Tc(IV) in Dilute to Concentrated Nacl and Mgcl2 Solutions

Radiochim. Acta 2015; 103(1): 57–72

Ezgi Yalçıntaş*, Xavier Gaona, Andreas C. Scheinost, Taishi Kobayashi, Marcus Altmaier, and Horst Geckeis Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions

3 −1 Abstract: The redox behaviour of Tc(VII)/Tc(IV) was in- all cases (𝑅d ≥10 Lkg ), although 𝑅d values are signif- vestigated within the pHc range 2–14.6 in (0.5 M and icantly lower in 4.5 M MgCl2 solutions. XANES (X-ray ab- 5.0 M) NaCl and (0.25 M, 2.0 M and 4.5 M) MgCl2 sorption near edge spectroscopy) evaluation of these sam- solutions in the presence of different reducing agents ples confirms that Tc(VII) is reduced to Tc(IV) by Fe(II)

(Na2S2O4, Sn(II), Fe(II)/Fe(III), Fe powder) and macro- minerals also in concentrated NaCl and MgCl2 brines. scopic amounts of Fe minerals (magnetite, mackinawite, Keywords: Technetium, redox reactions, salt brines, ther- siderite: S/L =20–30 g L−1). In the first group of samples, modynamics, Fe(II) minerals, XANES. the decrease of the initial Tc concentration (1×10−5 M, as Tc(VII)) indicated the reduction to Tc(IV) according to the chemical reaction TcO − +4H+ +3e− ⇔ DOI 10.1515/ract-2014-2272 4 Received April 1, 2014; accepted August 11, 2014 TcO 2 ⋅1.6H2O(s) + 0.4H2O. Redox speciation of Tc in the aqueous phase was further confirmed by solvent extraction. A good agreement is obtained between the experimentally determined Tc redox distribution and thermody- 1 Introduction namic calculations based on NEA–TDB (Nuclear Energy Agency, Thermochemical Database) and ionic strength 99Technetium is a 𝛽-emitting fission product primarily corrections by SIT or Pitzer approaches. These observa- produced in nuclear reactors by the fission of 235U and 239 tions indicate that experimental pHc and 𝐸h values in Pu. Due to its high fission yield, long half-life𝑡 ( 1/2 ∼ buffered systems can be considered as reliable param- 2.1 × 105 y) and redox-sensitive character, a reliable pre- eters to predict the redox behaviour of Tc in dilute to diction of the chemical behaviour of technetium is neces- highly concentrated NaCl and MgCl2 solutions. 𝐸h of sary for the long-term safety assessment of nuclear waste the system and aqueous concentration of Tc(IV) in equi- repositories. In the case of repositories in salt-rock forma- librium with TcO 2 ⋅1.6H2O(s) are strongly affected by tions, the boundary conditions may potentially include elevated ionic strength, especially in the case of 4.5 M saturated NaCl and MgCl2 brines with salt concentrations MgCl2 solutions. In such concentrated brines and un- above ∼5Mand ∼4.5M, respectively. Technetium is also der alkaline conditions (pHc =pHmax ∼9), kinetics play a relevant contaminant associated with sites for nuclear a relevant role and thermodynamic equilibrium for the sys- fuel reprocessing or plutonium production. Among the lat- tem Tc(IV)(aq) ⇔ Tc(IV)(s) was not attained from over- ter, the Hanford site (Washington State, USA) probably rep- saturation conditions within the timeframe of this study resents the largest remediation effort in the United States. (395 days). Tc(VII) is reduced to Tc(IV) by magnetite, mack- In this site, Tc is mainly found as soluble species in the inawite and siderite suspensions at pHc =8–9 in concen- supernatant of underground storage tanks, which mostly trated NaCl and MgCl2 solutions. Sorption is very high in consist of concentrated aqueous solutions of sodium ni- trate/nitrite salts and sodium hydroxide [1, 2]. The chemical behaviour of radionuclides in concentrated salt brines can significantly differ from observations *Corresponding author: Ezgi Yalçıntaş, Institute for Nuclear Waste Disposal, Karlsruhe Institute of Technology, Germany, made in dilute solutions. Strong interaction processes be- e-mail: [email protected] tween dissolved radionuclide species and the main back- Xavier Gaona, Marcus Altmaier, Horst Geckeis: Institute for Nuclear ground electrolyte ions can importantly modify the stabil- Waste Disposal, Karlsruhe Institute of Technology, Germany ity of charged species at elevated ionic strength. In some Andreas C. Scheinost: Institute of Resource Ecology, Helmholtz- cases, the formation of new aquatic species and complexes Zentrum Dresden-Rossendorf, Germany; and Rossendorf Beamline (ROBL) at ESRF, Grenoble, France unknown in dilute systems takes place in concentrated so- Taishi Kobayashi: Department of Nuclear Engineering, Kyoto lutions, as recently reported for An(III/IV/V) in concen- University, Japan trated CaCl2 brines [3–5]. Hence, the chemical behaviour 58 | E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions

and mobility of radionuclides under high ionic strength netite in the presence of synthetic ground water. The au- conditions cannot be simply extrapolated from dilute sys- thors assessed the effect of ionic strength and pH (7.8– tems but need to be investigated for particular saline sys- 9.5), investigated the rate of the sorption reaction, and tems. concluded that the uptake was controlled by a ligand ex- The mobility of Tc is strongly dependent on its oxida- change mechanism. Geraedts et al. (2002) [16]andMaes tion state. Although several oxidation states of Tc are re- et al. (2004) [17] studied the system magnetite–Tc in the ported in the literature (ranging from +II to +VII) [6–9], presence of natural and synthetic Gorleben groundwa-

Tc(VII) and Tc(IV) are the prevailing long-term stable re- ter. The authors concluded that TcO 2 ⋅ xH2O(s) formed dox states in the absence of any complexing ligand other in this system, and suggested that Tc(IV) polymers or than water under non reducing and reducing conditions, colloids were responsible for the observed increase in respectively. Heptavalent Tc exists as highly soluble and solubility (∼10−6 M). Wharton et al. (2000) [18]studied − mobile pertechnetate anion (TcO 4 ), whereas Tc(IV) forms the coprecipitation of Tc(VII) and Tc(IV) with mackinaw- sparingly soluble hydrous oxides (TcO 2 ⋅ xH2O(s)). Due ite (FeS) and characterized the resulting solid phases by to the very different mobility and chemical characteris- X–ray absorption spectroscopy. Tc was immobilized as IV tics of both oxidation states, an appropriate knowledge of a Tc S2-like phase regardless of the initial oxidation state the redox chemistry of Tc(VII)/Tc(IV) and Tc(IV) solubil- of Tc. Similar observations were reported by Livens et al. ity/sorption processes is of special relevance in the context (2004) [19], who investigated the interaction between Tc of radioactive waste disposal, as well as in the assessment and mackinawite using both +VII and +IV as initial re- of Tc behaviour in contaminated sites. dox state of Tc. Liu et al. (2008) [20] performed comprehen- The redox transformations of Tc(VII)/Tc(IV) in the sive immobilization experiments with Tc in the presence presence of different reducing systems were previously as- of mackinawite. The authors assessed the effect of ionic sessed in a number of experimental studies mostly in low strength (≤1.0MNaCl)andpH(6.1–9.0)ontheuptake ionic strength systems. Owunwanne et al. (1977) [10]and of Tc, and observed a strong pH–dependence and the in- Warwick et al. (2007) [11] used Sn(II) to study the reduc- crease of the uptake rate with increasing ionic strength. tion of Tc(VII) under highly acidic (pH < 2) and hyper- In contrast to Livens and co-workers, TcO 2-like instead alkaline (pH > 13.3) conditions, respectively. A fast and of TcS 2-like phases were reported to form on the sur- complete reduction of Tc(VII) was observed in both cases. face of mackinawite. Llorens et al. (2008) [21]coprecipi-

Hess et al. [12] obtained a well-defined TcO 2 ⋅ xH2O solid tated Tc(VII) with siderite (FeCO3), an Fe(II) solid phase phase by reducing Tc(VII) with Na2S2O4.Theresulting forming in the presence of carbonate-rich groundwaters. Tc(IV) solid phase was used in undersaturation solubil- The reduction of Tc(VII) and incorporation of Tc(IV) into ity experiments under saline (≤5.0MNaCl) and acidic the structure of siderite were confirmed by X–ray diffrac- (≤6.0MHCl) conditions. Based on their solubility and tion and transmission electron microscopy. Peretyazhko spectroscopic (UV–VIS/NIR, EXAFS) data, the authors de- et al. (2012) [22] studied the reduction of Tc by reactive rived comprehensive thermodynamic and activity models ferrous iron forming in naturally anoxic, redox transi- for Tc(IV) valid up to pHc ∼7.Cuiet al. (1996a) [13]re- tion zone sediments from the Hanford site. A number of ported a kinetically hindered reduction of Tc(VII) to Tc(IV) Fe(II) minerals were identified in the sediments, includ- in the presence of aqueous Fe(II), whereas Fe(II) precipi- ing Fe(II)–phyllosilicates, pyrite, magnetite and siderite. tated or sorbed on the vessel walls led to a rapid reduction The authors observed TcO 2 ⋅ xH2O-like phases regardless of Tc(VII) at pH > 7.5.Zacharaet al. (2007) [14]alsostud- of Fe(II) mineralogy or reduction rate. ied the reduction of Tc(VII) in the presence of Fe(II)(aq) un- Recently, Kobayashi and co–workers [23]systemati- der near-neutral pH conditions (6≤pH≤8). In contrast cally assessed the reliability of 𝐸h–pH measurements to Cui and co-workers, the authors observed the reduction and reaction kinetics to explain the redox behaviour of Tc(VII) at pH ≥ 6.8, although reduction kinetics were of Tc(VII)/Tc(IV) in dilute solutions (0.1 M NaCl) in strongly dependent on pH. the presence of homogeneous and heterogeneous re- The role of Fe solid phases in the reduction of Tc(VII) ducing systems. The authors compared their experi- has been intensively studied within the last decades. mental data with available thermodynamic predictions

Among Fe phases, magnetite (Fe3O4) is often considered for Tc [24], and reported an experimental borderline for as one of the most relevant corrosion product of steel Tc(VII)/Tc(IV) reduction about 100 mV below the bor- forming under the strongly reducing conditions expected derline calculated thermodynamically. This observation

in underground repositories for nuclear waste disposal. was explained by the formation of TcO 2 ⋅ xH2O(coll, hyd) Cui et al. (1996b) [15] studied the uptake of Tc by mag- colloidal particles with higher solubility than the solid E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions | 59

phase TcO 2 ⋅1.6H2O(s) selected in the NEA–TDB review. 0.8 In heterogeneous systems (presence of mineral phases), 12 1 bar O2(g) the reduction of Tc(VII) occurred only in the presence of (pe+pH)=20.77 0.6 magnetite, mackinawite and siderite phases. 8 TcO- Inspiteofthelargenumberofexperimentalstud- 4 0.4

ies focussing on Tc redox reactions, the understanding of 2+ 4 + "Redox neutral" Tc(VII)/Tc(IV) redox processes and Tc(IV) uptake by Fe TcO (pe+pH)=13.8 0.2 minerals is mostly limited to dilute aqueous systems, and B TcO(OH) C investigations addressing Tc redox chemistry in concen- 0 0.0 pe (V)

trated salt brine systems remain very scarce. In this work, h TcO(OH) (aq)

2 E the redox behaviour of Tc(VII)/Tc(IV) couple was inves- -0.2 -4 tigated in NaCl ( and )and ( , x A 0.5 M 5.0 M MgCl2 0.25 M TcO2. H2O(s) 2.0 M and 4.5 M) solutions. The reduction of Tc(VII) to -0.4 Tc(IV) was assessed in the presence of different reducing -8 - 3 I = 0 (A) 1 bar H2(g) systems, e.g. Na2S2O4, Sn(II), Fe(II)/Fe(III), Fe powder, I = 5.0 M (SIT) (B) (pe+pH)=0 -0.6 magnetite, mackinawite and siderite. The results are sys- I = 5.0 M (Pitzer) (B) TcO(OH) -12 I = 13.5 M (SIT) (C) tematised according to the experimental 𝐸h and pH val- I = 13.5 M (Pitzer) (C) -0.8 ues of the solution/suspension, the measured Tc concen- 02468101214 tration and redox state and compared to thermodynamic pH calculations (using SIT (Specific Ion Interaction Theory) c or Pitzer ion interaction models for ionic strength correc- Fig. 1: Pourbaix diagram of Tc calculated for 𝐼=0with tions). XANES was used to characterize the redox state thermodynamic data selected in NEA–TDB [25]. Thin dark grey lines of Tc sorbed by magnetite, mackinawite and siderite in correspond to 50 : 50 distribution borderlines between aqueous Tc concentrated brine solutions. Using a similar experimen- species at 𝐼=0. Black lines indicate upper and lower tal and conceptual approach, the studies reported here ex- decomposition lines of water and “redox neutral” (pe + pH = 13.8) conditions. Colored lines (A, B, C) correspond to Tc(VII)/Tc(IV) tend the study published by Kobayashi et al. to other elec- borderlines considering reaction (1), calculated for (A) 𝐼=0(blue trolyte systems and high ionic strength conditions. line), (B) 𝐼=5.0M(green dashed line: SIT, green solid line: Pitzer; 5.0 M NaCl) and (C) 𝐼 = 13.5 M (red dashed line: SIT, red solid line: −5 Pitzer; 4.5 M MgCl2). All calculations performed at [Tc]tot =10 M. 2 Thermodynamic background Ionic strength corrections performed by SIT and Pitzer approaches as described in Sects. 2.1 and 2.2.

The thermochemical database project of the Nuclear En- ergy Agency (NEA–TDB) comprises the most comprehen- species (not currently selected in the NEA–TDB) have been sive selection of thermodynamic data currently available reported to form under very acidic and reducing condi- for actinides and fission products. Technetium was ini- tions [9, 26]. Due to the very low solubility of Tc(IV), one of tially reviewed in the volume 3 of the NEA–TDB series [25], the most relevant redox reactions of Tc corresponds to the − although the thermodynamic selection was later revisited reduction of TcO 4 to the solid TcO 2 ⋅1.6H2O(s) phase in the update volume by Guillaumont and co-authors [24]. according to The outcome is a critically reviewed selection of Tc thermo- − + − dynamic data, rather complete for Tc(VII)/Tc(IV) redox TcO 4 +4H +3e ⇔TcO2 ⋅1.6H2O(s) + 0.4H2O reactions and Tc(IV) hydrolysis, solubility and carbonate ∗ 󸀠 log 𝐾 VII–IVs (1) complexation. The data selection is limited or non-existent ∗ 󸀠 for other potentially relevant ligands such as chloride, where log 𝐾 VII–IVs is the conditional equilibrium con- − phosphate or sulphate. No ion interaction coefficients are stant for the reduction of TcO 4 to TcO 2 ⋅1.6H2O(s) at selected for ionic species of Tc, thus limiting the applica- 𝐼=0. The blue solid line (A) in Figure 1 corresponds to bility of the selected thermodynamic data to diluted sys- the thermodynamic borderline for Tc(VII)/Tc(IV) calcu- tems where no specific ion interaction must be considered. lated at 𝐼=0according to reaction (1) and represents 50% Figure 1 shows the Pourbaix diagram of Tc within Tc(VII) and 50% Tc(IV) redox state distribution. Note that

0≤pHc ≤14and −14≤pe≤14as calculated with the this borderline depends both upon ionic strength and to- current NEA–TDB data selection. Only Tc(VII) and Tc(IV) tal Tc concentration, the later held constant at [Tc]0 = species appear in the diagram, although aqueous Tc(III) 1×10−5 M in the calculations consistent with the exper- 60 | E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions

imental conditions selected in this study (see Sect. 3.3). 2.1 Thermodynamic SIT model At 𝐼 =0̸ , chemical equilibria are affected by ion interaction processes and activity coefficients need to be taken The specific ion interaction theory (SIT approach [27]) is into account. Because of the high ionic strength in the the method adopted by NEA–TDB [24]forthetreatment experiments described in this work (up to 𝐼=5.0Min of ion interaction processes and ionic strength effects.

NaCl and 𝐼 = 13.5 M in MgCl2 systems), SIT and Pitzer ap- The validity of SIT is normally limited to 𝐼m ≤3.0M[28], − + although recent publications have shown a good perfor- proaches were used to calculate 𝛾TcO 4 and 𝛾H and thus ∗ 󸀠 determine log 𝐾 VII–IVs for each ionic strength according mance of SIT far beyond this limit in concentrated MgCl2 with Eqs. (2)–(3). and CaCl2 solutions [4, 5]. Activity coefficients in SIT are calculated according to Eq. (4): ∗ ∘ − − log 𝐾 VII–IVs =−log([TcO4 ]⋅𝛾TcO 4 ) log 𝛾 =−𝑧2𝐷+∑ 𝜀 (𝑗, 𝑘)𝑚 (4) + 𝑗 𝑗 𝑘 𝑘 −4log([H ]⋅𝛾H+ )+3pe+0.4log𝑎w (2) where 𝑧𝑗 isthechargeofion𝑗, 𝐷 is the Debye–Hückel ∗ 󸀠 ∗ ∘ − + log 𝐾 VII–IVs =log 𝐾 VII–IVs +log𝛾TcO 4 +4log𝛾H term, 𝑚𝑘 is molality of the oppositely charged ion 𝑘 and 𝜀(𝑗, 𝑘) is the specific ion interaction parameter. Thus, the −0.4log𝑎w (3) ∗ 󸀠 ionic strength dependence of log 𝐾 VII–IVs in NaCl and

− + MgCl solutions can be calculated by SIT based on equa- Table 1 summarizes the values of log 𝛾TcO 4 , log 𝛾H and 2 ∗ 󸀠 tions (5)and(6), respectively: log 𝐾 VII–IVs in NaCl (0.5 M and 5.0 M)andMgCl2 (0.25 M, 2.0 M and 4.5 M) calculated by SIT and Pitzer ∗ 󸀠 ∗ ∘ models as described in Sects. 2.1 and 2.2, respectively. log 𝐾 VII–IVs =log 𝐾 VII–IVs −0.4log𝑎w −5𝐷 − + + − Tc(VII)/Tc(IV) redox borderlines calculated for NaCl and +𝜀(TcO4 ,Na )𝑚Na+ +4𝜀(H ,Cl )𝑚Cl− (5) solutions using ∗ 󸀠 are compared in the ∗ 󸀠 ∗ ∘ MgCl2 log 𝐾 VII–IVs log 𝐾 VII–IVs =log 𝐾 VII–IVs −0.4log𝑎w −5𝐷 following sections with Tc redox distribution determined − 2+ + − +𝜀(TcO ,Mg )𝑚 2+ +4𝜀(H ,Cl )𝑚 − (6) experimentally for each independent reducing system. 4 Mg Cl Borderlines for Tc(VII)/Tc(IV) reduction calculated for − No SIT ion interaction coefficients are selected for TcO 4 5.0 M NaCl and 4.5 M MgCl are also included in Fig- − 2 in the NEA–TDB. The chemical analogy of TcO 4 with ure 1. The comparison of thermodynamic calculations for − ClO4 is therefore considered in this work to esti- 𝐼=0 (solid blue line (A)), 𝐼=5.0M (green lines (B), mate the corresponding SIT coefficients with Na+ and 5.0 M NaCl)and𝐼 = 13.5 M (red lines (C), 4.5 M MgCl ) 2+ − + − + 2 Mg , yielding 𝜀(TcO4 ,Na )≅𝜀(ClO4 ,Na )=0.01± clearly indicates that a stabilization of Tc(IV) relative to −1 − 2+ − 2+ 0.01 kg mol and 𝜀(TcO4 ,Mg )≅𝜀(ClO4 ,Mg )= Tc(VII) is expected with increasing ionic strength. Only 0.33 ± 0.03 kg mol−1 [24]. The activity of water in dilute minor differences≤0.3 ( pe-units) are observed between to concentrated NaCl and MgCl2 solutions, as well as Tc(VII)/Tc(IV) borderlines calculated by SIT and Pitzer, 𝜀(H+,Cl−) = 0.12±0.01 kg mol−1 were taken directly from indicating the validity of the SIT also in the extremely high the NEA–TDB [24]. saline conditions of this study.

2.2 Thermodynamic Pitzer model

− + Table 1: Activity coefficients of TcO 4 and H in (0.5 and 5.0 M) NaCl Due to the known limitations of the SIT approach at and (0.25, 2.0 and 4.5 M) MgCl2 solutions derived from SIT and Pitzer approaches (see Sects. 2.1 and 2.2). Conditional equilibrium high ionic strengths, the use of the Pitzer formulism is ∗ 󸀠 constants (log 𝐾VII–IVs) for reaction (1) are calculated for the strongly recommended for thermodynamic calculations corresponding ionic strengths conditions according to Eq. (3). and geochemical modelling in concentrated salt brine solutions [29]. The Pitzer approach describes the interactions ∗ 󸀠 Background log 𝛾 + log 𝛾 − log 𝐾 H TcO 4 VII–IVs between a cation 𝑐 and an anion 𝑎 by the binary parame- (0) (1) (𝜑) electrolyte SIT Pitzer SIT Pitzer SIT Pitzer 󸀠 󸀠 ters 𝛽𝑐𝑎 , 𝛽𝑐𝑎 and 𝐶𝑐𝑎 and mixing parameters 𝜃𝑐𝑐 (or 𝜃𝑎𝑎 ) 0.5 M NaCl −0.11 −0.10 −0.17 −0.19 37.2 37.2 and 𝛹𝑐󸀠𝑎(or 𝛹𝑎𝑎󸀠𝑐). The latter account for interactions with 5.0 M NaCl 0.41 0.59 −0.21 −0.16 39.3 40.1 further cations c (or anions a) in ternary or higher mul- 0.25 M MgCl2 −0.13 −0.19 −0.11 −0.06 37.2 37.0 ticomponent electrolyte solutions. The values of log 𝛾H+ 2.0 M MgCl2 0.24 0.21 0.43 0.50 39.2 39.2 and 𝑎w in equation (2) used in this study are calculated 4.5 M MgCl2 0.95 1.06 1.41 1.51 43.2 43.7 from the parameters reported by Harvie et al. [30]. Binary E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions | 61

− ∘ (0) (1) −1 (𝜙) Table 2: Pitzer ion interaction coefficients for TcO 4 in NaCl and MgCl2 media at 25 C: 𝛽𝑐𝑎 , 𝛽𝑐𝑎 and 𝜃𝑎𝑎󸀠 in [kg mol ]; 𝐶𝑐𝑎 and 𝛹𝑎𝑎󸀠𝑐 in 2 −2 [kg mol ].

Binary Pitzer parameters Ternary Pitzer parameters (0) (1) (𝜙) 󸀠 𝑎𝑐𝑐𝑎 𝛽 𝑐𝑎 𝐶𝑐𝑎 𝑎 𝜃𝑎𝑎󸀠 𝛹𝑎𝑎󸀠𝑐 Ref. − + − TcO 4 Na 0.01111 0.1595 0.00236 Cl 0.067 −0.0085 [31] − 2+ − TcO 4 Mg 0.3138 1.84 0.0114 Cl 0.067 −0.0115 [32]

− + − 2+ parameters for (TcO 4 , Na )and(TcO 4 , Mg ), mixing Orion) calibrated against standard pH buffers (pH 1–12, − − parameter for (TcO 4 , Cl ) and ternary parameters for Merck). The values of pHc =pHexp +𝐴c were calcu- − − + − − 2+ (TcO 4 , Cl , Na )and(TcO 4 , Cl , Mg ) were taken lated from the operational “measured” pHexp using em- from Könnecke et al. [31]andNecket al. [32] as summa- pirical corrections factors (𝐴c), which entail both the liq- rized in Table 2. uid junction potential and the activity coefficient of H+.

𝐴c values determined as a function of NaCl and MgCl2 concentration are reported in the literature [34]. In NaCl– 3 Experimental NaOH solutions with [OH−] > 0.03 M,theH+ concentra- tionwascalculatedfromthegivenfixedhydroxideconcen- − 󸀠 3.1 Chemicals tration [OH ] and the conditional ion product 𝐾w of water. In MgCl2 solutions, the highest pH is fixed by the pre- All solutions were prepared with purifiedater w (Milli- cipitation of Mg(OH)2(cr) (or Mg2(OH)3Cl ⋅ 4H2O(cr) Q academic, Millipore) and purged with Ar before use in MgCl2 ≥2m) which buffers pH conditions at pHc = to remove traces of oxygen. All sample preparation and pHmax ∼9[34]. handling was performed in an Ar-glovebox at 22 ± 2 ∘C. Redox potentials were measured with Pt combination A purified and radiochemically well-characterized 99Tc electrodes with Ag/AgCl reference system (Metrohm) and stock solution (1.3 M NaTcO4) was used for the experi- converted to 𝐸h vs. the standard hydrogen electrode by correction for the potential of the reference elec- ments. NaCl (p.a.), MgCl2 ⋅6H2O (p.a.), Mg(OH)2(cr), Ag/AgCl ∘ Na2S,hydroquinone(C6H4(OH)2), Na2S2O4,andmetal- trode (+208 mV for 3M KCl at 25 C). Stable 𝐸h read- licironpowder(grainsize10 𝜇m)wereobtainedfrom ings were obtained within 10 min in most of the samples,

Merck; FeCl3 ⋅6H2O, SnCl2, Fe(NH4)2(SO4)2 ⋅6H2O, although in some cases longer equilibration times (up to 󸀠) were required. The apparent electron activity (pe pH buffers MES (pH 5–7) and PIPES (pH 7–9), NH4OH 30 = and tetraphenylphosphonium chloride (TPPC) were ob- −log𝑎e−) was calculated from 𝐸h =−(𝑅𝑇/𝐹)ln𝑎e−,according to the relation pe (V). The performance tained from Sigma-Aldrich, and FeCl2 from Alfa Aesar. = 16.9 𝐸h HCl and NaOH Titrisol® (Merck) were used for adjust- of the redox electrode was tested with a standard redox buffer solution (Schott, . ), which ing the pH of solutions. Magnetite (Fe3O4), mackinawite +220 mV vs Ag/AgCl (FeS) and siderite (FeCO3) were synthesized following the provided readings within ±10 mV of the certified value. protocol described elsewhere [33]. Magnetite and macki- Previous studies [35, 36] have suggested the need of (exper- nawite were characterized by high energy powder XRD imentally determined) correction factors for 𝐸h measured (D8 ADVANCE, Bruker). No XRD characterization was per- at high ionic strengths, which should mostly account for formed for siderite due to the rapid oxidation of the dry variations in the liquid junction potential. Liquid junction powder during measurement. However, the characteris- potentials below 50 mV are expected in the conditions tic light-gray colour of siderite was retained in the aque- of this study [37]. These values are well within the uncer- ous samples throughout the equilibration time and subse- tainty considered for 𝐸h measurements, and thus the use quent XANES sample preparation performed in the glove- of such corrections was disregarded in the present work. box.

3.3 Sample preparation and E 3.2 pH and h measurements characterization

+ The hydrogen ion concentration (pHc =−log[H ])was Batch experiments were performed in NaCl (0.5 M and measured using combination pH electrodes (type ROSS, 5.0 M)andMgCl2 (0.25 M, 2.0 M and 4.5 M)inthe 62 | E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions

presence of different reducing systems1mMNa ( 2S2O4, (21 044 eV) at the Rossendorf Beamline (ROBLII), ESRF, 1mMSnCl2, 1 mM/0.1 mM Fe(II)/Fe(III), 1mg/15mL Grenoble, France. The energy of the Si(111) double-crystal Fe powder). In some of these systems, the formation of monochromator was calibrated using a Mo foil (edge en- a solid phase occurred at near-neutral to alkaline pH con- ergy 20 000 eV). Two Rh-coated mirrors were used to col-

ditions (SnCl2, Fe(II)/Fe(III) and Fe powder). In each ex- limate the beam into the monochromator crystal and perimental series, pH values were adjusted with either to reject higher-order harmonics. Fluorescence spectra

HCl–NaCl–NaOH or HCl–MgCl2–Mg(OH)2 of identi- were collected using a 13-element, high-purity, solid-state cal ionic strength. The initial Tc(VII) concentration was Ge detector (Canberra) with digital spectrometer (XIA −5 set to 1×10 M by addition of NaTcO4 stock solution to XMAP). During the measurement, the samples were kept the pre-equilibrated system. at 15 K with a closed-cycle He cryostat to avoid changes Tcconcentrationinsolutionwasmonitoredatregu- of oxidation state and to reduce thermal disorder in the lar time intervals (up to 395 days) by Liquid Scintillation samples [33]. The XANES spectra were compared with − Counting (LSC, Quantulus, Perkin Elmer) after 10 kD ul- a reference spectra of TcO 4 [41]. trafiltration2 ( –3nm, Pall Life Sciences). Samples for LSC analysis were mixed with 10 mL of LSC–cocktail (Ultima Gold XR, Perkin–Elmer), resulting in a detection limit of ∼ 4 Results and discussion 10−9 M. Although several measurements were performed for each independent batch sample throughout the time- 4.1 Tc redox chemistry in reducing aqueous frame of the experiment, note that Figs. 2–5 show only systems equilibrium values (except indicated otherwise). The oxidation state of Tc in the aqueous phase was determined by Experimental 𝐸h and pHc values measured in NaCl and solvent extraction as reported elsewhere [38, 39]. Briefly, MgCl2 solutions in the presence of reducing chemicals the supernatant of the sample was contacted with an an- are summarized in the 𝐸h–pH diagrams in Figs. 2–5.Col- ion exchanger 50 mM tetraphenylphosphonium chloride ored dashed and solid lines in the figures represent the (TPPC) in chloroform. After vigorous mixing for 1minute thermodynamic equilibrium between Tc(VII) and Tc(IV) and subsequent separation of the aqueous and organic (50 : 50 distribution borderline) calculated for reaction phases by centrifugation, Tc concentration in the aqueous (1) by SIT and Pitzer approaches, respectively. These cal- phase was determined by LSC and attributed to presence culations were performed assuming a content of 1.6 H2O of Tc(IV). molecules in the hydration sphere of TcO 2 ⋅ xH2O(s) ac- Six additional samples in 5.0 M NaCl and 4.5 M cording to literature [42]. The samples were regularly mon- MgCl2 were prepared in the presence of macroscopic itored for pHc, 𝐸h and [Tc] for up to 395 d.Inordertoclar- amounts of magnetite (solid-to-liquid ratio, S : L = ify the graphs, only average values are plotted for samples −1 −1 21 g L ), mackinawite (S : L =30gL )andsiderite where stable readings indicate that equilibrium has been (S : L =20gL−1). Initial concentration of Tc(VII) was set reached. Tc concentration as a function of pHc for each re- to 2×10−4 M to allow XANES analysis, and a contact ducing system is provided besides the corresponding 𝐸h– time of 4 weeks was adopted in all cases. pHc, 𝐸h and pH diagrams. Uncertainty associated to [Tc] is lower than [Tc] (after 10 kD ultrafiltration) were determined before 0.1 log-units. Only for a limited number of systems, kinetic phase separation using centrifugation at 4020 𝑔.Thewet processes are specifically addressed. The decrease of the paste resulting after phase separation was placed into Tc concentration in the aqueous phase is interpreted as double confined sample holders, heat-sealed inside the the reduction of Tc(VII) and consequent formation of the Ar-glovebox and stored in a N2 Dewar (Voyager 12, Air TcO 2 ⋅1.6H2O(s) phase, which is sparingly soluble under Liquide – DMC, France) until the collection of XANES the adopted conditions. Table 3 summarizes the redox dis- spectra. This method has been previously proven to avoid tribution of Tc in the aqueous phase of selected samples as changes in oxidation state of redox sensitive samples (e.g. quantified by solvent extraction. Np, Pu and Tc) [23, 33, 40].

4.1.1 Na S O systems 3.4 XANES measurements 2 2 4

𝐸h–pH diagrams of Tc in 1mMNa2S2O4 are shown in Fig- X-ray absorption near-edge structure (XANES) spectra ure 2a(NaCl)and2c(MgCl2). In all samples, measured were acquired in fluorescence mode at the Tc 𝐾-edge E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions | 63

-4 a) 0.8 b) 12 1 bar O (g) 2 Initial [TcO- ] 0.6 4 (pe+pH)=20.77 -5 8 0.4 "Redox neutral" (pe+pH)=13.8 -6 4 0.2

0.5 M NaCl 0 0.0 (V) -7

pe 5.0 M NaCl h E log[Tc]

1 bar H (g) -0.2 -4 2 (pe+pH)=0 -8 -0.4 -8 -9 limit of detection -0.6

-12 0.5 M NaCl 0.5 M NaCl 5.0 M NaCl 5.0 M NaCl -0.8 -10 02468101214 02468101214 pH pHc c

-4 c) 0.8 d) 12 1 bar O (g) 2 - (pe+pH)=20.77 0.6 Initial [TcO ] -5 4 8 0.4 "Redox neutral" (pe+pH)=13.8 -6 4 0.2

0.25 M MgCl2 0 0.0 -7 pe 2.0 M MgCl2 (V) log[Tc] 4.5 M MgCl2 h E 1 bar H (g) -0.2 -4 2 (pe+pH)=0 -8 -0.4 -8 -9 limit of detection -0.6 0.25 M MgCl2 0.25 M MgCl2 -12 2.0 M MgCl2 2.0 M MgCl2 4.5 M MgCl pH max 9 4.5 M MgCl 2 c -0.8 -10 2 02468101214 02468101214 pH pHc c

Fig. 2: Tc(VII)/Tc(IV) redox behaviour in the presence of 1mMNa2S2O4. NaCl solutions: (a) 𝐸h–pH diagram, (b) aqueous concentration of Tc in NaCl solutions. MgCl2 solutions: (c) 𝐸h–pH diagram, (d) aqueous concentration of Tc in MgCl2 solutions. Dashed and solid lines corresponding to the Tc(VII)/Tc(IV) redox borderline calculated for reaction (1) using NEA–TDB with SIT and Pitzer ionic strength corrections, respectively.

𝐸h values are below the thermodynamically expected large differences (up to 5 pe-units) arise between 𝐸h val- Tc(VII)/Tc(IV) borderline. 𝐸h values with large uncer- ues measured in dilute and concentrated MgCl2 solutions, tainties (up to ±100 mV) are observed in near-neutral to hence hinting towards a strong impact of ionic strength, slightly alkaline pH conditions. This observation is likely [Cl−]and/or [Mg2+] on the redox couple controlling the related with the known degradation of Na2 S2O4 under less 𝐸h in this system. alkaline to acidic pH conditions [43, 44]. No significant dif- A significant decrease of Tc concentration indicating ferences were observed between the 𝐸h values measured the reduction of Tc(VII) to Tc(IV) is observed in all sys- in dilute and concentrated NaCl solutions. On the contrary, tems with Na2S2O4 (Figs. 2band2d). The predominance 64 | E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions

Table 3: Tc redox distribution in the aqueous phase of selected stable colloidal Tc–S species has been previously reported samples in NaCl and MgCl2 solutions, as quantified by solvent in the literature [45]. extraction after 10 kD ultrafiltration. Very high [Tc(IV)] (up to 10−5.5 M)isretainedin

solution in concentrated MgCl2 systems, even in the Reducing Background a % c pHc 𝐸h Tc(IV) slightly alkaline pH conditions buffered by the presence system electrolyte (mV)b of Mg(OH)2(cr) or Mg2(OH)3Cl ⋅ 4H2O(cr).Thecom- Na2S2O4 0.5 M NaCl 7.5 −270 98 parison of these results with Sn(II) and Fe(II)/Fe(III) Na2S2O4 0.5 M NaCl 6.6 −120 99 systems suggests that [Tc(IV)] in equilibrium with Na2S2O4 0.5 M NaCl 12.0 −435 92 TcO ⋅1.6H O(s) in the presence of Na S O can be Na2S2O4 5.0 M NaCl 12.7 −445 99 2 2 2 2 4 Sn(II) 0.5 M NaCl 1.9 30 98 overestimated, especially at near-neutral conditions. Sn(II) 0.5 M NaCl 13.7 −760 99 Sn(II) 5.0 M NaCl 2.9 80 92 4.1.2 Sn(II) systems Sn(II) 5.0 M NaCl 14.5 −760 99 Fe(II)/Fe(III) 0.5 M NaCl 2.0 645 0.4 Tin(II) is a very strong reducing agent which keeps sta- Fe(II)/Fe(III) 5.0 M NaCl 2.8 400 0.1 Fe(II)/Fe(III) 5.0 M NaCl 4.5 635 0.4 ble 𝐸h values slightly above the border of water reduction Na2S2O4 2.0 M MgCl2 7.0 30 99 (Figs. 3aand3c). The redox couple Sn(II)/Sn(IV) control- Na2S2O4 4.5 M MgCl2 9.0 −55 99 ling the 𝐸h in this system is strongly impacted by ionic Sn(II) 2.0 M MgCl2 3.7 5 73 − 2+ strength, [Cl ]and/or [Mg ], with pe + pHc values rang- Sn(II) 4.5 M MgCl2 4.0 140 62 ing from 2±1to 7±1in 0.5 M NaCl and 4.5 M MgCl2,re- Sn(II) 4.5 M MgCl2 6.4 −10 85 spectively. Very similar pe + pH values like in 0.5 M NaCl Sn(II) 4.5 M MgCl2 9.0 −215 99 c were recently reported for Sn(II) solutions in NaCl Fe(II)/Fe(III) 0.25 M MgCl2 3.4 205 0.1 0.1 M Fe(II)/Fe(III) 2.0 M MgCl2 3.8 485 0.2 (pe + pHc =2±1)[23]. The very reducing 𝐸h values fixed Fe(II)/Fe(III) 4.5 M MgCl2 4.2 610 0.3 by Sn(II) promote a fast (𝑡1/2 =7d)andcompletereduction Fe(II)/Fe(III) 4.5 M MgCl2 6.4 365 0.6 of Tc(VII) to Tc(IV) in NaCl solutions (Figure 3b). This ob- Fe powder 4.5 M MgCl2 9.0 −195 99 servation is confirmed by solvent extraction, which indi- Fe powder 4.5 M MgCl2 8.9 −125 99 cates the predominance of Tc(IV) (≥ 92%)inallNaClsam- a: ±0.05;b:±50 mV;c:±10%. ples analysed (Table 3). The shape of the Tc(IV) solubility curve obtained from oversaturation in these systems is in excellent agreement with thermodynamic calculations of Tc(IV) in the aqueous phase is confirmed for selected performed at 𝐼=0according with NEA–TDB data selec- samples by solvent extraction (Table 3). This is in very tion (solid line in Figure 3b). This observation strongly sug- good agreement with thermodynamic calculations per- gests that equilibrium conditions have been attained in formed using NEA–TDB and SIT/Pitzer ionic strength cor- the system, as well as hinting towards the expected pre- 2+ + rections as summarized in Table 1. Higher concentrations dominance of TcO 2 ⋅1.6H2O(s) and TcO , TcO(OH) , − of Tc(IV) are retained in the aqueous phase in NaCl solu- TcO(OH) 2(aq) and TcO(OH) 3 as solubility controlling tions at pHc >12. This is consistent with the expected for- solid phase and Tc(IV) hydrolysis species in equilibrium, − mation of an anionic hydrolysis species (e.g. TcO(OH) 3 ), respectively. Tc concentrations determined in the acidic which increases the solubility of TcO 2 ⋅1.6H2O(s) in the pH range are also in very good agreement with the un- strongly alkaline pH region. Similar to previous observa- dersaturation solubility data reported by Hess and co- tions by Kobayashi and co-workers in diluted NaCl so- workers [12] (also shown in Figure 3b). This indicates that −7 lutions and Na2S2O4 [23], ∼10 M Tc(IV) concentra- the same solid phase has been obtained from oversatura- tions are observed in 0.5 M and 5.0 M NaCl solutions tion (present work) and undersaturation [12] conditions. under near-neutral pH conditions. A concentration level Our data also confirm the strong effect of ionic strength of 10−8 –10−9 M is expected for Tc(IV) solubility under on Tc(IV) solubility under acidic conditions, where [Tc]

these conditions according to the equilibrium reaction in equilibrium with TcO 2 ⋅1.6H2O(s) increases 1.5–2 log- TcO 2 ⋅1.6H2O(s) ⇔ TcO(OH)2(aq) + 0.6H2O [24]. This units between 0.5 M and 5.0 M NaCl. Note that in the pH- observation is also in disagreement with data obtained in range of our study (pHc ≥2), the thermodynamic model the present work in Sn(II) systems (see Figure 3b), thus by Hess and co-workers explains the increase of solubility hinting towards a possible impact of Na2S2O4 degradation by ion interaction phenomena, and disregards the forma- on the behaviour of Tc(IV). Note that the formation of very tion of Tc(IV)–Cl complexes. E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions | 65

-4 a) 0.8 b) 12

1 bar O2(g) - (pe+pH)=20.77 0.6 Initial [TcO ] -5 4 8 0.4 "Redox neutral" (pe+pH)=13.8 -6 4 0.2

0.5 M NaCl 0 5.0 M NaCl 0.0 -7 pe (V) h log[Tc] E -0.2 -4 1 bar H (g) 2 -8 Solubility of TcO ⋅ xH O (pe+pH)=0 2 2

(pe+pH)=4.01 -0.4 -8 -9 (pe+pH)=2.0±1 -0.6 limit of detection 0.8 M NaCl (Hess et al., 2004) -12 0.5 M NaCl 5.0 M NaCl (Hess et al., 2004) 0.5 M NaCl, (p.w.) 5.0 M NaCl ) -0.8 -10 5.0 M NaCl, (p.w. 02468101214 02468101214

pHc pHc

-4 c) 0.8 d) 12 1 bar O (g) 2 Initial [TcO- ] (pe+pH)=20.77 0.6 4 -5 8 0.4 12 d "Redox neutral" (pe+pH)=13.8 -6 150 d 4 (pe+pH)=6.5±1 0.2 0.25 M MgCl (pe+pH)=3.8±1 2 395 d 0 2.0 M MgCl 0.0 -7 pe 2 (V) log[Tc] 4.5 M MgCl h

2 E 1 bar H (g) -0.2 -4 2 (pe+pH)=0 -8 -0.4

-8 (pe+pH)=2.7±1 -9 limit of detection -0.6 0.25 M MgCl 0.25 M MgCl2 2 2.0 M MgCl -12 2.0 M MgCl2 2 4.5 M MgCl pH max 9 4.5 M MgCl 2 c -0.8 -10 2 02468101214 02468101214 pH pHc c

Fig. 3: Tc(VII)/Tc(IV) redox behaviour in the presence of 1mMSn(II). NaCl solutions: (a) 𝐸h–pH diagram, (b) aqueous concentration of Tc in NaCl solutions. MgCl2 solutions: (c) 𝐸h–pH diagram, (d) aqueous concentration of Tc in MgCl2 solutions. Dashed and solid lines corresponding to the Tc(VII)/Tc(IV) redox borderline calculated for reaction (1) using NEA–TDB with SIT and Pitzer ionic strength corrections, respectively. Solid grey line in (b) represents the solubility curve of TcO 2 ⋅1.6H2O(s) calculated at 𝐼=0using NEA–TDB. Symbol ⊕ in (d) indicates [Tc(IV)]in4.5 M MgCl2 measured from undersaturation conditions (see text).

Very low [Tc] are observed in 0.25 M MgCl2 at 3.5 ≤ prepared from undersaturation. Within this approach, pHc ≤9(Figure 3d), in good agreement with data in 0.5 M the solubility experiment is started with a macroscopic NaCl. Significantly higher [Tc] are observed in 2.0 M and amount of solid phase contacted with Tc-free super-

4.5 M MgCl2. Both systems are affected by strong kinet- natant solution. A faster attainment of thermodynamic ics (especially for pHc =pHmax ∼9), which likely hinders equilibrium is expected under these conditions, com- thermodynamic equilibrium even after 395 d. This hypoth- pared to the oversaturation approach where precipita- esis was further evaluated by Tc(IV) solubility samples tion of rather ill-defined microcrystalline phases may oc- 66 | E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions

-4 a) 0.8 b) 12

1 bar O2(g) - (pe+pH)=20.77 0.6 Initial [TcO ] -5 4 8 0.4 "Redox neutral" (pe+pH)=13.8 -6 4 0.2

0.5 M NaCl 0 5.0 M NaCl 0.0 -7 pe (V) h log[Tc] E

1 bar H (g) -0.2 -4 2 (pe+pH)=0 -8 -0.4 -8 -9 -0.6 limit of detection

-12 0.5 M NaCl 0.5 M NaCl 5.0 M NaCl 5.0 M NaCl -0.8 -10 02468101214 02468101214

pHc pHc

-4 c) 0.8 d) 12 1 bar O (g) 2 Initial [TcO- ] (pe+pH)=20.77 0.6 4 -5 8 0.4 "Redox neutral" (pe+pH)=13.8 -6 4 0.2

0.25 M MgCl2

0 2.0 M MgCl 0.0 pe 2 -7 (V) h 4.5 M MgCl log[Tc] 12 d 2 E 1 bar H (g) -0.2 60 d -4 2 (pe+pH)=0 -8 150 d -0.4 395 d -8 -9 limit of detection -0.6 0.25 M MgCl 2 0.25 M MgCl2 -12 2.0 M MgCl 2 pH max 9 2.0 M MgCl2 4.5 M MgCl2 c -0.8 4.5 M MgCl -10 2 02468101214 02468101214 pH c pHc

Fig. 4: Tc(VII)/Tc(IV) redox behaviour in the presence of Fe(II)/Fe(III). NaCl solutions: (a) 𝐸h–pH diagram, (b) aqueous concentration of Tc in NaCl solutions. MgCl2 solutions: (c) 𝐸h–pH diagram, (d) aqueous concentration of Tc in MgCl2 solutions. Dashed and solid lines corresponding to the Tc(VII)/Tc(IV) redox borderline calculated for reaction (1) using NEA–TDB with SIT and Pitzer ionic strength corrections, respectively.

cur. Hence, Tc(IV) prepared electrochemically was precip- 4.5 M MgCl2 support the trend observed from oversatu- itated as TcO 2 ⋅1.6H2O(s) [46] in a dilute NaOH solution ration and confirm that no thermodynamic equilibrium is (pHc ∼11), separated from the supernatant and contacted attained after 395 d in these conditions. On the contrary, with two 4.5 M MgCl2 solutions (pHc =4and pHmax)in very similar [Tc] are measured at pHc =4from undersat- the presence of 2mMSn(II). The aqueous concentration uration and oversaturation approaches, thus indicating of Tc(IV) in equilibrium with TcO 2 ⋅1.6H2O(s) as deter- that the same solid phase (expectedly TcO 2 ⋅1.6H2O(s)) mined from undersaturation conditions is marked as ⊕ in is controlling the solubility in both cases and confirming

Figure 3. Lower Tc concentrations observed at pHmax in E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions | 67

-4 a) 0.8 b) 12 1 bar O (g) 2 Initial [TcO-] 0.6 4 (pe+pH)=20.77 -5 8 0.4 "Redox neutral" (pe+pH)=13.8 -6 4 0.5 M NaCl 0.2 5.0 M NaCl

0 0.0 -7 pe

(pe+pH)=4.0±1 (V) log[Tc]

(pe+pH)=2.0±1 h E 1 bar H (g) -0.2 -4 2 (pe+pH)=0 -8 -0.4 -8 -9 -0.6 limit of detection

-12 0.5 M NaCl 0.5 M NaCl 5.0 M NaCl 5.0 M NaCl -0.8 -10 02468101214 02468101214 pH pHc c

-4 c) 0.8 d) 12 1 bar O (g) 2 - (pe+pH)=20.77 0.6 Initial [TcO ] -5 4 8 0.4 "Redox neutral" (pe+pH)=13.8 12 d 4 -6 60 d 0.2

0.25 M MgCl2

0 2.0 M MgCl 0.0 pe -7

2 (V) h 4.5 M MgCl log[Tc] 2 E 1 bar H (g) -0.2 150 d -4 2 (pe+pH)=0 -8 -0.4 -8 -9 limit of detection -0.6 0.25 M MgCl2 0.25 M MgCl2 -12 2.0 M MgCl 2.0 M MgCl 2 pH max 9 2 4.5 M MgCl c 4.5 M MgCl 2 -0.8 -10 2 02468101214 02468101214 pH c pHc

Fig. 5: Tc(VII)/Tc(IV) redox behaviour in the presence of Fe powder. NaCl solutions: (a) 𝐸h–pH diagram, (b) aqueous concentration of Tc in NaCl solutions. MgCl2 solutions: (c) 𝐸h–pH diagram, (d) aqueous concentration of Tc in MgCl2 solutions. Dashed and solid lines corresponding to the Tc(VII)/Tc(IV) redox borderline calculated for reaction (1) using NEA–TDB with SIT and Pitzer ionic strength corrections, respectively. that thermodynamic equilibrium has been also attained in ing UV–VIS/NIR. Note that the solvent extraction method the latter case. used in this work has been reported to extract into the − Solvent extraction analysis of the supernatant solu- organic phase not only MO4 species (M = Re, Mn, Tc, 2− tion in 2.0 M and 4.5 M MgCl2 indicates 62%–99% con- I) but also anionic chloride complexes such as SnCl6 2− tent in Tc(IV), depending upon pH and ionic strength and ZnCl4 [47]. Thus, the hypothetical formation of an- (Table 3). Similar observations in concentrated NaCl so- ionic Tc(IV)–Cl and/or Tc(IV)–OH–Cl species forming in lutions were reported by Hess and co-workers, who con- concentrated Cl-brines may lead to an underestimation firmed the predominance of Tc(IV) in the same samples us- of Tc(IV) concentration in the aqueous phase. Dedicated 68 | E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions

undersaturation solubility studies with Tc(IV) in dilute to ment with the thermodynamic borderline for 50:50%

concentrated NaCl, MgCl2 and CaCl2 are underway to Tc(VII)/Tc(IV) redox distribution calculated with NEA– assess these uncertainties and derive complete chemical, TDB and SIT/Pitzer ionic strength corrections as described thermodynamic and activity models valid also in concen- in Sect. 2. These observations are also in line with data re- trated brines. ported by Zachara and co-workers, who observed a rapid reduction of Tc(VII) in the presence of Fe(II) at pH > 6.8 [14]. In contrast to this, Cui et al. [13]reportedtheab- 4.1.3 Fe(II)/Fe(III) systems sence of Tc(VII) reduction in Fe(II) systems with pH ≤ 7.5. In contrast to the investigated NaCl systems,

Figure 4 shows the redox behaviour of Tc(VII)/Tc(IV) in a significant increase of experimental 𝐸h is observed the presence of Fe(II)/Fe(III) in dilute to concentrated with increasing MgCl2 concentration (Figure 4c). As oc- NaCl and MgCl2 solutions. In analogy to Na2S2O4 sys- curring for Sn(II) systems, this observation indicates that tems, the increase of ionic strength induces only a minor the Fe(II)/Fe(III) redox couple is significantly influenced − 2+ effect on the 𝐸h values measured in NaCl solutions. For this by ionic strength, [Cl ]and/or [Mg ]. Concentration background electrolyte system, the Fe(II)/Fe(III) redox of Tc in solution remains unaffected for samples in the pair define 𝐸h values above the calculated Tc(VII)/Tc(IV) acidic pH region with 𝐸h values above the calculated borderline under acidic pH conditions. At pHc =2–3, Tc(VII)/Tc(IV) borderline (Figure 4d), thus indicating most of the initial Fe(II)aq and Fe(III)aq are retained the predominance of Tc(VII). These observations are in solution. For these conditions, 𝐸h values can be further confirmed by solvent extraction analysis as thermodynamically calculated considering the chemi- summarized in Table 3.ThesampleatpHc ∼3.5 and 2+ 3+ − cal reaction Fe ⇔Fe +e. Assuming [Fe(II)]0 = 𝐸h =0.2Vin 0.25 M MgCl2 is the only sample in this [Fe(II)]aq and [Fe(III)]0 = [Fe(III)]aq , and considering pH region below the Tc(VII)/Tc(IV) borderline for which 2+ 3+ 3−n [Fe(II)]aq =[Fe ] and [Fe(III)]aq =[Fe ]+𝛴[FeCln ]+ unexpectedly no decrease of [Tc] is observed within 395 d. 3−n therm 𝛴[Fe(OH)n ] at pHc =2in 0.5 M NaCl,an𝐸h value In general, a very good agreement between experimental of 0.63 V can be calculated based on Eq. (7) and using data and thermodynamic calculations is observed for ∗ ∘ log 𝐾 and SIT ion interaction parameters summarized in MgCl2 solutions under near-neutral to alkaline pH condi- Tables A1 and A2 in the Appendix. tions (6≤pHc ≤9). All samples in this pH region with 𝐸h values below the Tc(VII)/Tc(IV) borderline show a clear pe = log(Fe(III)tot ⋅𝛾Fe3+) − log(Fe(II)tot ⋅𝛾Fe2+) decrease in [Tc] and consequent reduction to Tc(IV). As ∘ ∗ − 𝑛 + −𝑚 in the case of Sn(II) systems, slow kinetics are observed −log𝐾Fe2+Fe3+ −log(1+∑ 𝛽1,𝑛,𝑚 ⋅[Cl ] ⋅[H ] ) (7) in 4.5 M MgCl2 at pHc =pHmax,where[Tc]aq decreases from ∼10−7.3 M to ∼10−8.3 M within 395 d.Theverylow ∗ where 𝛽1,𝑛,𝑚 are the conditional equilibrium constants concentration of Tc reached in these conditions compared for the formation of Fe(III) hydroxo and chloro complexes to Sn(II) systems and undersaturation solubility experi- − under given pHc and [Cl ] conditions. This value is in ments (see Figure 3) indicate that sorption of Tc(IV) on the excellent agreement with the measured redox potential forming magnetite may have occurred.

(0.64 ± 0.05 V), thus supporting the validity of the 𝐸h values experimentally determined in the Fe systems. 𝐸h values below the calculated Tc(VII)/Tc(IV) borderline are 4.1.4 Corroding Fe powder systems measured in the near-neutral and alkaline pH regions. Un- der these conditions, the formation of a black Fe precip- Very reducing 𝐸h values are measured in the presence of itate (likely magnetite) occurs. Thermodynamic calcula- Fe corroding powder in diluted NaCl and MgCl2 systems at tions considering the formation of this solid phase predict pHc ≤9(Figs. 5aand5c). These values are in good agree- the decrease of 𝐸h towards more reducing conditions, al- ment with previous studies under analogous experimen- though the exact redox potential is strongly dependent on tal conditions at pe + pH = 2 ± 1 [48]. A fast reduction of

the particle size of the solid forming (and thus on the se- Tc(VII) to Tc(IV) is observed for all samples at pHc ≤9,as ∗ ∘ lected log 𝐾 𝑠,0). indicated by the decrease of [Tc] (Figure 5b) and confirmed A clear decrease of Tc concentration occurs under by solvent extraction (Table 3). Above pHc ∼10,thereduc- near-neutral to hyperalkaline pH conditions indicating the ing capacity of Fe powder is retained only for ∼60days. reduction of Tc(VII), whereas [Tc] remains ∼10−5 M in A similar behaviour of Fe powder in 0.1 M NaCl solutions the acidic pH region (Figure 4b). This is in good agree- was recently reported by Kobayashi et al. (2013) [23]. 𝐸h E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions | 69 values measured after this time are situated above the 4.1). This is consistent with the expected shift of the sorp-

Tc(VII)/Tc(IV) redox borderline, corresponding to the tion edge of Tc(IV) towards higher pHc values with increas- predominance of Tc(VII) in solution with [Tc] ∼ 10−5 M. ing ionic strength, similar to observations made for hydro-

These results again show that reliable 𝐸h values can be lysis ([12], p.w.). Note that a similar effect of ionic strength measured in Fe redox-buffered systems (also at high 𝐼). on sorption was recently reported by Schnurr et al. for the They also indicate that the NEA–TDB selection of thermo- uptake of Eu(III) by illite [49], where a much stronger de- dynamic data in combination with SIT/Pitzer ion interac- crease of sorption in the presence of divalent cations (Ca2+ tion parameters reported in the present study (see Sect. 2) and Mg2+) was observed. offer an accurate tool to predict Tc redox behaviour in dilute to concentrated brine solutions. 4.2.2 XANES analysis

4.2 Tc reduction and sorption by Fe phases Figure 6 shows the X-ray absorption near-edge structure (XANES) spectra measured at the Tc 𝐾-edge, for Fe(II) 4.2.1 Wet chemistry data minerals reacted with Tc(VII). The spectra were collected at a sample temperature of 10–15 K in He atmosphere Magnetite and mackinawite synthesized in this work were to prevent eventual changes of Tc oxidation state by at- positively identified by their XRD patterns (JCPDS–files mospheric O2 or by O-radicals produced by the high X- 19-0629 and 24-0073) after the equilibration with the cor- ray photon flux. All mineral samples have an edge posi- responding NaCl and MgCl2 solutions for 2 weeks (data tion near 21.058 keV and a white line position at 21.065 to not shown). This confirms that no solid phase transforma- tion took place in the concentrated brines at the given pH conditions. 𝐸h and pHc values measured in the Fe min- VII IV eral suspensions (magnetite, mackinawite and siderite) after 4 weeks equilibration time are summarized in Ta- Magnetite 5.0 M NaCl 4.5 M MgCl ble 4.Inallcases,experimental𝐸h values are below the 2 observed Tc(VII)/Tc(IV) reduction borderline. Analog to previous observations for other reducing systems reported Mackinawite 5.0 M NaCl

4.5 M MgCl2 in this work, 𝐸h values in 4.5 M MgCl2media are significantly higher than in 5.0 M NaCl (∼2pe-units) at the same Siderite 5.0 M NaCl pH , characteristic for a redox couple strongly impacted c 4.5 M MgCl2 by ionic strength, [Cl−]and/or [Mg2+]. 𝑅 values (𝑅 = [Tc]s ⋅ 𝑉 ,inLkg−1) were derived from d d [Tc]aq 𝑚 experimental data to assess Tc sorption properties in TcO - saline samples (see Table 4). A stronger uptake is observed 4 −1 in 5.0 M NaCl (4.6 ≤ log 𝑅d(L kg )≤7.2)comparedto −1 sorption samples in 4.5 M MgCl2 (3.0 ≤ log 𝑅d(L kg )≤ Normalized Absorption

Table 4: pHc, 𝐸h and log 𝑅d values determined for the uptake of Tc by Fe minerals (after 4 weeks of equilibration time).

a Fe mineral Background pHc 𝐸h log 𝑅d electrolyte (mV)b (Lkg−1)c

Magnetite 5.0 M NaCl 9.6 −140 4.7 21.00 21.05 21.10 21.15 21.20 Magnetite 4.5 M MgCl2 8.7 10 3.0 Energy [keV] Mackinawite 5.0 M NaCl 8.7 −290 7.2 Mackinawite 4.5 M MgCl2 8.3 −150 4.1 Fig. 6: Tc 𝐾-edge XANES spectra of Tc(VII) reacted with magnetite, Siderite 5.0 M NaCl 8.7 −175 6.0 mackinawite and siderite (see Table 4 for sample description). − Siderite 4.5 M MgCl2 8.3 −25 3.8 Bottom XANES spectrum corresponding to a TcO 4 reference sample. The reader is referred to previous publications for the a: ±0.05;b:±100 mV;c:±10% for log 𝑅d ≤3; ±50% for log 𝑅d ≥3. XANES spectra of a TcO 2(s) reference [20, 50, 51]. 70 | E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions

21.070 eV in line with Tc(IV), while the distinct pre-edge of new Tc(IV) aqueous species (i.e. Tc(IV)–OH–Cl or Mg– peak of Tc(VII) at 21.050 keV is absent in these samples. Tc(IV)–OH) cannot be ruled out.

Accordingly, Tc(VII) has been reduced to Tc(IV) in all the Tc(VII) is reduced and strongly sorbed (log 𝑅d ≥ samples. The edge and white line positions as well as the 3Lkg−1) by magnetite, mackinawite and siderite un- fine structure is furthermore suggesting a coordination to der mildly alkaline conditions in concentrated NaCl and

O atoms; therefore, we find no evidence for the (partial) MgCl2 solutions. Analog to the behaviour observed in coordination of Tc(IV) by S atoms in the high-salt macki- homogeneous systems, Tc(IV) is stabilized in the aque- nawite systems. This is in contrast to previous work, where ous phase in concentrated MgCl2 solutions resulting in formation of a TcS 2-like phase was found after precipi- a weaker sorption compared to dilute systems or concen- tating mackinawite in the presence of pertechnetate [18, trated NaCl solutions. To our knowledge, this is the first 19]. Tc(IV) coordinated to S was also found after sorp- experimental study providing quantitative data on the up- tion of pertechnetate to mackinawite at an ionic strength take of Tc by corrosion products of Fe in concentrated brine of 0.1 M [18, 19, 23], pointing to a decisive role of ionic solutions. strength on the reaction product, but this needs confirma- tion by more detailed investigations. Appendix

5Conclusion Table A1: Thermodynamic data for Fe(II)/Fe(III) used to calculate Fe systems under acidic conditions. A comprehensive understanding of technetium redox processes is essential for the reliable estimation of solubil- Reaction log ∗𝐾∘ Reference ity limits and technetium source term in the context of 2+ 3+ − Fe ⇔Fe +e −13.05 NEA-TDB [52] 3+ − 2+ nuclear waste disposal. In the case of nuclear waste dis- Fe +Cl ⇔FeCl 1.52 NEA-TDB [52] 3+ − + posal in rock salt or other sites with potentially high ionic Fe +2Cl ⇔FeCl2 2.22 NEA-TDB [52] 3+ 2+ + strength solutions present, a detailed understanding of Fe +H2O⇔Fe(OH) +H −2.15 NEA-TDB [52] Tc redox processes under these specific conditions is re-

quired. Table A2: SIT ion interaction coefficients for Fe species in NaCl Tc(VII)/Tc(IV) redox behaviour was systematically solutions.

studied in dilute to concentrated NaCl and MgCl2 solutions. In the investigated redox-buffered systems, the as- 𝑖𝑗𝑖𝑗 References 𝜀 (mol kg−1) sessment of experimental pH and 𝐸h values by thermody- + − namic calculations using NEA–TDB data provides an ac- H Cl 0.12 NEA-TDB [24] − curate tool to predict the redox distribution of Tc in aque- OH Na 0.04 NEA-TDB [24] 3+ − 3+ − a ous solutions. The use of SIT and Pitzer ion interaction Fe Cl 0.24 estimated from 𝜀(Fe ,ClO4 ) = 0.56 2+ − 2+ − a Fe Cl 0.16 estimated from 𝜀(Fe ,ClO4 ) = 0.38 coefficients proposed in this work allows to extend the 2+ − FeCl Cl 0.15 estimated by charge correlationb predictions to concentrated NaCl and MgCl brine sys- + − b 2 FeCl2 Cl 0.05 estimated by charge correlation 2+ − tems. A systematic increase of experimental 𝐸h values is FeOH Cl 0.15 estimated by charge correlationb related to increasing ionic strength, especially in concen- − a: The SIT coefficients for the interaction with Cl are estimated from trated MgCl2 brines. This observation reflects the impact − the corresponding SIT coefficients for the interaction with ClO4 [24] of ionic strength, choride- and/or magnesium concentra- z+ − z+ − according with the correlation: 𝜀(M ,Cl ) = 0.38 𝜀(M ,ClO4 )± −1 tion on the redox couple controlling the redox potential 0.02 mol kg [53]; of the solution. In redox-buffered systems where the two b: [54]. members of the redox couple (and corresponding thermodynamics) are known (i.e. Fe(II)/Fe(III) under acidic conditions), measured and thermodynamically calculated 𝐸h Acknowledgement: This work was partially supported by values show a very good agreement. the German Federal Ministry of Economics and Technol-

The solubility of Tc(IV) in concentrated MgCl2 solu- ogy (BMWi) under the project of VESPA. The staff of the tions (2.0 M and 4.5 M) is significantly enhanced com- ROBL beamline at ESRF and Dr. M. Marques Fernandes pared with diluted NaCl and MgCl2 systems. This can (PSI–LES, Switzerland) are kindly acknowledged for the be explained by the stabilization of charged hydrolysis assistance during EXAFS measurements. Thanks are ex- species at elevated ionic strengths, although the formation tended to M. Böttle and Dr. D. Fellhauer (KIT–INE) for sup- E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions | 71 port in the preparation of the Fe phases and Pitzer calcu- nature of solid phase redox products. Geochim. Cosmochim. lations. Acta 71, 2137–2157 (2007). 15. Cui, D. Q., Eriksen, T. E.: Reduction of pertechnetate in solution by heterogeneous electron transfer from Fe(II)-containing geo- logical material. Environ. Sci. Technol. 30, 2263–2269 (1996). References 16. Geraedts, K., Bruggeman, C., Maes, A., Van Loon, L. R., Ross- berg, A., Reich, T.: Evidence for the existence of Tc(IV)-humic 1. Blanchard, D. L., Brown, G. N., Conradson, S. D., Fadeff, S. K., substance species by X-ray absorption near-edge spec- Golcar, G. R., Hess, N. J., Klinger, G. S., Kurath, D. E.: Tech- troscopy. Radiochim. Acta 90, 879–884 (2002). netium in alkaline, high-salt radioactive tank waste supernate: 17. Maes, A., Geraedts, K., Bruggeman, C., Vancluysen, J., Ross- preliminary chracterization and removal, PNNL-11386 UC-2030 berg, A., Hennig, C.: Evidence for the interaction of technetium (1997). colloids with humic substances by X-ray absorption spec- 2. Kupfer, M. J., Boldt, A. L., Hodgson, K. M., Shelton, L. W., Simp- troscopy. Environ. Sci. Technol. 38, 2044–2051 (2004). son, B. C., Watrous, R. A., Leclaire, M. D., Borsheim, G. L., 18. Wharton, M. J., Atkins, B., Charnock, J. M., Livens, F. R., Pat- Winward, R. T., Higley, B. A., Orme, R. M., Colton, N. G., Lam- trick, R. A. D., Collison, D.: An X-ray absorption spectroscopy bert, S. L., Place, D. E., Schulz, W. W.: Standart inventories study of the coprecipitation of Tc and Re with mackinawite of chemicals and radionuclides in Hanford site tank wastes, (FeS). Appl. Geochem. 15, 347–354 (2000). HNF-SD-WM-740 (1999). 19. Livens, F. R., Jones, M. J., Hynes, A. J., Charnock, J. M., 3. Altmaier, M., Neck, V., Lützenkirchen, J., Fanghänel, T.: Solubil- Mosselmans, J. F. W., Hennig, C., Steele, H., Collison, D., ity of plutonium in MgCl2 and CaCl2 solutions in contact with Vaughan, D. J., Pattrick, R. A. D., Reed, W. A., Moyes, L. N.: metallic iron. Radiochim. Acta 97, 187–192 (2009). X-ray absorption spectroscopy studies of reactions of tech- 4. Neck, V., Altmaier, M., Rabung, T., Lützenkirchen, J., netium, uranium and neptunium with mackinawite. J. Environ. Fanghänel, T.: Thermodynamics of trivalent actinides and Radioact. 74, 211–219 (2004). neodymium in NaCl, MgCl2,andCaCl2 solutions: Solubility, 20. Liu, Y., Terry, J., Jurisson, S.: Pertechnetate immobilization with hydrolysis, and ternary Ca-M(III)-OH complexes. Pure Appl. amorphous iron sulfide. Radiochim. Acta 96, 823–833 (2008). Chem. 81, 1555–1568 (2009). 21. Llorens, I. A., Deniard, P., Gautron, E., Olicard, A., Fattahi, M., 5. Fellhauer, D., Neck, V., Altmaier, M., Lützenkirchen, J., Jobic, S., Grambow, B.: Structural investigation of coprecipita- Fanghänel, T.: Solubility of tetravalent actinides in alkaline tion of technetium-99 with iron phases. Radiochim. Acta 96, 4+ CaCl2 solutions and formation of Ca4[An(OH)8] complexes: 569–574 (2008). a study of Np(IV) and Pu(IV) under reducing conditions and 22. Peretyazhko, T. S., Zachara, J. M., Kukkadapu, R. K., the systematic trend in the An(IV) series. Radiochim. Acta 98, Heald, S. M., Kutnyakov, I. V., Resch, C. T., Arey, B. W., 541–548 (2010). Wang, C. M., Kovarik, L., Phillips, J. L., Moore, D. A.: Pertech- − 6. Russell, C. D., Cash, A.: Products of pertechnetate reduction in netate (TcO 4 ) reduction by reactive ferrous iron forms in natu- complexing media. J. Nucl. Med. 19, 694–694 (1978). rally anoxic, redox transition zone sediments from the Hanford 7. Grassi, J., Rogelet, P., Devynck, J., Tremillon, B.: Radiopo- Site, USA. Geochim. Cosmochim. Acta 92, 48–66 (2012). larography of technetium(VII) in acidic medium. J. Electroanal. 23. Kobayashi, T., Scheinost, A. C., Fellhauer, D., Gaona, X., Alt- Chem. 88, 97–103 (1978). maier, M.: Redox behavior of Tc(VII)/Tc(IV) under various 8. Grassi, J., Devynck, J., Tremillon, B.: Electrochemical studies reducing conditions in 0.1 M NaCl solutions. Radiochim. Acta of technetium at a mercury-electrode. Anal. Chimica Acta 107, 101, 323–332 (2013). 47–58 (1979). 24. Guillaumont, R., Fanghänel, T., Neck, V., Fuger, J., Palmer, D. A., 9. Poineau, F., Fattahi, M., Den Auwer, C., Hennig, C., Gram- Grenthe, I., Rand, M. H.: Update on the Chemical Thermody- bow, B.: Speciation of technetium and rhenium complexes namics of Uranium, Neptunium, Plutonium, Americium and by in situ XAS-electrochemistry. Radiochim. Acta 94, 283–289 Technetium. OECD Nuclear Energy Agency (ed.), Vol. 5, North- (2006). Holland, Amsterdam, Elsevier (2003). 10. Owunwanne, A., Marinsky, J., Blau, M.: Charge and nature of 25. Rard, J. A., Rand, M. H., Anderegg, G., Wanner, H.: Chem- technetium species produced in reduction of pertechnetate by ical Thermodynamics of Technetium.(Sandino,M.C.A., stannousion.J.Nucl.Med.18, 1099–1105 (1977). Östhols, E., eds.) Vol. 3, North-Holland, Amsterdam, Elsevier 11. Warwick, P., Aldridge, S., Evans, N., Vines, S.: The solubility (1999). of technetium(IV) at high pH. Radiochim. Acta 95,709–716 26. Chotkowski, M., Czerwinski, A.: Electrochemical and spectro- (2007). electrochemical studies of pertechnetate electroreduction in 12. Hess, N. J., Xia, Y. X., Rai, D., Conradson, S. D.: Thermodynamic acidic media. Electrochim. Acta 76, 165–173 (2012). model for the solubility of TcO 2 ⋅ xH2O(am) in the aqueous 27. Ciavatta, L.: The specific interaction theory in equilibrium- + − + − Tc(IV)-Na -Cl -H -OH -H2O system. J. Solut. Chem. 33, 199– analysis – some empirical rules for estimating interaction co- 226 (2004). efficients of metal-ion complexes. Ann. Chim. 80, 255–263 13. Cui, D. Q., Eriksen, T. E.: Reduction of pertechnetate by ferrous (1990). iron in solution: Influence of sorbed and precipitated Fe(II). 28. Grenthe, I., Wanner, H., Östhols, M.: Guidelines for the Extrap- Environ. Sci. Technol. 30, 2259–2262 (1996). olation to Zero Ionic Strength. Nuclear Energy Agency, OECD 14. Zachara, J. M., Heald, S. M., Jeon, B. H., Kukkadapu, R. K., (2000). Liu, C. X., McKinley, J. P., Dohnalkova, A. C., Moore, D. A.: Re- 29. Pitzer, K. S.: Activity Coefficients in Electrolyte Solutions. duction of pertechnetate [Tc(VII)] by aqueous Fe(II) and the Chap. 3, Boca Raton, FL (1991). 72 | E. Yalçıntaş et al., Redox chemistry of Tc(VII)/Tc(IV) in dilute to concentrated NaCl and MgCl2 solutions

30. Harvie, C. E., Moller, N., Weare, J. H.: The prediction of mineral (methylimino)bis(𝑁,𝑁-dioctylacetamide) complex with − − 1 solubilities in natural-waters – the Na-K-Mg-Ca-H-Cl-SO4-OH- Re(VII)O4 and Tc(VII)O4 by H NMR, EXAFS, and IR spec- ∘ HCO3-CO3-CO2-H2O system to high ionic strengths at 25 C. troscopy. Inorg. Chem. 51, 5814–5821 (2012). Geochim. Cosmochim. Acta 48,723–751(1984). 42. Meyer, R. E., Arnold, W. D.: The electrode potential of the Tc(IV)- 31. Konnecke, T., Neck, V., Fanghänel, T., Kim, J. I.: Activ- Tc(VII) couple. Radiochim. Acta 55, 19–22 (1991). ity coefficients and Pitzer parameters in the systems 43. Lem, W. J., Wayman, M.: Decomposition of aqueous dithionite. + + − − − ∘ Na /Cs /Cl /TcO4 or ClO4 /H2O at 25 C.J.Solut.Chem. 1. Kinetics of decomposition of aqueous sodium dithionite. 26, 561–577 (1997). Can. J. Chem. 48, 776 (1970). 32. Neck, V., Konnecke, T., Fanghänel, T., Kim, J. I.: Pitzer 44. Wayman, M., Lem, W. J.: Decomposition of aqueous dithionite. parameters for the pertechnetate ion in the system 2. A reaction mechanism for decomposition of aqueous sodium + + 2+ 2+ − 2− − ∘ Na /K /Mg /Ca /Cl /SO4 /TcO4 /H2O at 25 C.J.Solut. dithionite. Can. J. Chem. 48, 782 (1970). Chem. 27, 107–120 (1998). 45. Stern, H. S., Mcafee, J. G., Subraman, G.: Preparation distribu- 33. Kirsch, R., Fellhauer, D., Altmaier, M., Neck, V., Rossberg, A., tion and utilization of technetium-99m-sulfur colloid. J. Nucl. Fanghänel, T., Charlet, L., Scheinost, A. C.: Oxidation state and Med. 7, 665 (1966). local structure of plutonium reacted with magnetite, macki- 46. Eriksen, T. E., Ndalamba, P., Bruno, J., Caceci, M.: The solubility

nawite and chukanovite. Environ. Sci. Technol. 45, 7267–7274 of TcO 2 ⋅ nH2O in neutral to alkaline-solutions under constant

(2011). 𝑝CO2 . Radiochim. Acta 58–59, 67–70 (1992). 34. Altmaier, M., Metz, V., Neck, V., Muller, R., Fanghänel, T.: Solid- 47. Morrison, G. H., Freiser, H.: Solvent Extraction in Analytical

liquid equilibria of Mg(OH)2(cr) and Mg2(OH)3Cl ⋅ 4H2O(cr) Chemistry. John Willey & Sons, Inc., London (1957). ∘ in the system Mg-Na-H-OH-O-Cl-H2O at 25 C. Geochim. Cos- 48. Felmy, A. R., Rai, D., Schramke, J. A., Ryan, J. L.: The solubility mochim. Acta 67, 3595–3601 (2003). of plutonium hydroxide in dilute-solution and in high-ionic- 35. Bischofer, B. P., Hagemann, S., Hühne, C., Schönwiese, D., strength chloride brines. Radiochim. Acta 48, 29–35 (1989). Scharge, T.: Influence of chloride concentrations on the signal 49. Schnurr, A., Marsac, R., Rabung, T., Lützenkirchen, J., Geck- of redox electrode and on UV-spectrophotometric determina- eis, H.: Investigation of actinide and lanthanide sorption on 𝑡ℎ tion of Fe species. In 1st Annual Workshop Proceedings 7 EC clay minerals under saline conditions, in: Migration Confer- FP-Recosy CP, Barcelona (2009). ence, Brighton, United Kingdom (2013). 36. Scharge, T., Bischofer, B. P., Hagemann, S., Schönwiese, D.: 50. Liu, Y., Terry, J., Jurisson, S.: Pertechnetate immobilization in Spectrophotometric and potentiometric determination of the aqueous media with hydrogen sulfide under anaerobic and redox potential in solutions of high ionic strength. In: 2nd An- aerobic environments. Radiochim. Acta 95, 717–725 (2007). nual Workshop Proceeding 7th EC FP-Recosy CP, Cyprus (2010). 51. Um, W., Chang, H. S., Icenhower, J. P., Lukens, W. W., 37. Barry, P. H.: JPCALC, a software package for calculating liquid- Serne, R. J., Qafoku, N. P., Westsik, J. H., Buck, E. C., junction potential corrections in patch-clamp, intracellular, Smith, S. C.: Immobilization of 99-technetium(VII) by Fe(II)- epithelial and bilayer measurements and for correcting junc- goethite and limited reoxidation. Environ. Sci. Technol. 45, tion potential measurements. J. Neurosci. Methods 51, 107–116 4904–4913 (2011). (1994). 52. Lemire, R. J., Berner, U., Musikas, C., Palmer, D. A., Taylor, P., 38. Omori, T., Muraoka, Y., Suganuma, H.: Solvent-extraction mech- Tochiyama, O.: Chemical thermodynamics of iron, ed. OECD anism of pertechnetate with tetraphenylarsonium chloride. J. Nuclear Energy Agency. Vol. 13a. Issy-les-Moulineaux, France Radioanal. Nucl. Chem.-Art. 178, 237–243 (1994). (2013). 39. Kopunec, R., Abudeab, F. N., Skraskova, S.: Extraction of 53. Neck, V., Altmaier, M., Fanghänel, T.: Ion interaction (SIT) co- 4+ pertechnetate with tetraphenylphosphonium in the pres- efficients for the Th ion and trace activity coefficients in ence of various acids, salts and hydroxides. J. Radioanal. Nucl. NaClO4, NaNO3 and NaCl solution determined by solvent ex- Chem. 230, 51–60 (1998). traction with TBP. Radiochim. Acta 94, 501–507 (2006). 40. Gaona, X., Dähn, R., Tits, J., Scheinost, A. C., Wieland, E.: Up- 54. Hummel, W.: Ionic strength corrections and estimation of SIT take of Np(IV) by C-S-H phases and cement paste: an EXAFS ion interaction coefficients, PSI report, TM-44-09-01, Paul study. Environ. Sci. Technol. 45, 8765–8771 (2011). Scherrer Institut, Villigen (Switzerland) (2009). 41. Saeki, M., Sasaki, Y., Nakai, A., Ohashi, A., Banerjee, D., 󸀠 Scheinost, A. C., Foerstendorf, H.: Structural study on 2,2 -