- Email: [email protected]
The use of a high-FeO olivine rock as a redox buffer in a nuclear waste repository
Journal of Contaminant Hydrology 83 (2006) 42 – 52 www.elsevier.com/locate/jconhyd
The use of a high-FeO olivine rock as a redox buffer in a nuclear waste repository J. Gime´nez a,*, M. Rovira b, F. Clarens a, I. Casas a, L. Duro a, M. Grive´ c, J. Bruno c, J. de Pablo a,b a
Department of Chemical Eng. (ETSEIB-UPC), Av. Diagonal 647, 08028 Barcelona, Spain b CTM Centre Tecnolo`gic, Avda. Bases de Manresa, 1. 08240 Manresa, Spain c ENVIROS-SPAIN S.L. Passeig de Rubı´, 29-31, 08197 Valldoreix, Spain
Received 24 February 2005; received in revised form 20 October 2005; accepted 30 October 2005 Available online 13 December 2005
Abstract Due to the higher stability of the spent nuclear fuel (mainly composed of UO2) under reducing conditions, and in order to enhance the retention/retardation of some key radionuclides, the olivine rock from the Lovasja¨rvi intrusion has been proposed as a potential redox-active backfill-additive in deep highlevel nuclear waste (HLNW) repositories. In this work, two different approaches have been undertaken in order to establish the redox buffer capacity of olivine rock: (1) The capacity of the rock to respond to changes in pH or pe has been demonstrated and the final (pH, pe) coordinates agree with the control exerted by the system Fe(II)/Fe(III). (2) The rate of consumption of oxygen has been determined at different pH values. These rates are higher than the ones reported in the literature for other solids, what would point to the possibility of using this rock as an additive to the backfill material in a HLNW. D 2005 Elsevier B.V. All rights reserved. Keywords: High-level nuclear waste; Radionuclide migration; Olivine; Redox buffer capacity
1. Introduction The aim of the different engineered barriers for the long term disposal of high level nuclear waste (HLNW) is to avoid or to retard the dissemination of the radionuclides released from the spent nuclear fuel to the environment. The spent nuclear fuel is mainly composed of UO2 and its * Corresponding author. Departament Enginyeria Quı´mica, H-4 ETSEIB, Universitat Polite´cnica de Catalunya (UPC), Avda. Diagonal 647, 08028-Barcelona, Spain. Tel.: +34 934017388; fax: +34 934015814. E-mail address: [email protected] (J. Gime´nez). 0169-7722/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jconhyd.2005.10.011
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
43
corrosion increases with the redox potential, because uranium(VI) species are more mobile than the uranium(IV) ones (Shoesmith, 2000; de Pablo et al., 1999). The groundwaters that can arrive to the spent fuel are usually reducing but certain processes as water radiolysis or glacial meltwater intrusion can generate oxidizing species in the near-field of the fuel (Eriksen et al., 1995). For this reason, in the design of the repository, the function of the buffer material is to comprise a protective zone with a suitable environment around the canister (SKB91, 1992) and, in this context, the addition of chemical getters is suggested as a potential optimization of current designs of backfill material which are based upon the use of bentonite clay. The high-FeO olivine-rich rock from Lovasja¨rvi intrusion has been proposed as a potential redox-active backfill additive in deep nuclear waste repositories (Hellmuth, 1991; Hellmuth et al., 1992). In this sense, the main characteristics of this olivine rock are related to: (1) its ability to retain radionuclides dissolved from the nuclear waste, (2) its thermodynamic and kinetic stability, and (3) its redox buffer capacity. The retention capacity of the Lovasja¨rvi olivine rock is very important since it may control the migration of some significant radionuclides coming from the nuclear waste proximity. In fact, good sorption capacity of this rock has been found for some trace elements (Hellmuth et al., 1994; Rodrigues et al., 1998; Suksi et al., 1998; El Aamrani et al., 2002) and the uranium(VI) transport through columns filled with this rock was studied by our group (Rovira et al., 2000). The kinetics of dissolution of the olivine rock from Lovasja¨rvi has been recently studied as a function of pH in our department (Duro et al., 2005) and a similar behaviour in front of pH to that of the forsterite (Mg2SiO4) has been found. At acidic pH, the reaction order with respect to the proton concentration was 0.5, which agrees with values found in the literature for forsterite dissolution (Blum and Lasaga, 1988; Westrich et al., 1993; Wogelius and Walther, 1991, 1992; Pokrovsky and Schott, 2000a,b). At alkaline pH, considering the Si-based dissolution rates, the reaction order would be 0.3, similar to the reaction order obtained by Blum and Lasaga (1988) and Wogelius and Walther (1991). However, considering Mg and Fe data, the dissolution rates would not depend on the proton concentration. Both possibilities are discussed in Duro et al. (2005). In this work, we have studied the redox buffer capacity of the olivine rock, because it was thought that this capacity would be high considering its content on iron(II) present mainly in olivine (Mg,Fe)SiO4 and in less extent in magnetite (Fe3O4) mineral that is also present in the rock (see Table 1). Given that the buffering of the redox potential seems to be plausible, it is of the utmost importance the determination of the rate at which this olivine rock can consume a given amount of oxidants. In this sense, we have studied (1) the capacity of the olivine rock to respond to changes in pH and E h, and (2) the rate of oxygen consumption of the rock. The oxygen uptake capacity of the rock has also been compared to other solids that can be used as a redox buffer. 2. Experimental The mineral content of the high-FeO olivine-rich rock used in this study was: 65% of olivine (with 57% of forsterite, Mg2SiO4, and 43% of fayalite, Fe2SiO4), 20% of plagioclase (with 50% of anortite, CaAl2Si3O8, and 50% of albite, NaAlSi3O8), 8% of magnetite (Fe3O4) and 7% of pyroxene and serpentine. The chemical composition of the olivine rock is shown in Table 1 (Hellmuth et al., 1992). The solid was crushed and sieved to a 120–160 Am particle size, the specific surface area of the solid was determined to be 0.17 F 0.02 m2 g 1 by the BET method using a gas mixture of 30% N2/Ar.
44
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
Table 1 Composition of the olivine rock Wt.% SiO2 MgO FeO Fe2O3 MnO Cr2O3 Al2O3 TiO2 V2O5 CaO K2O Na2O NiO CoO P2O5 H2O
34.80 20.16 31.58 4.24 0.34 0.07 2.43 3.66 0.09 1.39 0.33 0.56 0.06 0.05 0.12 0.38
All the experiments were performed at 25.0 F 0.1 8C. Two different kinds of experiments were carried out in order to determine (1) the pe/pH buffering capacity and (2) the oxygen uptake capacity of the olivine rock, respectively. 2.1. Study of the E h/pH buffering capacity The experiments were performed in two identical batch reactors (I and II) as shown in Fig. 1. 200 cm3 of NaClO4 10 2 mol dm 3 were placed in each reactor and were continuously purged with N2(g), supplied by Air Liquide. The gas stream was bubbled through a solution of Cr(II) in contact with a Zn/Hg amalgam in order to avoid the introduction of oxidants into the system. The solutions were saturated by N2(g) bubbling overnight.
Fig. 1. Batch reactor used in the E h/pH buffering experiments.
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
45
Table 2 Experimental conditions used in pe/pH buffering capacity of olivine-rock tests
Part Part Part Part
A B C D
Time (h)
Solution composition
Final pH
Final pe
0–334 334–695 695–1682 1682–2400
NaClO4 10 2 M +HClO4 +NaOH +O2 (g)
8.2 2.6 6.13 5.0
3.5 10 63 5.5
2.667 g of the olivine rock was placed in reactor I while reactor II was used as a blank. Both reactors were continuously stirred during the tests by means of an orbital stirrer. The pH of the solutions was monitored by means of a calibrated combined-glass electrode. Redox potentials were measured with a platinum electrode, an additional Au redox electrode was introduced in reactor I. Redox potentials were measured against the Ag/AgCl(s) and KCl saturated reference of the combined glass electrode. During the experiments aqueous samples were taken periodically for iron analysis. These samples were immediately filtered through 0.22-Am-pore-size filters and acidified by adding a small volume of concentrated HNO3. Every sample was replaced by an equivalent volume of N2(g)-saturated ionic medium in reactor I, to keep constant the total volume of the solution inside the reactor. Iron was determined by the ferrozine method by means of UV–Vis spectrophotometry. Iron(II) concentration was determined spectrophotometrically in aqueous solution as the iron(II) complex with the disodium salt of ferrozine. Total iron concentration was determined by previous reduction of iron (III) to iron(II) using the same analytical method (Gibbs, 1976). The experiments were divided in four parts, A, B, C and D, depending on the experimental conditions employed, as can be seen in Table 2. These experiments were carried out in this way in order to force each time only one of the master variables (pH or pe) and to see the response of the other one. Once pH and pe reached steady state values, the conditions of the experiment were artificially changed in order to force the response of the olivine rock. Iron was analysed before and after each disturbance.
Fig. 2. Reactor scheme used in the oxygen consumption measurements.
46
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
Table 3 Initial conditions of the experiments of the olivine-rich rock redox buffer capacity study
Exp. Exp. Exp. Exp. Exp. Exp. Exp.
A B C D E F G
pHinitial
Weight of solid (g)
Volume (dm3)
Ionic medium
2.0 4.0 7.9 9.2 6.2 8.4 9.4
4.00 4.67 4.00 4.00 4.67 4.67 4.67
0.300 0.350 0.300 0.300 0.350 0.350 0.350
NaClO4, 0.01 mol dm 3 NaClO4, 0.01 mol dm 3 NaClO4, 0.01 mol dm 3 NaClO4, 0.01 mol dm 3 NaClO4, 0.01 mol dm 3 NaHCO3, 0.01 mol dm 3 NaHCO3, 0.01 mol dm 3
2.2. Study of the oxygen uptake capacity A scheme of the reactor used in this study is shown in Fig. 2. The evolution of the dissolved oxygen concentration with time was monitored by means of a dissolved oxygen meter (Orion Model 850) with a precision of 0.1 ppm.
Fig. 3. Evolution of (a) pH and (b) pe with time in the experiments related to the E h/pH buffer capacity of the olivine-rich rock.
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
47
4.67 g of the 120- to 160-Am-particle-size olivine rock was introduced into the reactor with a solution containing an initial dissolved oxygen concentration of 2.7 10 4 mol dm 3 in 0.01 mol dm 3 NaClO4. During the experiments, the pH was measured by means of a glass electrode. Samples of each experiment were withdrawn for iron quantification as in Section 2.1. A series of blank experiments were also performed in order to take into account the oxygen consumption by the oxygen meter due to the process of the measure itself. The initial conditions of the experiments are summarised in Table 3. 3. Results and discussion 3.1. Study of the E h/pH buffering capacity The evolution of the pH and pe as a function of time for all the experiment is shown in Fig. 3. In order to test whether the iron system is effectively buffering the pe/pH values, we have plotted all the experimental data on predominance diagrams of the iron system (see Fig. 4).
Fig. 4. Evolution of the experimental pe and pH superimposed to a predominance pe/pH diagram of the iron system. Solid arrows stand for the artificial disturbances imposed to the system.
48
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
As we can see, any artificial disturbance introduced in the system (pH decrease in part B, pH increase in part C, and pe increase in part D) induces a response in the system to re-establish the chemical equilibrium between Fe(II) and Fe(III).
Fig. 5. Evolution of the O2(aq) concentration and pH with time.
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
49
We can deduce, then, that the olivine rock is effectively responding to disturbances of either pH or redox potential externally induced. When we add HClO4 (part B), the pH decreases and the system evolves towards an increase in pe. Once the system is again in equilibrium, we add NaOH (part C) and the response of the system is a decrease of the redox potential. Finally, once the equilibrium is reached again we introduce O2(g) in the system in order to increase the redox potential (part D) which is contra arrested by a decrease in the pH of the system. As we can see also in Fig. 4, the final (pH,pe) coordinates are in agreement with the control exerted by the system Fe(II)/Fe(III) and that the precipitation of amorphous Fe(OH)3 solid can explain this behaviour. Therefore, it would be logical to assume that the surface of the sample of olivine rock is coated with this solid, which forms as a result of the oxidation of the structural Fe(II) of the solid, either of the olivine itself or of the Fe(II) contained in the magnetite of the rock. 3.2. Study of the oxygen uptake by the olivine rock In these experiments, we have measured the capacity of the olivine rock to consume oxygen introduced in the reactors. In all experiments, the oxygen concentration decreased with time (the oxygen concentrations measured can be seen in Fig. 5). In addition, the pH varied during the first hours of the experiments until it reached a constant value. We have calculated the rate of oxygen consumption in the different experiments from the decrease of the oxygen concentration in solution with time. The results obtained are shown in Table 4, the oxygen consumption rates have been calculated with the values measured when the pH was constant and are shown in Fig. 6. In this figure, we have also included the data from Duro et al. (2005) as a solid line: the olivine rock dissolution rates obtained based on the iron concentrations measured in solution as well as the dissolution rates obtained by applying a semi-empirical model. From the comparison between both rates, we can see that in the pH range of 6–10, representative of most groundwaters, the oxygen consumption rates are higher than the iron release rates, indicating that the oxygen is being consumed by the structural iron present in the rock either contained in the olivine or in the magnetite, according to the reaction: ½FeðIIÞsurface þ 1=4O2 þ Hþ Z ½FeðIIIÞsurface þ 1=2H2 O: This preferential reaction of the oxygen with the structural iron(II) would agree with the observations made by Wogelius and Walther (1992), who found that in the case of pure olivine dissolution, the release of iron to the solution is always slower than the oxygenation of iron and its subsequent precipitation as iron(III). In addition, studies focused on O2(aq) uptake by ironTable 4 Rates of oxygen consumption
Exp. Exp. Exp. Exp. Exp. Exp. Exp.
A B C D E F G
pHinitial
pHfinal
Rate (mg O2 dm 3 h 1)
2.0 4.0 7.9 9.2 6.1 8.4 9.4
2.5 5.7 6.0 6.0 6.3 8.3 9.3
0.059 F 0.003 0.015 F 0.002 0.009 F 0.001 0.052 F 0.001 0.007 F 0.001 0.023 F 0.003 0.062 F 0.002
50
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
Fig. 6. Experimental rates of oxygen consumption together with the olivine rock dissolution rates and theoretical dissolution rates determined in Duro et al. (2005, see the text).
containing silicates reported by White and Yee (1985) suggest surface oxidation of iron silicates as a significant mechanism for the consumption of O2(aq). In the case of the high-FeO olivine-rich rock studied in our work as previously stated, we have to consider also the presence in the solid of a 8% of magnetite, which would have even a higher oxidation rate than the olivine (White et al., 1994). As we said above, in experiments at pH 6.3 and 8.3, the evolution of total iron concentration in solution as well as the evolution of iron(II) concentration in solution was determined as a function of time. With these values, iron(II) consumption rates of 5 10 8 mol dm 3d h 1 and
Fig. 7. Rates of oxygen uptake determined in this work for the olivine rock compared to the rates of oxygen uptake by augite, biotite and hornblende (data from White and Yee (1985)).
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
51
2 10 7 mol dm 3d h 1 are calculated for experiments at pH 6.3 and 8.3, respectively. By calculating mass balances from the iron analyses, we have obtained that a maximum of only a 6% of the oxygen consumption in the system is due to homogeneous oxidation of aqueous Fe(II). In Fig. 7, we have compared the oxygen consumption rates obtained in this work with the Lovasja¨rvi olivine rock with rates of oxygen consumption obtained for other solids taken from the literature (White and Yee, 1985). From this comparison, we can conclude that the rate of oxygen uptake by the olivine rock is two orders of magnitude higher than the oxygen depletion caused by the oxidation of augite, biotite and hornblende, other silicate minerals that contain iron(II). 4. Conclusion The main conclusion deduced from our results is that the olivine rock used in this work is able to buffer the redox potential and to deplete oxygen at a higher rate than other silicates such as hornblende, biotite or augite; therefore, this rock could be used as a redox-active backfill additive in deep nuclear waste repositories due to its redox buffer capacity in the groundwater in case of an oxidant intrusion, reaching anoxic conditions, as well as its capacity to retain different radionuclides. Acknowledgments We would like to thank Dr. Karl-Heinz Hellmuth for his interest in our work. This work has been supported by STUK (Finnish Centre for Radiation and Nuclear Safety), CICYT (Spain) (REN2000-0201-P4-05), and the Spanish dMinisterio de Ciencia y Tecnologı´a (MCyT)T by means of the FIATE project (REN2002-02971/TECNO) and the dRamo´n y CajalT programme. References Blum, A., Lasaga, A., 1988. Role of surface speciation in the low-temperature dissolution of minerals. Nature 4, 431 – 433. de Pablo, J., Casas, I., Gime´nez, J., Molera, M., Rovira, M., Duro, L., Bruno, J., 1999. The oxidative dissolution mechanism of uranium dioxide: I. The effect of temperature in hydrogen carbonate medium. Geochim. Cosmochim. Acta 63, 3097 – 3103. Duro, L., El Aamrani, F., Rovira, M., Gime´nez, J., Casas, I., de Pablo, J., Bruno, J., 2005. The dissolution of high-FeO olivine rock from Lovasja¨rvi intrusion (SE-Finland) at 25 8C as a function of pH. Appl. Geochem. 20, 1284 – 1291. El Aamrani, F.Z., Duro, L., de Pablo, J., Bruno, J., 2002. Experimental study and modelling of the sorption of uranium(VI) onto olivine-rock. Appl. Geochem. 17, 399 – 408. Eriksen, T.E., Eklund, U.-B., Werme, L., Bruno, J., 1995. Dissolution of irradiated fuel: a radiolytic mass balance study. J. Nucl. Mater. 227, 76 – 82. Gibbs, C., 1976. Characterization and application of ferrozine iron reagent as ferrous iron indicator. Anal. Chem. 48, 1197 – 1200. Hellmuth, K.-H., 1991. The existence of native iron—implications for nuclear waste management: Part II. Evidence from investigations of samples of native iron. Report STUK-B-VALO 68, Finnish Centre for Radiation and Nuclear Safety, Helsinki. Hellmuth, K.-H., Lindberg, A., Tullborg, E.-L., 1992. Water–rock interaction in a high-FeO olivine rock in nature. Report STUK-YTO-TR 43, Finnish Centre for Radiation and Nuclear Safety, Helsinki. Hellmuth, K.-H., Siitari-Kauppi, M., Rauhala, E., Johansson, B., Zilliacus, R., Gijbels, R., Adriaens, 1994. Reaction of high FeO olivine rock with groundwater and redox sensitive elements studied by surface analytical methods and autoradiography. Mater. Res. Soc. Symp. Proc. 333, 947 – 953.
52
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
Pokrovsky, O.S., Schott, J., 2000a. Forsterite surface composition in aqueous solutions: a combined potentiometric, electrokinetic, and spectroscopic approach. Geochim. Cosmochim. Acta 64, 3299 – 3312. Pokrovsky, O.S., Schott, J., 2000b. Kinetics and mechanism of forsterite dissolution at 25 8C and pH from 1 to 12. Geochim. Cosmochim. Acta 64, 3313 – 3325. Rodrigues, E., El Aamrani, F.Z., Gime´nez, J., Casas, I., Torrero, M.E., de Pablo, J., Duro, L., Hellmuth, K.-H., 1998. Surface characterization of olivine-rock by X-ray photoelectron spectroscopy (XPS). Leaching and U(VI)-sorption experiments. Mater. Res. Soc. Symp. Proc. 506, 321 – 327. Rovira, M., El Aamrani, F.Z., Duro, L., Casas, I., de Pablo, J., Bruno, J., Dome`nech, C., Ayora, C., 2000. Experimental study and modeling of uranium(VI) transport through ferrous olivine rock columns. Radiochim. Acta 88, 1 – 7. Shoesmith, D.W., 2000. Fuel corrosion processes under waste disposal conditions. J. Nucl. Mater. 282, 1 – 31. SKB91, 1992. Final disposal of spent nuclear fuel. Importance of the bedrock for safety, SKB Technical Report 92-20. SKB, Stockholm, Sweden. Suksi, J., Upero, M., Hellmuth, K.H., 1998. Water–rock interaction and sorption of redox-sensitive elements: experiments on olivine and uranium. In: Arehart, Hulston (Eds.), Water–Rock Interaction. Balkema, Rotterdam, pp. 943 – 946. White, A.F., Yee, A., 1985. Aqueous oxidation–reduction kinetics associated with coupled electron–cation transfer from iron-containing silicates at 25 8C. Geochim. Cosmochim. Acta 49, 1263 – 1275. White, A.F., Peterson, M.L., Hochella Jr., M.F., 1994. Electrochemistry and dissolution kinetics of magnetite and ilmenite. Geochim. Cosmochim. Acta 58, 1859 – 1875. Westrich, H.R., Gygan, R.T., Casey, W., Zemitis, C., Arnold, G.W., 1993. The dissolution kinetics of mixed-cation orthosilicate minerals. Am. J. Sci. 293, 869 – 893. Wogelius, R.A., Walther, J.V., 1991. Olivine dissolution at 25 8C: effects of pH, CO2, and organic acids. Geochim. Cosmochim. Acta 55, 943 – 954. Wogelius, R.A., Walther, J.V., 1992. Olivine dissolution kinetics at near-surface conditions. Chem. Geol. 97, 101 – 112.
The use of a high-FeO olivine rock as a redox buffer in a nuclear waste repository J. Gime´nez a,*, M. Rovira b, F. Clarens a, I. Casas a, L. Duro a, M. Grive´ c, J. Bruno c, J. de Pablo a,b a
Department of Chemical Eng. (ETSEIB-UPC), Av. Diagonal 647, 08028 Barcelona, Spain b CTM Centre Tecnolo`gic, Avda. Bases de Manresa, 1. 08240 Manresa, Spain c ENVIROS-SPAIN S.L. Passeig de Rubı´, 29-31, 08197 Valldoreix, Spain
Received 24 February 2005; received in revised form 20 October 2005; accepted 30 October 2005 Available online 13 December 2005
Abstract Due to the higher stability of the spent nuclear fuel (mainly composed of UO2) under reducing conditions, and in order to enhance the retention/retardation of some key radionuclides, the olivine rock from the Lovasja¨rvi intrusion has been proposed as a potential redox-active backfill-additive in deep highlevel nuclear waste (HLNW) repositories. In this work, two different approaches have been undertaken in order to establish the redox buffer capacity of olivine rock: (1) The capacity of the rock to respond to changes in pH or pe has been demonstrated and the final (pH, pe) coordinates agree with the control exerted by the system Fe(II)/Fe(III). (2) The rate of consumption of oxygen has been determined at different pH values. These rates are higher than the ones reported in the literature for other solids, what would point to the possibility of using this rock as an additive to the backfill material in a HLNW. D 2005 Elsevier B.V. All rights reserved. Keywords: High-level nuclear waste; Radionuclide migration; Olivine; Redox buffer capacity
1. Introduction The aim of the different engineered barriers for the long term disposal of high level nuclear waste (HLNW) is to avoid or to retard the dissemination of the radionuclides released from the spent nuclear fuel to the environment. The spent nuclear fuel is mainly composed of UO2 and its * Corresponding author. Departament Enginyeria Quı´mica, H-4 ETSEIB, Universitat Polite´cnica de Catalunya (UPC), Avda. Diagonal 647, 08028-Barcelona, Spain. Tel.: +34 934017388; fax: +34 934015814. E-mail address: [email protected] (J. Gime´nez). 0169-7722/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jconhyd.2005.10.011
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
43
corrosion increases with the redox potential, because uranium(VI) species are more mobile than the uranium(IV) ones (Shoesmith, 2000; de Pablo et al., 1999). The groundwaters that can arrive to the spent fuel are usually reducing but certain processes as water radiolysis or glacial meltwater intrusion can generate oxidizing species in the near-field of the fuel (Eriksen et al., 1995). For this reason, in the design of the repository, the function of the buffer material is to comprise a protective zone with a suitable environment around the canister (SKB91, 1992) and, in this context, the addition of chemical getters is suggested as a potential optimization of current designs of backfill material which are based upon the use of bentonite clay. The high-FeO olivine-rich rock from Lovasja¨rvi intrusion has been proposed as a potential redox-active backfill additive in deep nuclear waste repositories (Hellmuth, 1991; Hellmuth et al., 1992). In this sense, the main characteristics of this olivine rock are related to: (1) its ability to retain radionuclides dissolved from the nuclear waste, (2) its thermodynamic and kinetic stability, and (3) its redox buffer capacity. The retention capacity of the Lovasja¨rvi olivine rock is very important since it may control the migration of some significant radionuclides coming from the nuclear waste proximity. In fact, good sorption capacity of this rock has been found for some trace elements (Hellmuth et al., 1994; Rodrigues et al., 1998; Suksi et al., 1998; El Aamrani et al., 2002) and the uranium(VI) transport through columns filled with this rock was studied by our group (Rovira et al., 2000). The kinetics of dissolution of the olivine rock from Lovasja¨rvi has been recently studied as a function of pH in our department (Duro et al., 2005) and a similar behaviour in front of pH to that of the forsterite (Mg2SiO4) has been found. At acidic pH, the reaction order with respect to the proton concentration was 0.5, which agrees with values found in the literature for forsterite dissolution (Blum and Lasaga, 1988; Westrich et al., 1993; Wogelius and Walther, 1991, 1992; Pokrovsky and Schott, 2000a,b). At alkaline pH, considering the Si-based dissolution rates, the reaction order would be 0.3, similar to the reaction order obtained by Blum and Lasaga (1988) and Wogelius and Walther (1991). However, considering Mg and Fe data, the dissolution rates would not depend on the proton concentration. Both possibilities are discussed in Duro et al. (2005). In this work, we have studied the redox buffer capacity of the olivine rock, because it was thought that this capacity would be high considering its content on iron(II) present mainly in olivine (Mg,Fe)SiO4 and in less extent in magnetite (Fe3O4) mineral that is also present in the rock (see Table 1). Given that the buffering of the redox potential seems to be plausible, it is of the utmost importance the determination of the rate at which this olivine rock can consume a given amount of oxidants. In this sense, we have studied (1) the capacity of the olivine rock to respond to changes in pH and E h, and (2) the rate of oxygen consumption of the rock. The oxygen uptake capacity of the rock has also been compared to other solids that can be used as a redox buffer. 2. Experimental The mineral content of the high-FeO olivine-rich rock used in this study was: 65% of olivine (with 57% of forsterite, Mg2SiO4, and 43% of fayalite, Fe2SiO4), 20% of plagioclase (with 50% of anortite, CaAl2Si3O8, and 50% of albite, NaAlSi3O8), 8% of magnetite (Fe3O4) and 7% of pyroxene and serpentine. The chemical composition of the olivine rock is shown in Table 1 (Hellmuth et al., 1992). The solid was crushed and sieved to a 120–160 Am particle size, the specific surface area of the solid was determined to be 0.17 F 0.02 m2 g 1 by the BET method using a gas mixture of 30% N2/Ar.
44
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
Table 1 Composition of the olivine rock Wt.% SiO2 MgO FeO Fe2O3 MnO Cr2O3 Al2O3 TiO2 V2O5 CaO K2O Na2O NiO CoO P2O5 H2O
34.80 20.16 31.58 4.24 0.34 0.07 2.43 3.66 0.09 1.39 0.33 0.56 0.06 0.05 0.12 0.38
All the experiments were performed at 25.0 F 0.1 8C. Two different kinds of experiments were carried out in order to determine (1) the pe/pH buffering capacity and (2) the oxygen uptake capacity of the olivine rock, respectively. 2.1. Study of the E h/pH buffering capacity The experiments were performed in two identical batch reactors (I and II) as shown in Fig. 1. 200 cm3 of NaClO4 10 2 mol dm 3 were placed in each reactor and were continuously purged with N2(g), supplied by Air Liquide. The gas stream was bubbled through a solution of Cr(II) in contact with a Zn/Hg amalgam in order to avoid the introduction of oxidants into the system. The solutions were saturated by N2(g) bubbling overnight.
Fig. 1. Batch reactor used in the E h/pH buffering experiments.
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
45
Table 2 Experimental conditions used in pe/pH buffering capacity of olivine-rock tests
Part Part Part Part
A B C D
Time (h)
Solution composition
Final pH
Final pe
0–334 334–695 695–1682 1682–2400
NaClO4 10 2 M +HClO4 +NaOH +O2 (g)
8.2 2.6 6.13 5.0
3.5 10 63 5.5
2.667 g of the olivine rock was placed in reactor I while reactor II was used as a blank. Both reactors were continuously stirred during the tests by means of an orbital stirrer. The pH of the solutions was monitored by means of a calibrated combined-glass electrode. Redox potentials were measured with a platinum electrode, an additional Au redox electrode was introduced in reactor I. Redox potentials were measured against the Ag/AgCl(s) and KCl saturated reference of the combined glass electrode. During the experiments aqueous samples were taken periodically for iron analysis. These samples were immediately filtered through 0.22-Am-pore-size filters and acidified by adding a small volume of concentrated HNO3. Every sample was replaced by an equivalent volume of N2(g)-saturated ionic medium in reactor I, to keep constant the total volume of the solution inside the reactor. Iron was determined by the ferrozine method by means of UV–Vis spectrophotometry. Iron(II) concentration was determined spectrophotometrically in aqueous solution as the iron(II) complex with the disodium salt of ferrozine. Total iron concentration was determined by previous reduction of iron (III) to iron(II) using the same analytical method (Gibbs, 1976). The experiments were divided in four parts, A, B, C and D, depending on the experimental conditions employed, as can be seen in Table 2. These experiments were carried out in this way in order to force each time only one of the master variables (pH or pe) and to see the response of the other one. Once pH and pe reached steady state values, the conditions of the experiment were artificially changed in order to force the response of the olivine rock. Iron was analysed before and after each disturbance.
Fig. 2. Reactor scheme used in the oxygen consumption measurements.
46
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
Table 3 Initial conditions of the experiments of the olivine-rich rock redox buffer capacity study
Exp. Exp. Exp. Exp. Exp. Exp. Exp.
A B C D E F G
pHinitial
Weight of solid (g)
Volume (dm3)
Ionic medium
2.0 4.0 7.9 9.2 6.2 8.4 9.4
4.00 4.67 4.00 4.00 4.67 4.67 4.67
0.300 0.350 0.300 0.300 0.350 0.350 0.350
NaClO4, 0.01 mol dm 3 NaClO4, 0.01 mol dm 3 NaClO4, 0.01 mol dm 3 NaClO4, 0.01 mol dm 3 NaClO4, 0.01 mol dm 3 NaHCO3, 0.01 mol dm 3 NaHCO3, 0.01 mol dm 3
2.2. Study of the oxygen uptake capacity A scheme of the reactor used in this study is shown in Fig. 2. The evolution of the dissolved oxygen concentration with time was monitored by means of a dissolved oxygen meter (Orion Model 850) with a precision of 0.1 ppm.
Fig. 3. Evolution of (a) pH and (b) pe with time in the experiments related to the E h/pH buffer capacity of the olivine-rich rock.
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
47
4.67 g of the 120- to 160-Am-particle-size olivine rock was introduced into the reactor with a solution containing an initial dissolved oxygen concentration of 2.7 10 4 mol dm 3 in 0.01 mol dm 3 NaClO4. During the experiments, the pH was measured by means of a glass electrode. Samples of each experiment were withdrawn for iron quantification as in Section 2.1. A series of blank experiments were also performed in order to take into account the oxygen consumption by the oxygen meter due to the process of the measure itself. The initial conditions of the experiments are summarised in Table 3. 3. Results and discussion 3.1. Study of the E h/pH buffering capacity The evolution of the pH and pe as a function of time for all the experiment is shown in Fig. 3. In order to test whether the iron system is effectively buffering the pe/pH values, we have plotted all the experimental data on predominance diagrams of the iron system (see Fig. 4).
Fig. 4. Evolution of the experimental pe and pH superimposed to a predominance pe/pH diagram of the iron system. Solid arrows stand for the artificial disturbances imposed to the system.
48
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
As we can see, any artificial disturbance introduced in the system (pH decrease in part B, pH increase in part C, and pe increase in part D) induces a response in the system to re-establish the chemical equilibrium between Fe(II) and Fe(III).
Fig. 5. Evolution of the O2(aq) concentration and pH with time.
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
49
We can deduce, then, that the olivine rock is effectively responding to disturbances of either pH or redox potential externally induced. When we add HClO4 (part B), the pH decreases and the system evolves towards an increase in pe. Once the system is again in equilibrium, we add NaOH (part C) and the response of the system is a decrease of the redox potential. Finally, once the equilibrium is reached again we introduce O2(g) in the system in order to increase the redox potential (part D) which is contra arrested by a decrease in the pH of the system. As we can see also in Fig. 4, the final (pH,pe) coordinates are in agreement with the control exerted by the system Fe(II)/Fe(III) and that the precipitation of amorphous Fe(OH)3 solid can explain this behaviour. Therefore, it would be logical to assume that the surface of the sample of olivine rock is coated with this solid, which forms as a result of the oxidation of the structural Fe(II) of the solid, either of the olivine itself or of the Fe(II) contained in the magnetite of the rock. 3.2. Study of the oxygen uptake by the olivine rock In these experiments, we have measured the capacity of the olivine rock to consume oxygen introduced in the reactors. In all experiments, the oxygen concentration decreased with time (the oxygen concentrations measured can be seen in Fig. 5). In addition, the pH varied during the first hours of the experiments until it reached a constant value. We have calculated the rate of oxygen consumption in the different experiments from the decrease of the oxygen concentration in solution with time. The results obtained are shown in Table 4, the oxygen consumption rates have been calculated with the values measured when the pH was constant and are shown in Fig. 6. In this figure, we have also included the data from Duro et al. (2005) as a solid line: the olivine rock dissolution rates obtained based on the iron concentrations measured in solution as well as the dissolution rates obtained by applying a semi-empirical model. From the comparison between both rates, we can see that in the pH range of 6–10, representative of most groundwaters, the oxygen consumption rates are higher than the iron release rates, indicating that the oxygen is being consumed by the structural iron present in the rock either contained in the olivine or in the magnetite, according to the reaction: ½FeðIIÞsurface þ 1=4O2 þ Hþ Z ½FeðIIIÞsurface þ 1=2H2 O: This preferential reaction of the oxygen with the structural iron(II) would agree with the observations made by Wogelius and Walther (1992), who found that in the case of pure olivine dissolution, the release of iron to the solution is always slower than the oxygenation of iron and its subsequent precipitation as iron(III). In addition, studies focused on O2(aq) uptake by ironTable 4 Rates of oxygen consumption
Exp. Exp. Exp. Exp. Exp. Exp. Exp.
A B C D E F G
pHinitial
pHfinal
Rate (mg O2 dm 3 h 1)
2.0 4.0 7.9 9.2 6.1 8.4 9.4
2.5 5.7 6.0 6.0 6.3 8.3 9.3
0.059 F 0.003 0.015 F 0.002 0.009 F 0.001 0.052 F 0.001 0.007 F 0.001 0.023 F 0.003 0.062 F 0.002
50
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
Fig. 6. Experimental rates of oxygen consumption together with the olivine rock dissolution rates and theoretical dissolution rates determined in Duro et al. (2005, see the text).
containing silicates reported by White and Yee (1985) suggest surface oxidation of iron silicates as a significant mechanism for the consumption of O2(aq). In the case of the high-FeO olivine-rich rock studied in our work as previously stated, we have to consider also the presence in the solid of a 8% of magnetite, which would have even a higher oxidation rate than the olivine (White et al., 1994). As we said above, in experiments at pH 6.3 and 8.3, the evolution of total iron concentration in solution as well as the evolution of iron(II) concentration in solution was determined as a function of time. With these values, iron(II) consumption rates of 5 10 8 mol dm 3d h 1 and
Fig. 7. Rates of oxygen uptake determined in this work for the olivine rock compared to the rates of oxygen uptake by augite, biotite and hornblende (data from White and Yee (1985)).
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
51
2 10 7 mol dm 3d h 1 are calculated for experiments at pH 6.3 and 8.3, respectively. By calculating mass balances from the iron analyses, we have obtained that a maximum of only a 6% of the oxygen consumption in the system is due to homogeneous oxidation of aqueous Fe(II). In Fig. 7, we have compared the oxygen consumption rates obtained in this work with the Lovasja¨rvi olivine rock with rates of oxygen consumption obtained for other solids taken from the literature (White and Yee, 1985). From this comparison, we can conclude that the rate of oxygen uptake by the olivine rock is two orders of magnitude higher than the oxygen depletion caused by the oxidation of augite, biotite and hornblende, other silicate minerals that contain iron(II). 4. Conclusion The main conclusion deduced from our results is that the olivine rock used in this work is able to buffer the redox potential and to deplete oxygen at a higher rate than other silicates such as hornblende, biotite or augite; therefore, this rock could be used as a redox-active backfill additive in deep nuclear waste repositories due to its redox buffer capacity in the groundwater in case of an oxidant intrusion, reaching anoxic conditions, as well as its capacity to retain different radionuclides. Acknowledgments We would like to thank Dr. Karl-Heinz Hellmuth for his interest in our work. This work has been supported by STUK (Finnish Centre for Radiation and Nuclear Safety), CICYT (Spain) (REN2000-0201-P4-05), and the Spanish dMinisterio de Ciencia y Tecnologı´a (MCyT)T by means of the FIATE project (REN2002-02971/TECNO) and the dRamo´n y CajalT programme. References Blum, A., Lasaga, A., 1988. Role of surface speciation in the low-temperature dissolution of minerals. Nature 4, 431 – 433. de Pablo, J., Casas, I., Gime´nez, J., Molera, M., Rovira, M., Duro, L., Bruno, J., 1999. The oxidative dissolution mechanism of uranium dioxide: I. The effect of temperature in hydrogen carbonate medium. Geochim. Cosmochim. Acta 63, 3097 – 3103. Duro, L., El Aamrani, F., Rovira, M., Gime´nez, J., Casas, I., de Pablo, J., Bruno, J., 2005. The dissolution of high-FeO olivine rock from Lovasja¨rvi intrusion (SE-Finland) at 25 8C as a function of pH. Appl. Geochem. 20, 1284 – 1291. El Aamrani, F.Z., Duro, L., de Pablo, J., Bruno, J., 2002. Experimental study and modelling of the sorption of uranium(VI) onto olivine-rock. Appl. Geochem. 17, 399 – 408. Eriksen, T.E., Eklund, U.-B., Werme, L., Bruno, J., 1995. Dissolution of irradiated fuel: a radiolytic mass balance study. J. Nucl. Mater. 227, 76 – 82. Gibbs, C., 1976. Characterization and application of ferrozine iron reagent as ferrous iron indicator. Anal. Chem. 48, 1197 – 1200. Hellmuth, K.-H., 1991. The existence of native iron—implications for nuclear waste management: Part II. Evidence from investigations of samples of native iron. Report STUK-B-VALO 68, Finnish Centre for Radiation and Nuclear Safety, Helsinki. Hellmuth, K.-H., Lindberg, A., Tullborg, E.-L., 1992. Water–rock interaction in a high-FeO olivine rock in nature. Report STUK-YTO-TR 43, Finnish Centre for Radiation and Nuclear Safety, Helsinki. Hellmuth, K.-H., Siitari-Kauppi, M., Rauhala, E., Johansson, B., Zilliacus, R., Gijbels, R., Adriaens, 1994. Reaction of high FeO olivine rock with groundwater and redox sensitive elements studied by surface analytical methods and autoradiography. Mater. Res. Soc. Symp. Proc. 333, 947 – 953.
52
J. Gime´nez et al. / Journal of Contaminant Hydrology 83 (2006) 42–52
Pokrovsky, O.S., Schott, J., 2000a. Forsterite surface composition in aqueous solutions: a combined potentiometric, electrokinetic, and spectroscopic approach. Geochim. Cosmochim. Acta 64, 3299 – 3312. Pokrovsky, O.S., Schott, J., 2000b. Kinetics and mechanism of forsterite dissolution at 25 8C and pH from 1 to 12. Geochim. Cosmochim. Acta 64, 3313 – 3325. Rodrigues, E., El Aamrani, F.Z., Gime´nez, J., Casas, I., Torrero, M.E., de Pablo, J., Duro, L., Hellmuth, K.-H., 1998. Surface characterization of olivine-rock by X-ray photoelectron spectroscopy (XPS). Leaching and U(VI)-sorption experiments. Mater. Res. Soc. Symp. Proc. 506, 321 – 327. Rovira, M., El Aamrani, F.Z., Duro, L., Casas, I., de Pablo, J., Bruno, J., Dome`nech, C., Ayora, C., 2000. Experimental study and modeling of uranium(VI) transport through ferrous olivine rock columns. Radiochim. Acta 88, 1 – 7. Shoesmith, D.W., 2000. Fuel corrosion processes under waste disposal conditions. J. Nucl. Mater. 282, 1 – 31. SKB91, 1992. Final disposal of spent nuclear fuel. Importance of the bedrock for safety, SKB Technical Report 92-20. SKB, Stockholm, Sweden. Suksi, J., Upero, M., Hellmuth, K.H., 1998. Water–rock interaction and sorption of redox-sensitive elements: experiments on olivine and uranium. In: Arehart, Hulston (Eds.), Water–Rock Interaction. Balkema, Rotterdam, pp. 943 – 946. White, A.F., Yee, A., 1985. Aqueous oxidation–reduction kinetics associated with coupled electron–cation transfer from iron-containing silicates at 25 8C. Geochim. Cosmochim. Acta 49, 1263 – 1275. White, A.F., Peterson, M.L., Hochella Jr., M.F., 1994. Electrochemistry and dissolution kinetics of magnetite and ilmenite. Geochim. Cosmochim. Acta 58, 1859 – 1875. Westrich, H.R., Gygan, R.T., Casey, W., Zemitis, C., Arnold, G.W., 1993. The dissolution kinetics of mixed-cation orthosilicate minerals. Am. J. Sci. 293, 869 – 893. Wogelius, R.A., Walther, J.V., 1991. Olivine dissolution at 25 8C: effects of pH, CO2, and organic acids. Geochim. Cosmochim. Acta 55, 943 – 954. Wogelius, R.A., Walther, J.V., 1992. Olivine dissolution kinetics at near-surface conditions. Chem. Geol. 97, 101 – 112.