) in aqueous solution at equilibrium would be orders of magnitude higher if the final Fe(III) reaction product was goethite rather than either 2LF or lepidocrocite (Figure 1). The impact of such surface-induced transformations has never been demonstrated for the reduction of Pu(IV). Although thermodynamics clearly predicts that the nature of the Fe(III) reaction product is critical for determining the extent of PuO2(am) reduction by Fe(II), it is not known, unlike for U(VI) and Tc(VII), whether the presence of an Fe(III) solid phase will facilitate Pu(IV) reduction by Fe(II). In this study we compared the reduction of PuO2(am) by Fe(II)aq in the absence and presence of goethite (R-FeOOH). As a first step, we prepared suspensions containing PuO2(am) and aqueous Fe(II) at the same concentrations previously used by Rai et al.26 to establish a baseline of Pu(III) concentrations and to ensure agreement with previous studies. After establishment of this baseline, goethite was added to selected samples to test for enhanced formation of Pu(III).
’ EXPERIMENTAL PROCEDURES Materials and Methods. The materials and methods for preparation of the PuO2(am) and Fe(II) solutions are similar to those used by Rai et al.26 and are described briefly here. A 106 g/L Pu stock solution consisting of 93% Pu239 and 7% Pu240 was prepared in 8 M HNO3. UVvis spectra of the solution showed only Pu(IV). Stock solutions of 0.25 M KBrO3 and 0.6 M FeCl2 in 0.001 M HCl were prepared by dissolving solid KBrO3 and FeCl2 3 4H2O in distilled deionized (DDI) water, respectively. A 0.5 M thenoyltrifluoroacetone (TTA) stock solution in toluene was prepared for use in oxidation-state analysis and stored in an amber bottle.
All experiments were conducted at room temperature (23 ( 2 °C) in a controlled atmosphere chamber of prepurified Ar (99.99%) with
