Using porous iron composite (PIC) material to immobilize rhenium as an analogue for technetium

Environment International 128 (2019) 379–389

Contents lists available at ScienceDirect

Environment International journal homepage: www.elsevier.com/locate/envint

Using porous iron composite (PIC) material to immobilize rhenium as an analogue for technetium

T

⁎

Fanny Coutelot , Robert J. Thomas, John C. Seaman Savannah River Ecology Laboratory, Aiken, SC, USA The University of Georgia, Athens, GA, USA

A R T I C LE I N FO

A B S T R A C T

Handling Editor: Adrian Covaci

Technetium (99Tc), a uranium-235 (235U) and plutonium-239 (239Pu) fission product, is a primary risk driver in low level radioactive liquid waste at U.S. Department of Energy sites. Previous studies have shown success in using Zero Valent Iron (ZVI) to chemically reduce and immobilize redox sensitive groundwater contaminants. Batch and column experiments were performed to assess the ability of a novel porous iron composite material (PIC) to immobilize Tc(VII) in comparison with two commercial Fe oxide sorbents and reagent grade ZVI in the presence and absence of NO3−, a competing oxidized species that is often found in high concentrations in liquid nuclear waste. Perrhenate (ReO4−) was used as a non-radioactive chemical analogue for pertechnetate (TcO4−) under both oxic and anoxic test conditions. The PIC powder was the most effective at immobilizing Re(VII) under all batch test conditions. The presence of nitrate (NO3−) slowed the removal of ReO4− from solution, presumably through chemical reduction and precipitation. Even so, the PIC and ZVI were effective at removing both Re(VII) and NO3− completely from solution. Nitrate was reduced to NH3 with very little nitrite (NO2−) buildup during equilibration. Significant Re immobilization was observed in the column tests containing PIC sorbent, even though inlet solutions were in equilibrium with O2. The presence of NO3− hastened Re breakthrough, while NO3− reduction to NH3 was observed. The results suggest that PIC and ZVI would be the most effective at the removal of TcO4− from contaminated groundwater sites.

Keywords: Technetium Rhenium Nitrate Zero-valent iron Permeable reactive barriers Column experiment

1. Introduction Technetium-99 (99Tc), one of several radioactive isotopes of technetium (Tc), is a beta emitter (β− ≈ 249 keV) with a half-life of 211,000 years that decays to form stable ruthenium-99 (99Ru). Technetium-99 and iodine-129 (129I, t1/2 = 15.7 million years) are long-term risk drivers at Department of Energy (DOE) facilities because of their long half-lives and mobility in the subsurface environment (Hu et al., 2010; Icenhower et al., 2010). Technetium-99 is a major fission product of 235U and 239Pu, with approximately 1 kg of 99Tc generated per ton of U in a conventional boiling water reactor (Artinger et al., 2003; Bruno and Ewing, 2006; Meena and Arai, 2017; Tagami, 2003). Significant amounts of 99Tc were also generated during the production of nuclear weapons materials, mainly Pu and 3H (Darab and Smith, 1996). Technetium is a redox-sensitive element that can be found in oxidation states from +2 to +7, of which Tc(IV) and Tc(VII) are the most common in the natural environment (Meena and Arai, 2017; Warwick et al., 2007). Under aerobic conditions the highly water-soluble

⁎

pertechnetate (TcO4−) anion will be the predominant species, while Tc (IV) can precipitate as oxides/oxy-hydroxides and sulfides, e.g., TcO2(s) and TcO2·nH2O(s), depending upon pH and the presence and absence of complexing agents (Cantrell and Williams, 2012; Cantrell and Williams, 2013; Coughtrey et al., 1983; Lieser, 1993). Pertechnetate [i.e., Tc(VII)] sorption under aerobic conditions is generally quite limited. Previous laboratory studies evaluating TcO4− sorption on highly weathered sediments reported Kd values ranging from −0.13 to 0.29 mL g−1 (Kaplan et al., 2000). Further, Tc(VII) Kd values increased with decreasing pH, which was attributed to greater sorption of the pertechnetate anion by amphoteric Fe and Al oxides, with negative Kd values due to anion exclusion effects for soils with appreciable cation exchange capacity (Kaplan et al., 1998). Another study evaluating Tc(VII) partitioning under oxic conditions for 20 sediment samples from the DOE Hanford Site reported Kd values ranging from −0.04 to 0.01 mL g−1, indicating no significant sorption under oxic conditions in less weathered subsurface materials (Kutynakov and Parker, 1998). Microbial activity can also affect Tc redox state and, thereby, affects mobility in the environment. A wide variety of anaerobic microbes have

Corresponding author at: Savannah River Ecology Laboratory, Aiken, SC, USA. E-mail address: [email protected] (F. Coutelot).

https://doi.org/10.1016/j.envint.2019.05.001 Received 21 December 2018; Received in revised form 30 April 2019; Accepted 1 May 2019 Available online 09 May 2019 0160-4120/ © 2019 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).

Environment International 128 (2019) 379–389

F. Coutelot, et al.

to reduction than pertechnetate. Thus, under reducing conditions, Re (VII) is more difficult to reduce than Tc(VII), and more readily subject to oxidation, making it a conservative chemical analogue for predicting Tc(VII) behavior in response to transient redox conditions. To address the limitations associated with ZVI, a high-surface area porous iron composite (PIC) material consisting of both Fe0 and Fe oxides was recently developed by Höganäs Environment Solutions Inc. (Cary, NC) for use as both a reactive filtration material for water treatment applications and possibly as an in situ PRB (Hu, 2016). The reactive mechanisms responsible for contaminant treatment are quite similar to conventional ZVI; however, the PIC materials have a larger reactive surface area, and appear to be more efficient at removing target contaminants. Using the PIC materials, Allred (2012) demonstrated the ability of the materials to reduce a significant amount of NO3−, sorb orthophosphate (PO43−), and chemically degrade the herbicide atrazine in agricultural drainage waters (Allred, 2012). More recently, Seaman et al. (2018) demonstrated the ability of PIC materials to effectively address a range of redox sensitive contaminants as a municipal water treatment strategy. The objective of this study was to assess the ability of four commercially available Fe materials (i.e., a novel porous iron composite (PIC) material, reagent grade Zero Valent Iron (ZVI), and two commercial Fe oxides (GFH1 and GFH2)) to immobilize Re(VII) from contaminated groundwater in the presence and absence of other common oxidants, such as nitrate (NO3−) and dissolved O2, that are likely present in 99Tc contaminated systems. To achieve this objective, a series of batch experiments were carried out first to determine which materials were best at immobilizing Re from contaminated water under extended, well-mixed conditions, followed by dynamic leaching experiments. These experiments were performed in both oxic and anoxic environments to determine the impact of O2, as well as in the presence and absence of NO3−.

been shown to reduce Tc(VII), even in heavily contaminated sediment, such as the Hanford Site, Washington State, viable populations of microorganisms have been documented. There is a prospect that the oxidized form of technetium, Tc(VII)O4, may be reduced to hydrous Tc(IV) O2 like solids and the mobility of technetium thus hindered (Abdelouas et al., 2005; Beasley and Lorz, 1986; Icenhower et al., 2010; Newsome et al., 2019). A great deal of research has focused on the remediation of groundwater contaminants (i.e., NO3− (Hwang et al., 2011; Jiang et al., 2011; Ryu et al., 2011), Cr(VI) (Guan et al., 2011; Lv et al., 2012; Mitra et al., 2011; Qiu et al., 2012), Pb (Cantrell et al., 1995; Zhang et al., 2011), U (Gu et al., 1998), Tc (Lawter et al., 2018; Newsome et al., 2019; Truex et al., 2017)) by using zero-valent iron (ZVI) as a reactive sorbent due to its non-toxicity, abundance, economic feasibility, ease of production, and low maintenance. However, several limitations to using ZVI for groundwater remediation as a permeable reactive barrier (PRB) or as a water treatment filter have been observed in previous studies. These limitations include: (1) the development of an oxidized passivation layer (e.g., metal hydroxides and metal carbonates) at the materials surface that hinders continued effectiveness and reduces hydraulic conductivity; (2) the narrow effective pH range for select target contaminants; (3) the low selectivity for certain contaminants of interest in the presence of alternate electron acceptors (i.e., NO3−, O2, etc.); (4) the limited effectiveness for certain contaminants; and (5) the potential for sorbed contaminants to be mobilized as the ZVI material ages (Calderon and Fullana, 2015; Fu et al., 2014; Guan et al., 2015; Mackenzie et al., 1999; Noubactep, 2014). Due to these limitations, major concerns remain in the broad application of ZVI based water treatment technologies. A common pollutant found in groundwater is nitrate (NO3−) due to chemical fertilizers, pesticides, animal-feeding operations, petroleum products, industrial uses, and waste contamination through storm and urban runoff (Follett and Hatfield, 2001). Recently, ZVI has been studied for its ability to reduce NO3− in water and groundwater. Nitrate can be reduced to NH3 and N2 as Fe0 is oxidized to Fe2+ and then to Fe3+, with the subsequent formation of Fe2O3 and Fe3O4 depending on the reaction conditions (Fu et al., 2014). Under slightly alkaline to acidic conditions, the reduction of NO3− to NH3 is spontaneous and rapid, even in aerobic systems at room temperature (Cheng et al., 1997). In unbuffered solutions, the pH can increase to > 8.0 rapidly, resulting in surface passivation and a dramatic decrease in the rate of NO3− reduction (Xu et al., 2012). The reduction of NO3− will continue in the presence of mixed valent Fe corrosion products; however, the increasing pH may facilitate the gaseous release of NH3 from alkaline solutions to complete the remediation process. Rhenium (Re) is often used as a chemical analogue for Tc because of its similar chemical and thermodynamic properties (Brookins, 1986; Ding et al., 2013; Kim, 2003; Kim and Boulègue, 2003; Lenell and Arai, 2017; Li et al., 2019; Liu et al., 2013; Pierce et al., 2014; Poineau et al., 2006). Both elements are found in the same column of the periodic table and have valence states ranging from 0 to +7. The most stable oxidation state is +7, with the soluble perrhenate (ReO4−) species being comparable to 99TcO4−, and, consistent with Tc, the next most stable oxidation state for Re is +4 (Darab and Smith, 1996). Using Re as a non-radioactive analogue for Tc, Liu et al. (2013) and Lenell and Arai (2017) investigated Tc(VII) reduction and sorption on ZVI as a possible remediation technique for contaminated groundwater. Batch and column experiments demonstrated the effective reduction of ReO4− by ZVI in simulated groundwater solutions. However, the observed rate of ReO4− immobilization/sorption decreased significantly as the pH increased from 8 to 10. In addition, Re(VII) has been observed to be more resistant to chemical reduction than Tc(VII) (Maset et al., 2006). Also, Tc(VII) reduction continued during nitrate (NO3−) reduction in soil microcosm experiments while Re(VII) reduction was minimized (Wharton et al., 2000). Both Tc(VII) and Re(VII) are subject to reduction by sulfide (S2−); however, perrhenate is still more resistant

2. Materials and methods 2.1. Materials Four Fe-based reactive sorbents were evaluated in the current study: (1) Cleanit®, a high-surface area PIC material consisting mainly of Fe0 and Fe oxides produced by Höganäs Environment Solutions (Cary, NC); (2) a reagent grade ZVI supplied by Sigma-Aldrich; (3) Bayoxide® E33 (GFH1), a poorly crystallized β-FeOOH supplied by Advant Edge Technologies, Inc. (Buford, GA); and (4) GFH®, a granular ferric hydroxide (GFH2) manufactured by Evoqua Water Technologies (Pittsburg, PA). The novel PIC material has been proposed for use in waste water treatment applications that are similar to ZVI (Hu et al., 2010; Seaman et al., 2018), while the GFH1 and GFH2 materials are both marketed for the removal of As(III), As(V), and other heavy metals from contaminated water (Badruzzaman et al., 2004; Evoqua, 2014). All test solutions were made using reagent grade chemicals and milli-Q water. 2.2. Batch test methods The two batch test solutions consisted of (i) 4 mg L−1 of Re as NaReO4 in a 0.01 M NaCl background solution, and (ii) 4 mg L−1 of Re (NaReO4) plus 100 mg L−1 NO3− (NaNO3) in a 8.9 mM NaCl background solution. The rhenium concentration was chosen based on previous sorption isotherm experiment (not presented), reflecting the maximum sorbed concentration. Each treatment was carried out in triplicate, including no sorbent control beakers in both oxic (i.e., lab atmosphere) and anoxic (5% H2(g)/95% N2(g) atmosphere; Coy Vinyl Anaerobic Chambers, Grass Lake, MI) test environments, with a solid/ solution ratio of 50 (g L−1). The reaction vessels were placed on an orbital shaker and equilibrated at 140 rpm. Both the oxic and anoxic treatment remained open to the equilibrating treatment atmosphere. 380

Environment International 128 (2019) 379–389

F. Coutelot, et al.

Fig. 1. Batch pH and ORP values for oxic and anoxic conditions without NO3−.

laboratory column system was designed to mimic conventional flowthrough filter applications. Clear polyvinyl chloride (PVC) tubing with an inner diameter of 2.66 cm and a length of 8 cm was used for the filter column. Plastic mesh was placed at the inlet and outlet of the column to retain the reactive materials. The filter column was packed with 30 g of Ottawa sand at the bottom followed by 50 g of PIC material, then another 30 g of Ottawa sand on the top to hold the reactive material in place and disperse flow across the cross section of the reactive filter material. An additional filter column containing only 110 g of Ottawa sand was evaluated as an experimental control. A peristaltic pump was used to maintain a constant inlet flow rate of 1 mL min−1, the flow rate was chosen to simulate an accelerated permeable reactive barrier. A flow through pH electrode was placed at the column outlet to monitor the pH of the effluent, and a fraction collector was used to collect effluent samples for additional chemical analysis. Four column experiments were conducted: (1) a control test using Ottawa sand as a non-reactive (limited) sorbent (4 mg L−1 of Re as NaReO4 in 0.01 M NaCl); (2) a Re breakthrough test using the PIC

Non-destructive samples (a total of 2.75% of the solution, 0.5 mL per sampling intervals) were at intervals ranging from 1 to 720 h, while the pH and oxidation reduction potential (ORP) of the residual solution in the reaction vessel was measured. The samples were filtered (0.22 μm pore size filter), and acidified (2% HNO3) for chemical analysis of Re and Fe concentrations by inductively coupled plasma mass spectrometry (ICP-MS) on a NexIon 300× (Perkin Elmer, Inc.) in accordance with the quality assurance (QA) and quality control (QC) protocols of EPA method 6020A (USEPA, 2007). Non acidified samples were analyzed as follows: NO3− was determined by the Chromotropic Acid test method, nitrite (NO2−) was determined by the Diazotization method (APHA, 1997b), and ammonia/ammonium (NH4+/NH3) was determined by the Phenate method (APHA, 1997a). 2.3. Column methods Column testing focused on the PIC material, which proved to be the most effective Re sorbent based on the batch tests described above. The 381

Environment International 128 (2019) 379–389

F. Coutelot, et al.

periodically throughout testing. Effluent samples were filtered (0.2 μm pore-size filter) and acidified (2% HNO3) for preservation in preparation for analysis of Re, Fe, Na, calcium (Ca), aluminum (Al), magnesium (Mg), and potassium (K) using ICP-MS as described above. Unacidified fractions were analyzed for other reactive solution components (i.e., NO3−, NO2−, and NH4+/NH3) using the colorimetric methods cited above. 2.4. Solid phase analysis Both the initial sorbent materials and the residual materials after testing were characterized by x-ray diffraction (XRD) analysis using a Bruker D-2 Phaser operating at 30 kV and 10 mA using Cu K-alpha radiation (λ = 1.5406 A) over an angular range of 10° to 90° 2θ with a step scan size of ≈0:014° sec−1 using the Bragg-Brentano geometry. Phase identification was conducted using the software Match! 3 (Match!, 2011) in conjunction with the Crystallography Open Database (Gražulis et al., 2012). 2.5. Statistics Statistical tests were conducted on the entire data set using ANOVA to determine whether or not the means of several treatment groups are significantly different, and therefore generalizes the Student t-test to more than two groups. In some cases, the probability that the means of two populations were equal was evaluated using a Tukey post-hoc test to address multiple comparisons. All of the statistical analyses were performed using R 2.14.0 (R Development Core Team, 2013). 3. Results Fig. 2. Rhenium in solution as a function of contact time in anoxic (A) and oxic (B) test environments without NO3− present. Error bars represent the standard deviation of the mean.

3.1. Materials characterization XRD analysis was used to investigate the mineral composition of all four iron materials. The ZVI was determined to be composed of 100% of zero-valent iron, while the PIC material was composed of 90% zerovalent iron, with ≈10% composed of the mixed-valence Fe oxides with a formula of Fe3O4. The GFH1 was composed of goethite (α-FeOOH) and antigorite [(Mg, Fe)3(Si2O5)(OH)4] and the GFH2 is composed of Fe oxy-hydroxide (FeOOH).

Table 1 ANOVA test results comparing the mean Re concentrations of all sorbent treatments compared to the Re concentration in the control for experiments without NO3. Groups

Estimate

SE

df

t.ratio

p.value

Oxic GFH1 GFH2 PIC ZVI

0.14 1.26 2.93 2.53

0.26 0.26 0.26 0.26

293 293 293 293

0.552 4.782 11.304 9.843

0.999 0.0001 < 0.0001 < 0.0001

Anoxic GFH1 GFH2 PIC ZVI

0.21 1.23 2.96 2.61

0.26 0.26 0.26 0.26

293 293 293 293

0.809 4.648 11.422 10.178

0.9949 0.0002 < 0.0001 < 0.0001

3.2. Batch tests 3.2.1. Nitrate free batch tests Fig. 1 shows the pH and Oxidation-Reduction Potential (ORP) of the treatment solutions versus the equilibration time (in hours) for the control, PIC, ZVI, GFH1, and GFH2 treatments in the oxic test environment. In general, the addition of the Fe-based sorbent materials significantly increased the pH of the solution compared to the control except for GFH2 (Fig. 1A). As expected, the greatest pH increase was associated with the consumption of acidity associated with the oxidation of the two Fe0 based sorbents, PIC and ZVI. The ORP for the Fe-based sorbents showed a large degree of variation at the beginning of the experiment (Fig. 1C). Under oxic conditions, the ORP in control samples fluctuated between ≈61.4 and 340 mV, which can be attributed to the fact that the no-solids controls were poorly redox poised systems. Compared to the control, the ZVI and PIC materials lowered the ORP over time, while the GFH1 and GFH2 materials showed a similar ORP the control. Under anoxic conditions, PIC and ZVI materials displayed a greater pH than the control (Fig. 1B). In contrast, the GFH1 and GHF2 treatments didn't significantly impact the pH. As expected, the ORP values for the anoxic treatments (Fig. 1D) were consistently lower than the values observed for the oxic test environment, regardless of the sorbent. In general, after 48 h of contact with the solution, all treatment and control samples exhibited negative ORP values ranging from −279 to

sorbent material (4 mg L−1 of Re as NaReO4 in 0.01 M NaCl); (3) a Re breakthrough test using the PIC material with NO3− also present in the inlet solution (4 mg L−1 of Re as NaReO4 plus 100 mg L−1 NO3− (NaNO3) in a 0.0089 M NaCl); and (4) a Re (4 mg L−1 of Re as NaReO4) breakthrough test using an Artificial Groundwater (AGW) surrogate. An artificial groundwater surrogate based on routine sampling of non-impacted water table wells on the SRS was used as the leachant (Strom and Kaback, 1992). The AGW consisted of 1.88 mg L−1 Na, 1.0 mg L−1 Ca, 0.66 mg L−1 Mg, 0.21 mg L−1 K 0.73 mg L−1 SO4 and 5.51 mg L−1 Cl. No effort was taken to restrict the exposure to dissolved O2 as all of the inlet test solutions were open to the laboratory atmosphere. Each test column was initially saturated with 0.01 M NaCl(l) for 3 h and then switched to the appropriate treatment solution. Filter plugging, as influenced by solution treatment, was monitored using a piezometer tube at the column inlet, with head buildup monitored 382

Environment International 128 (2019) 379–389

F. Coutelot, et al.

Fig. 3. Batch pH and ORP values for the sorbent treatments in the oxic and anoxic environment in the presence of NO3−.

−570 mV, indicative of anoxic conditions. Fig. 2 shows the ratio of residual Re in solution over the initial Re concentration in batch experiments throughout equilibration under anoxic (2A) and oxic (2B) conditions without NO3−. The Re results for a given sorbent were remarkably similar under both treatment atmospheres, with no decrease in soluble Re observed for the control treatment throughout equilibration. All treatments under anoxic conditions showed Re sorption except for GHF1, with full Re sorption occurring within the first 200 and 400 h of equilibration for the PIC and the ZVI, respectively. In contrast, a 50% decrease in soluble Re was observed for the GHF1 material in the first 12 h for both oxic and anoxic test conditions. Statistical analysis showed that there was no significant difference in the results from the oxic and anoxic test environments (Table 1). There was a statistically significant difference between the control samples and the PIC material (p value < 0.001 in both oxic and anoxic), the ZVI (p value < 0.001 in both oxic and anoxic), and the

GHF2 (p value = 0.0001 in oxic condition and p value = 0.0002 in anoxic). However, the results for GFH1 were not statistically different from the control (p value = 0.999 in oxic and 0.9949 in anoxic condition). 3.2.2. Batch tests with nitrate present Fig. 3 shows the pH and ORP of solution treatments with NO3− versus the contact time (in hours) for the no sorbent control, PIC, ZVI, GFH1, and GFH2 treatments in the oxic test environment. The pH for the no sorbent control starts at ≈5.7 and fluctuate between ≈5.8 and 7.1 throughout the experiment (Fig. 3A). The pH for the PIC treatment is above the control pH throughout the experiment. The pH is ≈9.3 at the first hour sampling event then increases to ≈10.4 at the 24-hour sampling event (Fig. 3A). The pH then oscillates between ≈8.5 and 10.3 for the remainder of the experiment. The pH for the ZVI treatment stays the same as the control until the 12-hour sampling time, with a pH 383

Environment International 128 (2019) 379–389

F. Coutelot, et al.

Fig. 5. Nitrate reduction (sorption) as a function of contact time in anoxic (A) and oxic (B) test environments. Error bars represent the standard deviation of the mean.

Fig. 4. Rhenium in solution as a function of contact time in anoxic (A) and oxic (B) test environments for solutions containing 100 mg L−1 NO3−. Error bars represent the standard deviation of the mean.

GFH1 and GFH2 materials displayed ORP values that were similar to the control. Under anoxic conditions (Fig. 3B and D) there was less scatter in both the pH and ORP values for the anoxic treatment atmosphere, consistent with results from the NO3−-free batch test. The pH of the control ranges between ≈8.0 and 8.8. For the PIC and ZVI material increases pH when compared to the control. As expected, the ORP values for the anoxic treatments were consistently lower than the values observed for the oxic environment, regardless of the sorbent, with final ORP values that were similar for all treatments. Fig. 4 shows the ratio of residual Re in solution over the initial concentration versus the contact time (in hours) for the no sorbent control, PIC, ZVI, GFH1, and GFH2 treatments in the oxic (4A) and anoxic (4B) environments for solutions containing NO3−. In general, there was more variation in the Re sorption results for NO3− containing treatments. However, the trends with respect to Re removal were generally consistent with those demonstrated in Fig. 2, with PIC being the most effective at sorbing Re followed by ZVI under both test conditions. The PIC treatment was able to fully sorb Re within the first 200 h under both anoxic and oxic conditions. For ZVI, the Re concentration decreased at a slower rate than in the absence of NO3−, but Re was still no longer detectable after 360 h in both test environments. With GFH2 material, the Re concentration decreases to 50 to 60% of the starting concentration within the first 100 h and remains fairly constant until the end of the experiment, consistent with previous observations. For the GFH1 treatment, the Re concentration remained similar to the control throughout the entire experiment under both oxic and anoxic test conditions. The statistical results show a significant difference in the mean Re concentration between the control and the PIC materials (p value < 0.001) and the GHF2 (p value < 0.001) in both oxic and anoxic conditions (Table 2). However, the ZVI shows a significant difference in

Table 2 ANOVA test results comparing the mean Re concentrations of all sorbent treatments compared to the Re concentration in the control for experiments with NO3. Groups

Estimate

SE

df

t.ratio

p.value

Oxic GFH1 GFH2 PIC ZVI

0.39 1.50 3.36 1.76

0.30 0.30 0.30 0.30

278 278 278 278

1.302 4.992 11.1 5.829

0.9526 < 0.0001 < 0.0001 < 0.0001

Anoxic GFH1 GFH2 PIC ZVI

0.53 1.69 2.15 0.87

0.30 0.30 0.30 0.30

278 278 278 278

1.750 5.616 7.137 2.888

0.7656 < 0.0001 < 0.0001 0.1138

of ≈7.3. The GFH1 treatment pH is above the control pH throughout the experiment, with a pH of ≈7.9 at the first hour and increasing to ≈8.0 by hour 24. The pH then oscillates between ≈7.4 and 7.8 till the end of the experiment. For the GFH2 treatment, the pH is the same as the control pH throughout the experiment. The pH then increases to the end of the experiment with a pH of ≈10.8. For the PIC and ZVI, the pH increased overall throughout the experiment. As observed above for the NO3− free treatments, there is more scatter in the ORP (Fig. 3C). values for the oxic treatment atmosphere than in the anoxic treatments. Such scatter may be indicative of the lack of full O2 equilibration throughout the batch reactor, resulting in redox gradients that make consistent, repeated ORP measurements difficult. The no sorbent control ORP fluctuates between ≈274 mV and 366 mV. Compared to the control, the two Fe0 materials (i.e., ZVI and PIC) reduce the ORP ending at ≈8 mV (PIC) and 13.1 mV (ZVI). However, 384

Environment International 128 (2019) 379–389

F. Coutelot, et al.

Fig. 6. Linear regression comparing NO3− in solution to Re in solution to test for correlation in the presence and absence of O2.

the GFH1 treatment was not statistically different from the no-sorbent control sample (p value = 0.9526 in oxic and 0.7656 anoxic conditions).

Table 3 Tukey simultaneous test for differences of Re concentration means with and without NO3−. Difference of levels (without NO3− − with NO3−) Control GHF1 GFH2 PIC ZVI

Difference of means

0.526 0.245 0.174 0.712 1.781

95% confidence interval

(−0.066; 1.119) (−0.445; 0.936) (−0.519; 0.867) (0.031; 1.394) (1.105; 2.457)

Adjusted Pvalue

3.2.3. Nitrate reduction Fig. 5 shows the NO3− concentration (in mg L−1) versus time (in hours) for the control, PIC, ZVI, GFH1, and GFH2 treatments in the anoxic (5A) and oxic (5B) test environments. Under both test conditions, the NO3− concentration in the no-sorbent control treatments remained stable throughout the experiment. For the PIC and ZVI treatments, the NO3− concentration decreased to the point where it was no longer detectable after 200 and 400 h, respectively. The GFH2 material also reduced the NO3− concentration when compared to the control, reaching ≈50 to 60 mg L−1 after 168 h of equilibration for the anoxic and oxic test environments. The NO3− concentration for the GFH1 treatment remained quite similar to that of the no-sorbent control throughout the experiment. There was no significant correlation between NO3− and Re

0.132 0.981 0.999 0.032 0

Individual confidence level 95%.

oxic conditions (p value < 0.001), but a weak difference in anoxic conditions (p value = 0.1138), and GHF2 (p value = 0.0001 in oxic and p value = 0.0002 in anoxic conditions). Further, the Re sorption on the PIC materials was not influenced by the presence of NO3− in the solution. Similar to the non-NO3− batch test, the Re concentration in

Fig. 7. Ratio of Re concentration of the column effluent (C) to the Re concentration in column influent (C0) over the liquid to solid ratio of the column. 385

Environment International 128 (2019) 379–389

F. Coutelot, et al.

Fig. 8. Effluent Fe concentration for each test treatment.

Fig. 9. Nitrate and ammonia in column effluent for the Re + NO3− treatment.

concentration, meaning that the level of NO3− in solution doesn't influence Re sorption (Fig. 6) for the range of concentration tested. However, a Tukey multiple comparison test (Table 3) was run comparing the means of Re concentration for each treatment materials with and without NO3−. The results show that the presence of NO3− in solution strongly influences Re sorption on ZVI (p-adjusted = 0), and to a lesser degree Re sorption on PIC material (p-adjusted = 0.03).

Rhenium breakthrough in the AGW solution was similar to that of the Re-only leaching test, if even a bit more delayed due to the lower ionic strength of the AGW compared to the 0.01 M NaCl leaching solution, with Re in the effluent becoming detectable at ≈208 L Kg−1. The Re concentration slowly increased reaching ≈1.0 mg L−1 around 906 L Kg−1. This is not surprising since the AGW leaching solution did not include any alternate oxidants, like NO3−, that might compete with Re(VII) for the reductive sorption capacity of the PIC material. In contrast, Re breakthrough in the presence of NO3− was achieved around 114 L Kg−1, showing that the presence of NO3− can lessen Re immobilization in the column, even though such an effect was not readily evident in the batch results. Soluble Fe present in the column effluent for each leachate treatment is shown in Fig. 8. As expected, no soluble Fe was detected in the effluent for the control treatment using sand as the filter material. Only limited dissolved Fe was detected in the Re plus NO3− leachate treatment at the time when Re first becomes detectable in the effluent despite the clear evidence of Re immobilization and NO3− reduction associated with the responsible Fe reactions. The lack of significant Fe in the effluent is presumably caused by the presence of NO3−, which serves as an additional oxidant to ensure more complete Fe oxidation and precipitation as Fe oxy-hydroxides within the filter matrix. For the Re leachates without NO3−, generally higher effluent Fe concentrations were observed. However, the effluent Fe concentrations appear to oscillate in a somewhat predictable manner, with the oscillations apparently shortening over time due to the logarithmic nature of the x-axis in

3.2.4. Column results For comparison, column results are presented in terms of the liquid/ solid ratio (i.e., L/S), where L is the volume of effluent and S is the mass of reactive material in kilograms. The effluent Re concentration (i.e., C) as a function of the inlet concentration (i.e., C0 ≈ 4 mg L−1) for the various column treatments in presented in Fig. 7. All column tests were conducted at a constant inlet flow rate of 1 mL min−1. The effluent Re breakthrough data are presented on a logarithmic scale because of the wide range of L/S ratios for the different tests. Rhenium breakthrough for the control column (i.e., sand) was quite rapid, with full breakthrough occurring at ≈2.4 L Kg−1. In contrast, full Re breakthrough in the 0.01 M NaCl background solution was never observed for the PIC material in the absence of the competing oxidant (i.e., NO3−) besides dissolved O2, despite the extended leaching duration (1509 h). Initial limited Re breakthrough was detected at around 149 L Kg−1, and the Re concentration slowly increased with continued leaching for a final effluent concentration that reached approximately 1 mg L−1 after 624 L Kg−1. 386

Environment International 128 (2019) 379–389

F. Coutelot, et al.

Fig. 10. Major elements detected in the column effluent for the Re + AGW leachate treatment, the red line represents the inlet concentration. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Re + AGW leachate treatment are shown in Fig. 10. For the major common cations (i.e., Na, Mg, K and Ca) there were some initial fluctuations at the beginning of the test that may be attributed to solutes associated with the sand and PIC materials. However, the effluent concentrations for the major cations were generally consistent with the inlet solution composition.

Fig. 8. This pattern is likely an artifact associated with how the samples were collected, filtered and then acidified for chemical analysis, with higher levels of dissolved Fe observed for samples that were immediately filtered and acidified for subsequent analysis as the samples were collected. In contrast, the level of soluble Fe clearly decreased if the effluent samples oxidized and Fe(III) precipitated with storage before filtration and acidification for analysis. The NO3− and NH3/NH4+ concentrations in the column effluent for the Re plus NO3− leachate treatment are shown in Fig. 9. Little or no detectable nitrite (NO2−) was ever present in the column effluent. Limited NO3− was initially detected in the column effluent, but it increases to ≈0.21 mg L−1 at around 7.8 L Kg−1. The presence of detectable NH3/NH4+ in the effluent occurs just before detection of any NO3− and increases quickly to ≈10.1 mg L−1 at 1.25 L Kg−1, and then remains between 9 and 14 mg L−1 through 20.5 L Kg−1, with limited NO3− present in the effluent. Eventually greater NO3− breakthrough occurred and a decrease in effluent NH3 levels coincided with Re breakthrough, indicating that the available reductive capacity had been largely consumed. The effluent leaching patterns for the major elements present in the

4. Discussion A series of batch and column experiments were carried out using various commercially available Fe sorbents to determine which materials are best at immobilizing Re as an analogue for 99Tc. Experiments were performed in both oxic and anoxic environments to determine the impact of O2, and in the presence and absence of NO3−, on the immobilization of the target contaminant, in this case Re serving as an analogue for Tc. In the batch experiments, the PIC material immobilized Re the most in the presence or absence of O2, and in the presence of NO3−. Perrhenate sorption by ZVI is generally fastest at near neutral pH (pH = 7), with sorption inhibited under alkaline conditions (Lenell and 387

Environment International 128 (2019) 379–389

F. Coutelot, et al.

References

Arai, 2017). However, dissolved ferrous species can further react with OH to form ferrous hydroxide species. This can be further oxidized to form oxyhydroxides such as magnetite (Fe3O4), which may further adsorb or entrap ReO4− (Guan et al., 2015; Newsome et al., 2019). In the study by Livens et al. (2004), reduced technetium is harbored by mackinawite (FeS) and EXAFS data indicates the presence of TcOS bonds. During re-oxidation mackinawite forms goethite, yet technetium remains in the reduced state. While the reduction of Tc(VII) can result in several potential Tc(IV) oxy-hydroxide products (i.e., TcO2-c(s), TcO2∙1.6H2O(s), and Tc2∙2H2O(s)) (Cantrell et al., 2013; Cantrell and Williams, 2012), additional spectroscopic studies using both XAFS or XANES are warranted to evaluate better the Re-Fe systems as an analogue for Tc-Fe systems. Part of the difference in behavior between the two elements may be a reflection of their chemical properties, such as their binding energies. The binding energy of TcO4_ and ReO4_ is _388.7 and _371.1 eV, respectively and this difference may be decisive in surface mediated reduction. The PIC material was also effective at fully reducing NO3− before Re immobilization. No significant buildup of nitrite (NO2−) was observed, with mass balance estimates for NH3 supporting full reduction. Nitrate reduction by Fe0 can produce multiple nitrogen products depending on the exact chemical conditions; however, the production of NH3 in the presence of ZVI is considered to be more thermodynamically favorable than N2 (Zhang et al., 2017). These results are consistent with the redox potential for the reduction of NO3− (E0 = 0.88 V) and ReO4− (E0 = −0.55 V), with the more positive NO3− having the higher affinity for electrons. While thermodynamic data exist for several potential Tc oxy-hydroxide reduction products, i.e., TcO2-c(s), TcO2∙1.6H2O(s), and Tc2∙2H2O(s) (Cantrell et al., 2013; Cantrell and Williams, 2012), this is not the case for Re. In addition, a recent study by Li et al. (2019) using synchrotron-based techniques to evaluate solid phase speciation found very little evidence for Re(VII) reduction compared to Tc(VII) under similar batch conditions using both the PIC and ZVI, despite considerable sequestration of both Tc(VII) and Re(VII). As demonstrated in the current study, Li et al. (2019) found that the PIC material was more effective than the ZVI at immobilizing both Tc and Re. However, these results suggest that the reactions for Re(VII) immobilization in the Fe0 system may be quite different than that of Tc (VII), despite batch evidence that ongoing Fe oxidation is required for Re(VII) immobilization. The ZVI also effectively removed Re from solution at a somewhat slower rate when compared to the PIC, and was more sensitive to the presence of NO3−. This difference could be explained by the presence of secondary formed iron oxides as explained earlier. In contrast, the two Fe oxide materials had limited impact on soluble Re, with GFH1 displaying no significant Re sorption when compared to the no-sorbent control and the GFH2 sorbing ≈20 to 40% of the Re. As noted above, the PIC and ZVI materials were both effective at reducing NO3− to NH3, with no significant buildup of nitrite (NO2−). In addition to the batch studies, dynamic column experiments were conducted using the PIC material under kinetically limited conditions that are more analogous to applications such as permeable reactive barriers or waste-water treatment filters. High levels of Re immobilization were observed in the absence of a competing oxidized species other than dissolved O2. When NO3− was present in the leaching solution, Re breakthrough was greatly hastened, indicating that NO3− was actively competing for the PIC reductive capacity. This was not fully evident from the batch results. Nitrate reduction to NH3 without NO2− buildup was also confirmed in column tests.

Abdelouas, A., Grambow, B., Fattahi, M., Andres, Y., Leclerc-Cessac, E., 2005. Microbial reduction of 99Tc in organic matter-rich soils. Sci. Total Environ. 336, 255–268. Allred, B., 2012. Laboratory evaluation of porous iron composite for agricultural drainage water filter treatment. Trans. ASABE 55, 1683–1697. APHA, 1997a. Method-4500-NH3 nitrogen. In: Standard Methods for the Examination of Water and Wastewater, pp. 20005 Washington, DC. APHA, 1997b. Method-4500-NO2− (nitrogen). In: Standard Methods for the Examination of Water and Wastewater, (Washington, DC 20005). Artinger, R., Buckau, G., Zeh, P., Geraedts, K., Vancluysen, J., Maes, A., Kim, J.-I., 2003. Humic colloid mediated transport of tetravalent actinides and technetium. Radiochim. Acta 91, 743–750. Badruzzaman, M., Westerhoff, P., Knappe, D.R., 2004. Intraparticle diffusion and adsorption of arsenate onto granular ferric hydroxide (GFH). Water Res. 38, 4002–4012. Beasley, T., Lorz, H., 1986. A review of the biological and geochemical behaviour of technetium in the marine environment. In: Technetium in the Environment. Springer, pp. 197–216. Brookins, D.G., 1986. Rhenium as analog for fissiogenic technetium: Eh-pH diagram (25 °C, 1 bar) constraints. Appl. Geochem. 1, 513–517. Bruno, J., Ewing, R.C., 2006. Spent nuclear fuel. Elements 2, 343–349. Calderon, B., Fullana, A., 2015. Heavy metal release due to aging effect during zero valent iron nanoparticles remediation. Water Res. 83, 1–9. Cantrell, K.J., Williams, B.D., 2012. Equilibrium Solubility Model for Technetium Release From Saltstone Based on Single-pass Flow Eperiments. Pacific Northwest National Laboratory, Richland, Washington. Cantrell, K.J., Williams, B.D., 2013. Solubility control of technetium release from saltstone by TcO2·xH2O. J. Nucl. Mater. 437, 424–431. Cantrell, K.J., Kaplan, D.I., Wietsma, T.W., 1995. Zero-valent iron for the in situ remediation of selected metals in groundwater. J. Hazard. Mater. 42, 201–212. Cantrell, K.J., Carroll, K.C., Buck, E.C., Neiner, D., Geiszler, K.N., 2013. Single-pass flowthrough test elucidation of weathering behavior and evaluation of contaminant release models for Hanford tank residual radioactive waste. Appl. Geochem. 28, 119–127. Cheng, I.F., Muftikian, R., Fernando, Q., Korte, N., 1997. Reduction of nitrate to ammonia by zero-valent iron. Chemosphere 35, 2689–2695. Coughtrey, P.J., Jackson, D., Thorne, M., 1983. Radionuclide Distribution and Transport in Terrestrial and Aquatic Ecosystems. A Critical Review of Data. Volume 3. AA Balkema. Darab, J.G., Smith, P.A., 1996. Chemistry of technetium and rhenium species during lowlevel radioactive waste vitrification. Chem. Mater. 8, 1004–1021. Ding, Q., Qian, T., Yang, F., Liu, H., Wang, L., Zhao, D., Zhang, M., 2013. Kinetics of reductive immobilization of rhenium in soil and groundwater using zero valent iron nanoparticles. Environ. Eng. Sci. 30, 713–718. Evoqua, 2014. GFH Media Removes Arsenic in Arizona Water Supply. vol. 2018 Evoqua Water Technologies (Internet). Follett, R.F., Hatfield, J.L., 2001. Nitrogen in the environment: sources, problems, and management. Sci. World J. 1, 920–926. Fu, F., Dionysiou, D.D., Liu, H., 2014. The use of zero-valent iron for groundwater remediation and wastewater treatment: a review. J. Hazard. Mater. 267, 194–205. Gražulis, S., Daškevič, A., Merkys, A., Chateigner, D., Lutterotti, L., Quirós, M., Serebryanaya, N.R., Moeck, P., Downs, R.T., Le Bail, A., 2012. Crystallography Open Database (COD): An Open-Access Collection of Crystal Structures and Platform for World-Wide Collaboration. Nucleic Acids Res. 40 (no. D1), D420–D427. https://doi. org/10.1093/nar/gkr900. (January 1). Gu, B., Liang, L., Dickey, M., Yin, X., Dai, S., 1998. Reductive precipitation of uranium (VI) by zero-valent iron. Environ. Sci. Technol. 32, 3366–3373. Guan, X., Dong, H., Ma, J., Lo, I.M., Dou, X., 2011. Performance and mechanism of simultaneous removal of chromium and arsenate by Fe (II) from contaminated groundwater. Sep. Purif. Technol. 80, 179–185. Guan, X., Sun, Y., Qin, H., Li, J., Lo, I.M., He, D., Dong, H., 2015. The limitations of applying zero-valent iron technology in contaminants sequestration and the corresponding countermeasures: the development in zero-valent iron technology in the last two decades (1994–2014). Water Res. 75, 224–248. Hu, B. (2016). Permeable porous composite. Google Patents. Hu, Q.-H., Weng, J.-Q., Wang, J.-S., 2010. Sources of anthropogenic radionuclides in the environment: a review. J. Environ. Radioact. 101, 426–437. Hwang, Y.-H., Kim, D.-G., Shin, H.-S., 2011. Mechanism study of nitrate reduction by nano zero valent iron. J. Hazard. Mater. 185, 1513–1521. Icenhower, J., Qafoku, N., Zachara, J., Martin, W., 2010. The biogeochemistry of technetium: a review of the behavior of an artificial element in the natural environment. American Journal of Science (1818–1819) 310, 721–752. Jiang, Z., Lv, L., Zhang, W., Du, Q., Pan, B., Yang, L., Zhang, Q., 2011. Nitrate reduction using nanosized zero-valent iron supported by polystyrene resins: role of surface functional groups. Water Res. 45, 2191–2198. Kaplan, D., Parker, K., Orr, R., 1998. Effects of High-pH and High-ionic-strength Groundwater on Iodide, Pertechnetate, and Selenate Sorption to Hanford Sediments: Final Report for Subtask 3a. Pacific Northwest National Laboratory, Richland, WA. Kaplan, D., Mattigod, S., Parker, K., Iversen, G., 2000. Experimental Work in Support of the 129I-Disposal Special Analysis. Radioiodine Biogeochemistry, vol. 2329 Westinghouse Savannah River Company, Aiken, SC. Kim, E., 2003. Chemistry of rhenium as an analogue of technetium: experimental studies of the dissolution of rhenium oxides in aqueous solutions. Radiochim. Acta 91, 211–216.

Acknowledgments This research is supported through a Cooperative Agreement (DEFC09-07-SR22506) between the Department of Energy and the University of Georgia Research Foundation. 388

Environment International 128 (2019) 379–389

F. Coutelot, et al.

117, 104–107. Pierce, E.M., Lukens, W.W., Fitts, J.P., Jantzen, C.M., Tang, G., 2014. Experimental determination of the speciation, partitioning, and release of perrhenate as a chemical surrogate for pertechnetate from a sodalite-bearing multiphase ceramic waste form. Appl. Geochem. 42, 47–59. Poineau, F., Fattahi, M., Den Auwer, C., Hennig, C., Grambow, B., 2006. Speciation of technetium and rhenium complexes by in situ XAS-electrochemistry. Radiochim. Acta 94, 283–289. Qiu, X., Fang, Z., Yan, X., Gu, F., Jiang, F., 2012. Emergency remediation of simulated chromium(VI)-polluted river by nanoscale zero-valent iron: laboratory study and numerical simulation. Chem. Eng. J. 193, 358–365. R Development Core Team, 2013. R Development Core Team. RA Lang Environ Stat Comput. vol. 55. pp. 275–286. Ryu, A., Jeong, S.-W., Jang, A., Choi, H., 2011. Reduction of highly concentrated nitrate using nanoscale zero-valent iron: effects of aggregation and catalyst on reactivity. Appl. Catal. B Environ. 105, 128–135. Seaman, J.C., Li, D., Dorward, E., Cochran, J., Chang, H., Tandukar, M., Coutelot, F.M., Kaplan, D.I., 2018. Immobilization of radioactive materials using porous iron composite media. In: Waste Managment (WM2018), (Phoenix, AZ). Strom, R.N., Kaback, D.S., 1992. SRP Baseline Hydrogeologic Investigation: Aquifer Characterization Groundwater Geochemistry of the Savannah River Site and Vicinity (U). Westinghouse Savannah River Company, Environmental Sciences Section, Aiken, SC, pp. 98. Tagami, K., 2003. Technetium-99 behavior in the terrestrial environment. J. Nucl. Radiochem. Sci. 4, A1–A8. Truex, M.J., Lee, B.D., Johnson, C.D., Qafoku, N.P., Szecsody, J.E., Kyle, J.E., Tfaily, M.M., et al., 2017. Conceptual Model of Iodine Behavior in the Subsurface at the Hanford Site. PNNL-24709 Rev. 2. Pacific Northwest National Laboratory, Richland, WA. USEPA, 2007. Method 6020A, Rev. 0. Inductively Coupled Plasma-mass Spectrometry. Office of Solid Waste, Washington, DC. Warwick, P., Aldridge, S., Evans, N., Vines, S., 2007. The solubility of technetium (IV) at high pH. Radiochim. Acta 95, 709–716. Wharton, M., Atkins, B., Charnockab, J., Livens, F., Pattrick, R., Collison, D., 2000. An Xray absorption spectroscopy study of the coprecipitation of Tc and Re with mackinawite (FeS). Appl. Geochem. 15, 347–354. Xu, J., Hao, Z., Xie, C., Lv, X., Yang, Y., Xu, X., 2012. Promotion effect of Fe2+ and Fe3O4 on nitrate reduction using zero-valent iron. Desalination 284, 9–13. Zhang, X., Lin, S., Chen, Z., Megharaj, M., Naidu, R., 2011. Kaolinite-supported nanoscale zero-valent iron for removal of Pb2+ from aqueous solution: reactivity, characterization and mechanism. Water Res. 45, 3481–3488. Zhang, Y., Douglas, G.B., Pu, L., Zhao, Q., Tang, Y., Xu, W., Luo, B., Hong, W., Cui, L., Ye, Z., 2017. Zero-valent iron-facilitated reduction of nitrate: chemical kinetics and reaction pathways. Sci. Total Environ. 598, 1140–1150.

Kim, E., Boulègue, J., 2003. Chemistry of rhenium as an analogue of technetium: experimental studies of the dissolution of rhenium oxides in aqueous solutions. Radiochim. Acta 91, 211–216. Kutynakov, D.K.I., Parker, K.E., 1998. Radionuclide Distribution Coefficients for Sediments Collected From Borehole 299-E17-21: Final Report for Subtask 1a. Pacific Northwest National Laboratory, Richland, WA. Lawter, A.R., Garcia, Whitney L., Kukkadapu, Ravi K., Qafoku, Odeta, Bowden, Mark E., Saslow, Sarah A., Qafoku, Nikolla P., 2018. Technetium and iodine aqueous species immobilization and transformations in the presence of strong reductants and calciteforming solutions: remedial action implications. Sci. Total Environ. 636, 588–595. https://doi.org/10.1016/j.scitotenv.2018.04.240. Lenell, B.A., Arai, Y., 2017. Perrhenate sorption kinetics in zerovalent iron in high pH and nitrate media. J. Hazard. Mater. 321, 335–343. Li, D., Seaman, J.C., Kaplan, D.I., Heald, S.M., Sun, C., 2019. Pertechnetate (TcO4−) sequestration from groundwater by cost-effective organoclays and granular activated carbon under oxic environmental conditions. Chem. Eng. J. 360, 1–9. Lieser, K., 1993. Technetium in the nuclear fuel cycle, in medicine and in the environment. Radiochim. Acta 63, 5–8. Liu, H., Qian, T., Zhao, D., 2013. Reductive immobilization of perrhenate in soil and groundwater using starch-stabilized ZVI nanoparticles. Chin. Sci. Bull. 1–7. Livens, F.R., Jones, M.J., Hynes, A.J., Charnock, J.M., Mosselmans, J.F.W., Hennig, C., Steele, H., Collison, D., Vaughn, D.J., Pattrick, R.A.D., Reed, W.A., Moyes, L.N., 2004. X-ray absorption spectroscopy studies of reactions of technetium, uranium and neptunium with mackinawite. J. Environ. Radioact. 74, 211–219. https://doi.org/10. 1016/j.jenvrad.2004.01.012. Lv, X., Xu, J., Jiang, G., Tang, J., Xu, X., 2012. Highly active nanoscale zero-valent iron (nZVI)–Fe3O4 nanocomposites for the removal of chromium (VI) from aqueous solutions. J. Colloid Interface Sci. 369, 460–469. Mackenzie, P.D., Horney, D.P., Sivavec, T.M., 1999. Mineral precipitation and porosity losses in granular iron columns. J. Hazard. Mater. 68, 1–17. Maset, E.R., Sidhu, S.H., Fisher, A., Heydon, A., Worsfold, P.J., Cartwright, A.J., KeithRoach, M.J., 2006. Effect of organic co-contaminants on technetium and rhenium speciation and solubility under reducing conditions. Environ. Sci. Technol. 40, 5472–5477. Match!, 2011. Phase Identification From Powder Diffraction; Crystal Impact GbR. Meena, A.H., Arai, Y., 2017. Environmental geochemistry of technetium. Environ. Chem. Lett. 15, 241–263. Mitra, P., Sarkar, D., Chakrabarti, S., Dutta, B.K., 2011. Reduction of hexa-valent chromium with zero-valent iron: batch kinetic studies and rate model. Chem. Eng. J. 171, 54–60. Newsome, L., Morris, K., Cleary, A., Masters-Waage, N.K., Boothman, C., Joshi, N., Atherton, N., Lloyd, J.R., 2019. The impact of iron nanoparticles on technetiumcontaminated groundwater and sediment microbial communities. J. Hazard. Mater. 364, 134–142. Noubactep, C., 2014. Flaws in the design of Fe(0)-based filtration systems? Chemosphere

389