- Email: [email protected]
Microbial reduction of 99Tc in organic matter-rich soils
Science of the Total Environment 336 (2005) 255 – 268 www.elsevier.com/locate/scitotenv
Microbial reduction of
99
Tc in organic matter-rich soils
A. Abdelouas a,*, B. Grambow a, M. Fattahi a, Y. Andre`s b, E. Leclerc-Cessac c a
Ecole des Mines de Nantes, SUBATECH (UMR 6457), 4, rue Alfred Kastler-La chantrerie, BP 20722, 44307 Nantes Cedex 3, France b Ecole des Mines de Nantes, GEPEA (UMR 6144), 4, rue Alfred Kastler-La chantrerie, BP 20722, 44307 Nantes Cedex 3, France c Andra, Parc de la Croix Blanche-1/7, rue Jean-Monnet-92298 Chaˆtenay-Malabry Cedex, France Received 4 March 2004; received in revised form 28 May 2004; accepted 7 June 2004
Abstract For safety assessment purposes, it is necessary to study the mobility of long-lived radionuclides in the geosphere and the biosphere. Within this framework, we studied the behaviour of 99Tc in biologically active organic matter-rich soils. To simulate the redox conditions in soils, we stimulated the growth of aerobic and facultative denitrifying and anaerobic sulphate-reducing bacteria (SRB). In the presence of either a pure culture of denitrifiers (Pseudomonas aeruginosa) or a consortium of soil denitrifiers, the solubility of TcO4 was not affected. The nonsorption of TcO4 onto bacteria was confirmed in biosorption experiments with washed cells of P. aeruginosa regardless of the pH. At the end of denitrification with indigenous denitrifiers in soil/water batch experiments, the redox potential (EH) dropped and this was accompanied by an increase of Fe concentration in solution as a result of reduction of less soluble Fe(III) to Fe(II) from the soil particles. It is suggested that this is due to the growth of a consortium of anaerobic bacteria (e.g., Fe-reducing bacteria). The drop in EH was accompanied by a strong decrease in Tc concentration as a result of Tc(VII) reduction to Tc(IV). Thermodynamic calculations suggested the precipitation of TcO2. The stimulation of the growth of indigenous sulphate-reducing bacteria in soil/water systems led to even lower EH with final Tc concentration of 10 8 M. Experiments with glass columns filled with soil reproduced the results obtained with batch cultures. Sequential chemical extraction of precipitated Tc in soils showed that this radionuclide is strongly immobilised within soil particles under anaerobic conditions. More than 90% of Tc is released together with organic matter (60 – 66%) and Feoxyhydroxides (23 – 31%). The present work shows that ubiquitous indigenous anaerobic bacteria in soils play a major role in Tc immobilisation. In addition, organic matter plays a key role in the stability of the reduced Tc. D 2004 Elsevier B.V. All rights reserved. Keywords: Technetium; Sulphate-reducing bacteria; Iron-reducing bacteria; Pseudomonas aeruginosa; Sequential extraction; Organic matter
1. Introduction Large quantities of 99Tc are produced annually during the operation of nuclear power reactors. 99Tc * Corresponding author. Tel.: +33-2-51-85-84-62; fax: +33-251-85-84-52. E-mail address: [email protected] (A. Abdelouas). 0048-9697/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2004.06.005
is a pure h-emitter (Emax = 0.29 keV) with a half-life of 2.1 105 years. In France, fission products originating from reprocessed spent nuclear fuel are being vitrified and stored on the surface before possible final disposal in a deep geological clay formation (Vernaz, 2002). Despite a multibarrier concept for radioactive waste confinement, migration of radionuclide toward the biosphere cannot be excluded. Therefore, for safety assessment purposes, it is necessary to model
256
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
the mobility of long-lived radionuclides in the geosphere and the biosphere. In deep geological formations, where reducing conditions prevail, Tc is expected to exist as Tc(IV), while at the surface the highly soluble pertechnetate ion (TcO4-) is expected to dominate Tc speciation (Lieser and Bauscher, 1987). In the absence of complexing ligands, Tc(IV) may precipitate as highly immobile oxides and sulphides (Bondietti and Francis, 1979; Hunter and Kosawa, 1985; Wildung et al., 2000) or it may become adsorbed on different mineral surfaces. The French agency for radioactive waste management (Andra) is currently supporting research to study the behaviour of radionuclides in soils around an underground laboratory that is being constructed in the East of France. Within this framework, we investigated the behaviour of Tc in clayey organic matter-rich soils. Tagami and Uchida (1996, 1998) pointed out the role of microorganisms in Tc immobilisation in soils under waterlogged conditions. They related the Tc immobilisation to its reduction and sorption onto soil particles, particularly onto organic matter as a precipitate including TcO2 and TcS2. However, the authors did not give any details on microorganisms isolated from the soils. In addition to reduction and precipitation, biosorption of Tc onto bacteria cells can enhance its retention. Biosorption takes place on the cell wall, which is a rigid layer around the cell and the process has fast kinetics (Gadd, 1993). To simulate redox conditions that may develop in hydromorphic soils used in this work, we selected two types of bacteria, commonly present in soils: denitrifying and sulphate-reducing bacteria (SRB) (Postgate, 1984; Ehrlich, 1996, 1997). These two types can grow under redox conditions, ranging from aerobic to strongly anaerobic. Denitrifiers, which are chemoheterotrophic bacteria, are facultative and capable of growing under surface aerobic and anaerobic conditions using either oxygen or nitrate as electron acceptors. In addition, denitrifying bacteria are abundant in agricultural soils investigated in the present work due to the use of fertilizers. We studied Pseudomonas aeruginosa because it is ubiquitous in agricultural areas (Davet, 1996). Sulphate-reducing bacteria (SRB) grow under anaerobic conditions and are ubiquitous in soils (Postgate, 1984; Lovley et al., 1993; Abdelouas et al., 2000). Furthermore, some
species such as Desulfovibrio vulgaris can tolerate the presence of some oxygen in the system (Ganesh et al., 1997). Hence, the growth of these two types of bacteria is expected to lead to a wide range of redox conditions commonly found in natural soils. In addition, a diverse microbial population exists in soils including metal-reducing bacteria (e.g., Shewanella putrefaciens, Lovley, 1995) or other anaerobic bacteria (e.g., Clostridium sp., Francis et al., 2002) and many fungi and yeast can contribute to regulate the physicochemical parameters balance in the soils. Preliminary microbial work done by Andra confirmed the presence of denitrifying, iron- and sulphate-reducing bacteria (Leclerc-Cessac, 2004). For the purpose of simplification, only denitrifiers and sulphate-reducing bacteria were used; with the assumption that they could induce the optimal range of EH conditions including those of metal reducers (Abdelouas et al., 1998). The objective of this work was to determine whether 99Tc could be immobilised in surface soils via reduction and sorption processes catalysed by the growth of natural microbial populations. We used natural soils without any treatment to simulate better events under environmental conditions. Tc speciation was studied in the aqueous phase. However, Tc concentrations in the experiments were too low to study its speciation in the solid phase. As a result, spectroscopic methods (such as X-ray absorption spectroscopy) were not applied in our study for Tc speciation in soils.
2. Experimental 2.1. Materials and methods Three types of soil (S-3, S-5 and S-10) were sampled around the Andra underground laboratory in order to study safety of the disposal of high activity and long-lived radionuclides. The samples were stored at 4 jC until use. Soil samples for batch and column experiments were processed in a laminar flow hood, weighed in sterilised plastic serum bottles and deoxygenated in a glove box filled with nitrogen. Soils used for mineralogical and chemical analyses and certain batches were air dried and disaggregated before passing through a 2-mm mesh sieve in order
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
to prepare homogenous soil samples, according to procedure by Paul and Clark (1989). The water content of soil was measured by further drying at 105 jC overnight. The average water content of soil was about 40%. The total organic carbon was measured by heating the soil at 550 jC for 5 h. The total volatile inorganic matter was determined by heating the soil at 900 jC. The soils were analysed by X-ray fluorescence, gamma spectrometry and by scanning/transmission electron microscopy. In batch and column experiments water samples were filtered through 0.22-Am filters to remove microorganisms and soil particles prior to analysis. The water samples were analysed by a PQ-Excell VGElemental ICP-MS for cations with a detection limit of 0.001 – 1 Ag/l and by a Waters Quanta 4000 capillary electrophoresis for anions with a detection limit of about 0.5 mg/l. Technetium was measured by liquid scintillation with a Packard 2500/TR/AB analyser with a detection limit of 0.01 Bq/l. The pH was measured with an Orion 520A glass electrode and EH was measured with an Orion platinum electrode combined with an Ag/AgCl reference electrode. The EH electrode was checked against a BS 870 oxidation –reduction buffer solution purchased from Radiometer Analytical company. The EH is reported versus normal hydrogen electrode (NHE). Solid samples were studied with a Philips CM30 transmission electron microscope equipped with a Kevax EDX element analyser. The microscope was operated at 300 keV. A few drops of water-containing soil were deposited onto a carboncoated grid and rinsed with deionised water to remove salts resulting from solution evaporation. The samples were subsequently introduced into the microscope to prevent oxidation. Geochemical modelling of the experimental data for Tc speciation was conducted with the code MEDUSA (Puigdomenech, 2001). Tc speciation depends on physicochemical conditions, including pH and EH. MEDUSA is a Windows interface to the MS-DOS programs INPUT-SED-PREDOM, which perform the calculations needed to create chemical equilibrium diagrams. MEDUSA can display several diagrams simultaneously. MEDUSA can also call Hydrochemical Equilibrium Constants Database (HYDRA) to create diagrams based on equilibrium constants retrieved from an electronic database.
257
2.2. Procedure A mixture of experiments was conducted to determine the role of inorganic and organic soils components on the solubility of Tc. Experiments were conducted with/without amendment to stimulate certain types of microorganisms living in the soils. All experiments, except those with columns, were conducted in duplicate. The sequence of experiments was: (1) batch tests with soils and deionised water to which Tc(VII) was added in the form of TcO4NH4, in order to study Tc sorption on soil components. (2) growth of denitrifying and sulphate-reducing bacteria in deionised water, equilibrated and separated from soils, to which Tc(VII) was added in the form of TcO4NH4. The goal was to study the direct effect of bacteria without interference from soil particles. (3) sorption with Tc(VII) on bacteria cells, in order to study the role of biosorption in Tc immobilisation. (4) batch and column tests with soils and deionised water to which Tc(VII) was added in the form of TcO4NH4 and amended with carbon and phosphate for indigenous bacteria growth. The goal was to study the combined effect of bacteria and soil components. (5) sequential leaching of the soils to determine the distribution of the precipitated Tc(IV) within the soil components. The goal was to study the chemical stability of the precipitated Tc(IV). 2.2.1. Batch experiments The purpose of these experiments was to determine whether the soil components could affect the solubility of Tc(VII) via sorption processes. The tests were performed at 22 jC in 160 ml serum bottles with 10 g of dry and sieved soils and 100 ml of deionised water. TcO4 was added from a stock solution to reach a concentration of about 10 6 M. Following degassing and sealing, half of the bottles were kept in anaerobic conditions in the glove box, while the rest were kept on the bench in contact with the air. Water aliquots were taken with syringes within 3 months and filtered through 0.22-Am filters before Tc analysis.
258
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
2.2.2. Growth of bacteria in deionised water equilibrated with soils Ten grams of dry and sieved soils was mixed with 100 ml of deionised water for 48 h. Three water aliquots were taken within this period to confirm the equilibrium by measuring pH and elements such as Ca, Mg, Na, K etc. At equilibrium, the solution was centrifuged, filtered through a 0.45-Am filter and autoclaved at 121 jC for 20 min. The sterilised solution was then stored at 4 jC. For denitrification, a pure culture of P. aeruginosa (CIP A22), purchased from Institut Pasteur (Paris, France), was cultured. P. aeruginosa was grown overnight in a liquid nutriment broth (pH 7) at 30 jC. The cells were then separated from the medium by centrifugation and washed three times using a sterilised 9 g/l NaCl solution. For sulphate reduction, an indigenous SRB consortium isolated from soil S-5 was cultured. The SRB were isolated as follow: first, the indigenous SRB were cultured at 22 jC in a serum bottle containing soil S-5 and water amended with sulphate, phosphate and lactate. Then, after 3 weeks, an inoculum from the suspension was added to the Desulfovibrio medium no. 63 (catalogue DSM, Braunschweig, Germany) at pH 7.6 and incubated at 30 jC for 2 days. The procedure was repeated three times in a modified medium no. 63 with less iron, to prevent precipitation of SRB cells with iron sulphide. To stimulate the growth of P. aeruginosa and SRB in water equilibrated and separated from soil, the solution was amended with carbon and phosphate sources. Hence, for denitrification, after addition of a suspension of P. aeruginosa (CIP A22), the solution was amended with sodium nitrate (5 mM), sodium succinate (5 mM) and sodium trimetaphosphate (20 mg/l). The pH of the mixture ranged between 6.4 and 7, depending on the nature of the soil used to prepare the equilibrated water. For sulphate reduction in the equilibrated water, the solution was amended with an inoculum of SRB, sodium sulphate (10 mM), sodium lactate (10 mM), and sodium trimetaphosphate (20 mg/l). The pH of the mixture was about 7.6, regardless the nature of the soil used to prepare the equilibrated water. Succinate and lactate were added in excess, compared to the stoichiometric equations of denitrification with succinate and sulphate reduction with lactate, as indicated by Abdelouas et al. (1998,
1999a, 2000). Nitrate, sulphate and Tc reduction reactions are given as follows: 14NO3 +5C4H4O24 +3H+ = 7N2+20HCO3 +2H2O 3SO24 +2C3H5O3 = 6HCO3 +3HS +H+ 4TcO4 +C3H5O3 +18H+ = 4Tc4++3HCO3 +10H2O
Denitrification Sulphate reduction Tc reduction
2.2.3. Tc(VII) sorption on bacteria A suspension of P. aeruginosa in 9 g/l NaCl solution was used to study the role of biosorption in Tc immobilisation. A known concentration of bacteria, determined using plate counting techniques, was added to 20 ml NaCl solution, containing Tc(VII), and the pH was adjusted with HCl (pH 2.5) or NaOH (pH 6.9). Sorption experiments lasted for about a month. Prior to Tc analysis, the solutions were filtered through 0.22-Am filters to remove bacteria cells. Tc speciation was determined by tetraphenylarsonium chloride (TPA) extraction (Omori et al., 1994; Bensaid et al., 1998; Vichot et al., 2002). 2.2.4. Growth of indigenous bacteria in soils/ deionised water batches and columns In batch experiments, 10 g of unsterilised soils were added to 100 ml of deionised water and amendment was added to stimulate either indigenous denitrifying bacteria or SRB. The soil/water batch was amended as described in Section 2.2.2 for denitrifying and SRB. For denitrification, the initial pH of the suspension ranged between 6.4 and 7.0, depending on the nature of the soil. For sulphate reduction, the initial pH of the suspension ranged between 7.2 and 8.3, depending on the nature of the soil. Glass columns were filled with S-5 sieved soil, saturated with sterilised pure water and placed in the glove box filled with N2. For the growth of SRB the water was amended as described in Section 2.2.2. The water flowed into the column by gravity with a flow rate near 0.5 ml/day. 2.2.5. Distribution of Tc among soil components: sequential extraction To determine the distribution of Tc among soil components we conducted sequential extraction of soil samples containing precipitated Tc(IV) with a procedure similar to that of Tessier et al. (1979), but modified to minimize oxidation of Tc(IV) with oxy-
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
gen. Thus, the leaching solutions were degassed with a vacuum pump, purged with N2 and kept in a glove box filled with N2. About 2.5 g of wet soil samples was weighed in centrifuge tubes and leaching solutions were added according to the procedure of Tessier et al. (1979). A 1 M magnesium chloride solution (pH 7) was used to extract exchangeable Tc (Tc-Ex) at 22 jC. To extract Tc together with carbonate, a 1 M sodium acetate solution adjusted to pH 5 with acetic acid was added to the residue from Tc-Ex experiment at 22 jC (Tc-Car). The resulting residue was then leached with 0.04 M hydroxylamine hydrochloride in acetic acid solution at 96 jC to dissolve Fe – Mn oxyhydroxides (Tc-Ox). After reaction, the residue was leached with hydrogen peroxide (25%) adjusted to pH 2– 3 with HNO3, heated to 85 jC, then leached with 3.2 M ammonium acetate in 20% (v/v) HNO3 to oxidize the organic matter (Tc-OM). Finally, a microwave digestion of the remaining residue with HF – HNO3 was conducted to extract the Tc held within the mineral matrix (Tc-Min). The Fe concentration was also analysed in the leaching solutions.
3. Results 3.1. Soil characterisation Chemical analyses of soils with X-ray fluorescence are given in Table 1. Soil S-3 is high in Ca and low in Si and Fe compared to S-5 and S-10. The concentrations of the rest of the elements are similar. Analyses under the scanning electron microscope showed the presence of clay minerals, calcite and quartz, in agreement with the chemical composition. A photograph of clay aggregate from soil S-5 is given in Fig. 1. The clay minerals consisted of Si, Al, Fe and little of Mg and K. Table 1 Chemical composition of soils obtained with X-ray fluorescence
Al Ca Fe K Mg Si
S-3
S-5
S-10
42.8 167.0 18.3 14.1 0.2 158.7
65.2 84.5 42.2 18.8 14.1 206.8
56.4 3.0 57.4 8.9 1.9 308.0
Data are given in mg/g.
259
Fig. 1. Scanning electron microscope photograph of a clay aggregate of soil S-5.
No reducing mineral phases such as iron carbonate or sulphide were detected in the samples. Gamma spectrometry analysis showed that the radioactivity levels of soils are typical of soils with low levels of contamination, with 137Cs specific activity between 7 and 28 Bq/kg (Aslani et al., 2003). Results of heat treatments of the soils are given in Table 2. Soil S-5 is rich in organic matter (18 wt.%), while soil S-3 has lost more than 18% in weight after heat treatment at 900 jC, due to decomposition of calcite. Under air, calcite begins to decompose at 740 – 750 jC and decomposition is terminated at 780 jC (Engler et al., 1988). 3.2. Batch experiments Tc concentrations in deionised water reacted with the three soils at 22 jC are given in Table 3. Sorption kinetics of anaerobic tests is plotted in Fig. 2. Soil S-5 showed the fastest decrease of Tc concentration with a final value of 5.10 8 M (Fig. 2a). Unlike the experiments with soils S-5 and S-3 where the pH was near to 8, sample S-10 showed the slowest reduction of Tc concentration, which is due to the low initial pH of
Table 2 Results of heat treatment of dry soils
% Organic matter % Volatile mineral matter
Heat treatment (jC)
S-3
S-5
S-10
550 900
5.64 18.64
18.00 6.17
3.78 0.58
260
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
Table 3 Tc concentrations in deionised water/soils S-3, S-5 and S-10 batch experiments under anaerobic and aerobic conditions Anaerobic (after 26 days) S-3 Tc initial (M) Tc final (M)
10
Aerobic (after 26 days) S-5
6
5.10
10 8
S-10 6
6.10
10 8
S-3
6
2.10
7
Aerobic (after 74 days)
S-5
S-10
10
6
10
10
6
3.10
5.5 (Fig. 2b). Low pH values inhibit development of environmental microorganisms such as denitrifying and sulphate-reducing bacteria (Maier, 2000). The decrease of Tc concentration correlates well with a drop in EH (from + 400 to + 40 mV) during water/soil interaction (Fig. 2b). To test whether the drop in EH is chemically or biologically controlled, we conducted experiments with sterilised batches. The results are given in Fig. 2c, which clearly shows that in the sterilised batch the Tc concentration remained unchanged while it significantly decreased in unsterilised batches. This result suggests that the decrease in EH is controlled by the growth of microorganisms that metabolise the natural organic matter present in sample S-5 (18%). The correlation of the decrease of Tc concentration with EH drop suggests the reduction of Tc(VII) into Tc(IV) by indigenous microorganisms such as metal-reducing bacteria. The unchanged Tc concentration in the sterilised sample suggests that Tc(VII) is not sorbed onto soil mineral particles. In addition, in the experiment where the Tc(VII) was reduced, the Fe concentration increased from 0.06 to 14.7 ppm at the end of the run, due to reduction of highly insoluble Fe(III) from the soil to Fe(II) with higher solubility. The increase in Fe concentration indicates the growth and activity of iron-reducing bacteria (Lovley, 1995). In the sterilised batches, the Fe concentration remained constant. This is in good agreement with Andra’s microbial work in soils similar to those used in our study showing the presence of a consortium of iron-reducing bacteria (Leclerc-Cessac, 2004). Data from batch experiments under aerobic conditions are given in Table 3. It took 74 days for Tc(VII) to be partially reduced. Hence, except for soil S-5 with high organic matter content, the Tc concentration remained unchanged after 26 days of reaction. In addition, the Tc concentration increased back from
6
7
S-3
S-5
10
6
10
6
10
6
8.10
10 8
S-10 6
6.10
10 7
6
8.10
7
3 10 7 to 6 10 7 M after 74 days of reaction, probably as a result of its reoxidation (Table 3). 3.3. Growth of bacteria in deionised water equilibrated with soils 3.3.1. Denitrifying bacteria The results of batch experiments described in Section 3.2 confirmed the implication of microorganisms in the speciation and immobilisationn of Tc. Thus, we tested the interaction of Tc(VII) with growing microorganisms including denitrifying bacteria and SRB, which are ubiquitous in subsurface soils and groundwaters (Abdelouas et al., 2000). We used only equilibrated water, as indicated in Section 2.2.2 to prevent interference from soil particles. The amendment of equilibrated water with nitrate, succinate and phosphate stimulated the growth of P. aeruginosa added to the water. Traces of certain elements in the equilibrated water were enough to sustain cell development. Results of denitrification at 28 jC showed the complete reduction of nitrate to nitrite within 26 h (Fig. 3). The accumulated nitrite was then reduced to N2 after 45 h. During denitrification the pH increased to pH 8, while the EH remained constant at about + 410 mV. The Tc(VII) concentration remained unchanged and the total amount of Tc was extracted with T.P.A., indicating no change in its speciation. Final cell concentrations, determined using plate-counting techniques, were of the order of 108 cells/ml. No biosorption of Tc on bacteria cells was observed. 3.3.2. Sulphate-reducing bacteria A suspension of SRB enriched from soil S-5 as described in Section 2.2.2 was inoculated at 28 jC in equilibrated water amended with sulphate, lactate, phosphate and Tc(VII). The results of the experiments are plotted in Fig. 4. We notice a reduction of Tc(VII) in
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
261
Fig. 3. Denitrification at 28 jC in deionised water equilibrated and separated from soil S-5. The solution was amended with nitrate, succinate and phosphate.
remained unchanged (7 10 7 M) between 39 and 49 days of reaction despite the low EH. This excludes the precipitation of phases with low solubility such as TcO2 but rather supports the sorption of Tc(IV) on SRB cells. The pH increased from 7.6 to 8.2. Extraction with T.P.A. showed that a significant fraction of soluble Tc is present in the reduced form Tc(IV), which is probably complexed with carbonate and organics released through bacteria metabolism. To study the adherence of Tc on the cells, the filtrate containing bacteria with precipitated Tc(IV) was leached with 1 M H2O2 as a strong oxidant. It took 24 h of reaction for Tc to be completely dissolved, which indicates that the sorbed Tc(IV) is strongly complexed with bacteria. In the autoclaved amended water there was no change in Tc
Fig. 2. (a) Tc concentration in deionised water/soil batch experiments under anaerobic conditions at 22 jC. (b) pH and EH in deionised water/soil batch experiments under anaerobic conditions at 22 jC. (c) Tc concentration in unsterilised and autoclaved (auto) deionised water/soil S-5 batch experiments under anaerobic conditions.
the pristine sample accompanied by a drastic drop in the EH down to 80 mV due to sulphate reduction to sulphide. Analysis of solution showed a drop in sulphate concentration. However, the Tc concentration
Fig. 4. Tc concentration in unsterilised and autoclaved (auto) deionised water equilibrated with S-5 soil at 28 jC. The water was amended with sulphate, lactate and phosphate and inoculated with a SRB consortium isolated from soil S-5.
262
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
concentration, in agreement with the high EH of + 420 mV. In addition, the experimental results show that equilibrated water alone is not able to reduce Tc(VII). 3.4. Biosorption of Tc on P. Aeruginosa Biosorption experiments with Tc(VII) onto P. aeruginosa, harvested at the stationary phase and washed, were conducted in 9 g/l NaCl at pH 2.5 and 6.9. The initial bacteria concentration, determined using plate-counting techniques, was 4.1 109 cells/ml. The experiments with Tc(VII) with initial concentrations of about 10-6 M showed no Tc sorption (Fig. 5), in agreement with literature data for experiments at pH >7 (Henrot, 1989; Abdelouas et al., 2002). In addition, experiments with higher Tc(VII) concentration (f 5.10 4 M) did not show any significant decrease in Tc concentration. It has been concluded that, under aerobic conditions where TcO4- is stable, soil biomass cannot contribute to Tc(VII) retention via biosorption and bioaccumulation. These results are in agreement with those described in Section 3.3.1. 3.5. Growth of indigenous bacteria in soil/deionised water batches and columns 3.5.1. Denitrifying bacteria An example of denitrification experiment with indigenous bacteria in soil S-5 at 25 jC is plotted in Fig. 6a. The results are similar to those obtained with a pure culture of P. aeruginosa but it took longer for denitri-
Fig. 5. Tc concentration in the presence of a suspension of P. aeruginosa at pH 2.5 and 6.9. The bacteria concentration was 4.1 109 cells/ml.
Fig. 6. (a) Denitrification at 25 jC in deionised water/soil S-5 batch amended with nitrate, succinate and phosphate. (b) Decrease of Tc concentration following denitrification at 25 jC in deionised water/ soils batches amended with nitrate, succinate and phosphate. Denitrification occurred during the first 96 h.
fication to be completed (108 h). This is due to the low initial bacteria concentration in the soils compared to the batch experiments with a pure culture of P. aeruginosa (Section 3.3). Autoclaved samples did not show any nitrate reduction because of sterilisation. Fig. 6b shows the Tc concentrations in batch experiments with soils S-3, S-5 and S-10, where denitrification was completed within 96 h. The figure shows that the Tc concentration remained fairly constant during denitrification (within the first 96 h). After denitrification, the Tc concentration decreased to about 10 8 M as a result of the EH decrease of > + 400 mV (at 96 h) to + 40 mV (at 306 h) (Fig. 6b). Sterilised batches did not show any reduction of Tc(VII). Geochemical modelling with MEDUSA code using the experimental data, including EH, pH and Tc concentration, suggested the precipitation of TcO2 at the end of the experiment (Fig. 7). The figure shows
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
263
t = 25°C Fig. 7. EH – pH stability diagram of Tc, including experimental data before and after denitrification in deionised water/soil S-5 batch experiments (see Fig. 5b).
that at the beginning of the reaction the experimental data (circles) fit in the stability domain of TcO4 , while at the end of the experiment the data fit well in the stability domain of TcO2. The drop in EH in experiments described in Fig. 6b correlates well with an increase in Fe concentrations (>10 ppm) as a result of Fe(III) reduction from soil particles to Fe(II) probably by a consortium of anaerobic bacteria. The results of the soil column experiment are similar to those from batch experiments. The Tc concentrations, as a function of time in the outlet
Fig. 8. Tc concentration in the outlet water of soil S-5 column at 25 jC. The column was amended with 10 6 M TcO4 only.
water, are plotted in Fig. 8. After 36 days, nearly one pore volume flowed through the column with a flow rate of 0.5 ml/day. After 36 days, the Tc concentration stabilised at about 2.5 10-8 M, indicating the precipitation of hydrated TcO2. The drop in Tc concentration was accompanied by the decrease of EH from > + 400 mV at the beginning of the experiment to 3 mV at 36 d. Furthermore, the experimental data using columns fit well in the stability domain of TcO2 as predicted by MEDUSA calculations. 3.5.2. Sulphate-reducing bacteria Fig. 9a shows the Tc concentrations in water/soil batch experiments amended with sulphate, lactate and phosphate to stimulate the growth of SRB at 25 jC. The initial Tc(VII) concentration was about 10 6 M. We sampled once the soil colour turned into black at day 10, due to iron sulphide precipitation as indicated by Abdelouas et al. (2000). It took 10 days for the Tc concentration to be reduced by a factor of 40, with final concentrations of the order of 10-8 M. The black precipitate was identified, as the iron sulphide mackinawite (FeS), by X-ray microanalysis under the transmission electron microscope. Mackinawite formed as a result of sulphate (solution) reduction to sulphide and Fe(III) (soil) reduction to Fe(II). In
264
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
Fig. 10. Tc concentration in the outlet water of soil S-5 column at 25 jC. The column was amended with sulphate, phosphate, lactate and TcO4 (10 6 M).
TcO2 formation, sorption and coprecipitation with iron sulphide. Coprecipitation of Tc(IV) with iron sulphide has been indicated by Wharton et al. (2000). Fig. 9. (a) Decrease of Tc concentration during sulphate reduction at 25 jC in deionised water/soil batch experiments amended with sulphate, lactate and phosphate. Auto = autoclaved. (b) A sulphatereducing bacterium covered with iron sulphide and isolated from soil S-5.
addition, mackinawite sometimes contains some reduced Tc on bacteria cells (Fig. 9b). The Fe concentration increase in solution was limited (from 0.07 to 0.77 ppm) due to mackinawite precipitation. Final pH values were about 8.2 and the EH was as low as 50 mV. The sulphate and Tc(VII) in the autoclaved sample were not reduced. Given the high concentration of organic matter in the soils (3.8 – 18%) it is likely that anaerobes other than SRB are also involved in Tc(VII) reduction. The presence of soil in the batch experiments, unlike in those using SRB alone (Section 3.3), enhanced the precipitation of Tc(IV) with final concentrations near the solubility of TcO2 after only 10 days of reaction. MEDUSA calculations using the experimental data of EH, pH and Tc concentration suggest the formation of TcO2. The results of the column experiment are similar to those of the batch experiments and are plotted in Fig. 10. The Tc immobilisation is even higher than in experiment without sulphate. The final Tc concentration is as low as 6.10 9 M, probably due to
3.6. Distribution of Tc among soil components: sequential extraction The sequential leaching was applied to two types of experiments with soil S-5: an experiment with SRB (described in Section 3.5.2) and an experiment without amendments (described in Section 3.2), where Tc(VII) was reduced by natural growth of anaerobic microorganisms. The amounts of Tc held in the different fractions of the soils are given in Table 4 and show good agreement between the nominal amount of Tc precipitated in each experiment and
Table 4 Results of the sequential extraction of 99Tc (Bq) from soil S-5 with (S5_SRB + Tc) or without (S5 + Tc) the growth of sulphate-reducing bacteria S5_SRB + Tc S5 + Tc Tc nominal Tc-Ex Tc-Car Tc-Ox Tc-OM Tc-Min Total % Error
6313 146 51 1359 3950 441 5947 6
5976 161 151 1987 3825 229 6353 6
In both experiments, 10 g of soil were mixed with 100 ml of pure water and TcO4 .
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
265
11b). The rest of the iron is present as oxyhydroxides (17 – 21%) and in organic matter (8– 9%). No effect of the presence of SRB on Tc distribution was observed. This confirms that the Tc is indeed strongly complexed with organic matter.
4. Discussion
Fig. 11. (a) Results of the sequential extraction of 99Tc from soil S-5 under anaerobic (S5 + Tc) and SRB (S5_SRB + Tc) conditions. ‘‘S5_SRB + Tc’’ is soil S-5 amended with deionised water containing sulphate, lactate, phosphate and TcO4 . ‘‘S5 + Tc’’ is soil S-5 amended with deionised water containing TcO4 . (b) Results of the sequential extraction of total Fe from soil S-5 under anaerobic (S5 + Tc) and SRB (S5_SRB + Tc) conditions. ‘‘S5_SRB + Tc’’ is soil S5 amended with deionised water containing sulphate, lactate, phosphate and TcO4 . ‘‘S5 + Tc’’ is soil S5 amended with deionised water containing TcO4 .
the total amount of Tc obtained during leaching. The error did not exceed 6%. The results are plotted in Fig. 11a. In both types of experiment more than 90% of the Tc is released together with Fe-oxyhydroxides (23 – 31%) and organic matter (60 – 66%). Much smaller proportions (3– 5%) are present as an exchangeable form and in carbonate minerals. The remaining 3– 7% is present in the residual fractions. Analysis of Fe concentration in the different fractions of the soils using ICP/MS showed that most of the iron is present in the mineralised residual fraction (60 –74%) (Fig.
This work shows that anaerobic microorganisms such as metal- and sulphate-reducing bacteria play a major role in Tc immobilisation in organic matter-rich subsurface environment if oxygen access is limited. Final Tc concentrations were of the order of 10 8 M, which reflects the solubility of amorphous and hydrated TcO2 (Burnett and Jobe, 1997; Meyer et al., 1991). In the absence of living microorganisms, organic matter was not able to immobilise Tc(VII). In a nitrate-rich environment the reduction of Tc(VII) was not possible, mainly as the EH remained largely positive (> + 400 mV). No significant bioaccumulation and sorption of TcO4 occurred on cells because of electrostatic repulsion at pH 6.9. Also at pH 2.5 Tc(VII) did not sorb onto cells, which underlines the noncomplexing nature of pertechnetate. The lack of TcO4 sorption was true for pure cultures of P. aeruginosa and mixed cultures grown from the soils. This result is in good agreement with published data with pure and mixed cultures of aerobic and denitrifying bacteria (Henrot, 1989; Abdelouas et al., 2002; Andre`s et al., 2001). In addition, inorganic soil particles including silicate (quartz, clays) and calcite did not retain pertechnetate ion. Under environmental conditions, where organic matter essential for microorganisms growth is lacking, pertechnetate is expected to migrate with surface and groundwater. However, the presence of plants, which are able to accumulate and retain TcO4 via biotransformation processes, may play an important role in the environmental fate of this radionuclide (Aarkrog et al., 1997; Uchida et al., 2000; Hattink et al., 2003). The growth of denitrifying bacteria had no effect on Tc speciation and its concentration remained unchanged in solution. This suggests that in nitrate- and organic matter-rich environments such as agricultural areas, denitrifiers may compete with other bacteria, including anaerobic microorganisms, for organic matter and inhibit the reduction and precipitation of Tc.
266
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
Under these conditions, TcO4 may be available for plant uptake or may be transported in flowing surface and possibly groundwaters. In the presence of a consortium of anaerobic bacteria, Tc(VII) was reduced and precipitated as hydrated TcO2. These anaerobes appear to use the soil organic matter to grow and reduce Tc(VII). The general assumption stating that under surface oxidizing conditions TcO4- is stable and Tc is largely mobile (Lieser and Bauscher, 1987), does not hold in our soils under oxygen poor conditions because of the high content in organic matter. Localized degradation of organic matter by indigenous bacteria creates localized anaerobic conditions that help reduction and precipitation of Tc. Biological reduction of Tc(VII) by anaerobic bacteria in particular iron-reducing bacteria including Shewanella putrefaciens and Geobacter metallireducens was shown by several authors (Lloyd and Macaskie, 1996; Lloyd et al., 2000; Wildung et al., 2000). In addition, Lloyd et al. (2000) showed that a mixed microbial culture enriched from radionuclide-contaminated sediment, containing a single Geobacter sp., were able indirectly to reduce and precipitate Tc via biogenic magnetite. They also showed the direct enzymatic reduction of Tc(VII) by the iron reducers. More recently, field work conducted by Istok et al. (2004) evaluated the potential of indigenous microbes to reduce U(VI) and Tc(VII) in the presence of high concentration of nitrate in a shallow aquifer. They showed that in the presence of added carbon sources reduction of nitrate and Tc(VII) was observed. They explained the reduction of Tc(VII) by the in situ growth of metal-reducing microorganisms. In amended soil/water batches SRB were able to grow and Tc(VII) was reduced. Abiotic reduction by Fe(II) is also expected (Lloyd et al., 2000). Similar results on the reduction of Tc(VII) by pure and mixed cultures of SRB were published by several authors (Lloyd et al., 1998, 1999; Abdelouas et al., 2002). Geochemical modelling of the experimental data in batch experiments with/without sulphate suggested the reduction of Tc(VII) to Tc(IV) and formation of TcO2. However, under denitrification conditions, the Tc speciation is dominated by TcO4 and no reduction of Tc(VII) occurred. Modelling confirmed the critical role of EH and to a lesser extent pH in Tc immobilisation.
Sequential leaching of the soils containing precipitated Tc(IV) showed that Tc was released at the same time as Fe-oxyhydroxides and organic matter including bacteria. The behaviour of Tc was comparable under iron- and sulphate-reducing bacteria conditions. Two hypotheses can be suggested to explain the leaching results: (1) dissolution of free TcO2, regardless of the nature of soil components (organic matter, iron sulphide, iron oxyhydroxides, etc.). This hypothesis is not supported by the work of Keith-Roach et al. (2003) who investigated Tc-binding in estuarine sediments from locations that have received 99Tc discharge from Sellafield. They found that, in certain sediments, the reduced Tc was indeed released together with Feoxyhydroxides and organic matter, while in others about 80% of the total Tc was released at the same time as carbonate. (2) Tc(IV) is mostly complexed with Fe-oxyhydroxides and organic matter and is released once these phases are dissolved. In both cases we can consider that the reduced Tc is strongly held in the soils. Under sulphate-rich environmental conditions, the reducing conditions due to sulphide formation are expected to contribute to Tc(VII) reduction and to the stability of Tc(IV)-rich precipitated phases. Furthermore, oxidation of iron sulphide leads to the formation of FeOOH, which strongly holds TcO2 (Wharton et al., 2000). Abdelouas et al. (1999b) showed that the presence of mackinawite together with bioprecipitated UO2 in soil columns ensured the long-term chemical stability of UO2. The authors indicated that 90% of the oxygen, in groundwater entering the column, was consumed by mackinawite oxidation.
5. Conclusion The main conclusions drawn from this work are: (1) In organic matter-rich soils Tc(VII) was reduced and probably precipitated as TcO2 with a final concentration of about 10 8 M. The Tc reduction resulted from the drop in EH under indigenous metal- and sulphate-reducing bacteria conditions. (2) Tc(VII) was not reduced in the presence of denitrifying bacteria. In addition, TcO4 did not sorb onto P. aeruginosa. (3) Organic matter and to a lesser extent iron oxyhydroxides play a crucial role in Tc immobi-
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
lisation in soils. Tc(IV) is strongly complexed with organic matter and bacteria. (4) Under sulphate-reducing conditions iron sulphides do not play a major role in Tc immobilisation but are expected to contribute to Tc(IV) stability by maintaining reducing conditions. The present study suggests that anaerobic conditions may very likely develop in the organic matterrich soils surrounding the Andra site due to the activity of anaerobic microorganisms. This may efficiently contribute to the reduction and immobilisation of Tc, hence reducing its bioavailability. However, fieldwork is needed to assess better the in situ geochemical conditions (EH, pH) of the soils in order to predict the behaviour of Tc.
Acknowledgements We would like to thank Andra for funding this work and for providing the soil samples. In addition, special thanks to A. Barreau for SEM and E. Gautier for TEM work performed at the Institut des Mate´riaux de Nantes Jean Rouxel (France).
References Aarkrog A, Chen QJ, Dahlgaard H, Nielsen SP, Trapeznikov A, Pozolotina V. Evidence of 99Tc in Ural River sediments. J Environ Radioact 1997;37:201 – 13. Abdelouas A, Lu Y, Nuttall E, Lutze W. Reduction of U(VI) to U(IV) by indigenous bacteria in contaminated groundwater. J Contam Hydrol 1998;35:217 – 33. Abdelouas A, Lutze W, Nuttall E. Uranium contamination in the subsurface: characterization and remediation. In: Finch R, Burns PC, editors. Reviews in Mineralogy: Uranium: Mineralogy, Geochemistry and the Environment, vol. 38. Washington, DC: Mineralogical Society of America; 1999a. p. 433 – 73. Abdelouas A, Lutze W, Nuttall E. Oxidative dissolution of uraninite precipitated on Navajo sandstone. J Contam Hydrol 1999b;36: 353 – 75. Abdelouas A, Lutze W, Gong W, Nuttall E, Strietelmeier BA, Travis BJ. Biological reduction of uranium in groundwater and subsurface soil. Sci Total Environ 2000;250:21 – 35. Abdelouas A, Fattahi M, Grambow B, Vichot L, Gautier E. Precipitation of technetium by subsurface sulfate-reducing bacteria. Radiochim Acta 2002;90:773 – 7. Andre`s Y, Redercher S, Gerente C, Thouand G. Contribution of
267
biosorption to the behavior of radionuclides in the environment. J Radioanal Nucl Chem 2001;247:89 – 93. Aslani MAA, Aytas S, Akyil S, Yaprak G, Yener G, Eral M. Activity concentration of caesium-137 in agricultural soils. J Environ Radioact 2003;65:131 – 45. Ben-said K, Fattahi M, Musikas Cl, Abbe´ JCh. A novel approach to the oxidation potential of the Tc(VII)/Tc(IV) couple in hydrochloric acid medium through the reduction of TcO4 by Fe2 + ion. Radiochim Acta 1998;83:195 – 203. Bondietti EA, Francis CW. Geologic migration potentials of Tc-99 and Np-237. Science 1979;203:1337 – 40. Burnett KB, Jobe DJ. Solubility and dissolution of TcO2(cr) in the absence and presence of CO23 (aq). Atomic Energy of Canada Limited (AECL). Final report RC-1664, COG-96-238-I; 1997. Davet P. Vie microbienne du sol et production ve´ge´tale. Paris: INRA Editions, 1996. p. 21–272. Ehrlich HL. Geomicrobiology of sulfur. Geomicrobiology 3rd ed. New York: Marcel Dekker; 1996. p. 508 – 59. Ehrlich HL. Microbes and metals. Appl Microbiol Biotechnol 1997;48:687 – 92. Engler P, Santana MW, Mittleman ML, Balazs D. Non-isothermal in situ XRD analysis of dolomite decomposition. Rigaku J 1988;5:3 – 8. Francis AJ, Dodge CJ, Meiken GE. Biotransformation of pertechnetate by clostridia. Radiochim Acta 2002;90:791 – 7. Gadd GM. Microbial formation and transformation of organometallic and organometaloid compounds. FEMS Microbiol Rev 1993;11:297 – 316. Ganesh R, Robinson KG, Reed GD, Sayler GS. Reduction of hexavalent uranium from organic complexes by sulfate-reducing and iron-reducing bacteria. Appl Environ Microbiol 1997;63: 4385 – 91. Hattink J, Harms AV, Goeij JJM. Uptake, biotransformation, and elimination of 99Tc in duckweed. Sci Total Environ 2003; 312:59 – 65. Henrot J. Bioaccumulation and chemical modification of Tc by soil bacteria. Health Phys 1989;57:239 – 45. Hunter JA, Kosawa A. Manganese, technetium and rhenium. In: Bard JA, Parsons R, Jordan J, editors. Standard potentials in aqueous solutions. New York: Institute of Pure and Applied Chemistry; 1985. p. 212. Istok JD, Senko JM, Krumholz LR, Watson D, Bogle MA, Peacock A, et al. In situ bioremediation of technetium and uranium in a nitrate-contaminated aquifer. Environ Sci Technol 2004;38: 468 – 75. Keith-Roach MJ, Morris K, Dahlgaard H. An investigation into technetium binding in sediments. Mar Chem 2003;81:149 – 62. Leclerc-Cessac E. Etude expe´rimentale du roˆle des microorganismes sur la mobilite´ du Se et Tc: Revue finale interlaboratoires. Report Andra No DS/TR/04-012; 2004. Lieser KH, Bauscher C. Technetium in the hydrosphere and in the geosphere: I. Chemistry of technetium and iron in natural waters and influence of the redox potential on the sorption of technetium. Radiochim Acta 1987;42:205 – 13. Lloyd JR, Macaskie LE. A novel PhosphorImager-based technique for monitoring the microbial reduction of technetium. Appl Environ Microbiol 1996;62:578 – 82.
268
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
Lloyd JR, Nolting HF, Sole VA, Bosecker K, Macaskie LE. Technetium reduction and precipitation by sulphate-reducing bacteria. Geomicrobiol J 1998;15:43 – 56. Lloyd JR, Ridley J, Khizniak T, Lyalikova NN, Macaskie LE. Reduction of technetium by Desulfovibrio desulfuricans: biocatalyst characterization and use in a flowthrough bioreactor. Appl Environ Microbiol 1999;65:2691 – 6. Lloyd JR, Sole VA, Van Praagh VG, Lovley DR. Direct and Fe(II)mediated reduction of technetium by Fe(III)-reducing bacteria. Appl Environ Microbiol 2000;66:3743 – 9. Lovley DR. Microbial reduction of iron, manganese, and other metals. In: Donald L, Sparks L, editors. Advances in agronomy, vol. 54. New York: Academic Press; 1995. p. 175 – 231. Lovley DR, Roden EE, Phillips EJP, Woodward JC. Enzymatic iron and uranium reduction by sulfate-reducing bacteria. Mar Geol 1993;113:41 – 53. Maier RM. In: Maier RM, Pepper IL, Gerba CP, editors. Environmental microbiology. San Diego, USA: Academic Press; 2000. p. 363 – 400. Meyer RE, Arnold WD, Case FI, O’Kelley GD. Solubilities of Tc(IV) oxides. Radiochim Acta 1991;55:11 – 8. Omori T, Muraoka Y, Suganuma H. Solvent extraction mechanism of pertechnetate with tetraphenylarsonium chloride. J Radioanal Nucl Chem 1994;178:237. Paul EA, Clark FE. Soil microbiology and biochemistry. Academic press, INC, Harcourt Brace Jovanovich, Publishers; 1989. Postgate JR. The sulfate-reducing bacteria. 2nd edn. Cambridge, UK: Cambridge Univ. Press; 1984 (208 pp.). Puigdomenech I. HYDRA MEDUSA; make equilibrium diagrams
using sophisticated algorithms. Stockholm, Sweden: Royal Institute of Technology; 2001. Tagami K, Uchida S. Microbial role in immobilization of technetium in soil under waterlogged conditions. Chemosphere 1996;33:217 – 25. Tagami K, Uchida S. Chemical transformation of technetium in soil during the change of soil water conditions. Chemosphere 1998;5:963 – 71. Tessier A, Campbell PGC, Bisson M. Sequential extraction procedure for the speciation of particulate trace metals. Anal Chem 1979;51:844 – 51. Uchida S, Tagami K, Ru¨hm W, Steiner M, Wirth E. Separation of Tc-99 in soil and plant samples collected around the Chernobyl reactor using a Tc-selective chromatographic resin and determination of the nuclide by ICP-MS. Appl Radiat Isotopes 2000;53:69 – 73. Vernaz EY. Estimating the lifetime of R7T7 glass in various media. C R - Acad Sci., Phys 2002;3:813 – 25. Vichot L, Ouvrard G, Montavan G, Fattahi M, Musikas C, Grambow B. XAS study of technetium(IV) polymer formation in mixed sulphate/chloride media. Radiochim Acta 2002;90:575 – 9. Wharton MJ, Atkins B, Charnock JM, Livens FR, Pattrick RAD, Collison D. An X-ray absorption spectroscopy study of the coprecipitation of Tc and Re with mackinawite (FeS). Appl Geochem 2000;15:347 – 54. Wildung RE, Gorby YA, Krupka KM, Hess NJ, Li SW, Plymale AE, et al. Effect of electron donor and solution chemistry and products of dissimilatory reduction of technetium by Shewanella putrefaciens. Appl Environ Microbiol 2000;66:2451 – 60.
Microbial reduction of
99
Tc in organic matter-rich soils
A. Abdelouas a,*, B. Grambow a, M. Fattahi a, Y. Andre`s b, E. Leclerc-Cessac c a
Ecole des Mines de Nantes, SUBATECH (UMR 6457), 4, rue Alfred Kastler-La chantrerie, BP 20722, 44307 Nantes Cedex 3, France b Ecole des Mines de Nantes, GEPEA (UMR 6144), 4, rue Alfred Kastler-La chantrerie, BP 20722, 44307 Nantes Cedex 3, France c Andra, Parc de la Croix Blanche-1/7, rue Jean-Monnet-92298 Chaˆtenay-Malabry Cedex, France Received 4 March 2004; received in revised form 28 May 2004; accepted 7 June 2004
Abstract For safety assessment purposes, it is necessary to study the mobility of long-lived radionuclides in the geosphere and the biosphere. Within this framework, we studied the behaviour of 99Tc in biologically active organic matter-rich soils. To simulate the redox conditions in soils, we stimulated the growth of aerobic and facultative denitrifying and anaerobic sulphate-reducing bacteria (SRB). In the presence of either a pure culture of denitrifiers (Pseudomonas aeruginosa) or a consortium of soil denitrifiers, the solubility of TcO4 was not affected. The nonsorption of TcO4 onto bacteria was confirmed in biosorption experiments with washed cells of P. aeruginosa regardless of the pH. At the end of denitrification with indigenous denitrifiers in soil/water batch experiments, the redox potential (EH) dropped and this was accompanied by an increase of Fe concentration in solution as a result of reduction of less soluble Fe(III) to Fe(II) from the soil particles. It is suggested that this is due to the growth of a consortium of anaerobic bacteria (e.g., Fe-reducing bacteria). The drop in EH was accompanied by a strong decrease in Tc concentration as a result of Tc(VII) reduction to Tc(IV). Thermodynamic calculations suggested the precipitation of TcO2. The stimulation of the growth of indigenous sulphate-reducing bacteria in soil/water systems led to even lower EH with final Tc concentration of 10 8 M. Experiments with glass columns filled with soil reproduced the results obtained with batch cultures. Sequential chemical extraction of precipitated Tc in soils showed that this radionuclide is strongly immobilised within soil particles under anaerobic conditions. More than 90% of Tc is released together with organic matter (60 – 66%) and Feoxyhydroxides (23 – 31%). The present work shows that ubiquitous indigenous anaerobic bacteria in soils play a major role in Tc immobilisation. In addition, organic matter plays a key role in the stability of the reduced Tc. D 2004 Elsevier B.V. All rights reserved. Keywords: Technetium; Sulphate-reducing bacteria; Iron-reducing bacteria; Pseudomonas aeruginosa; Sequential extraction; Organic matter
1. Introduction Large quantities of 99Tc are produced annually during the operation of nuclear power reactors. 99Tc * Corresponding author. Tel.: +33-2-51-85-84-62; fax: +33-251-85-84-52. E-mail address: [email protected] (A. Abdelouas). 0048-9697/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2004.06.005
is a pure h-emitter (Emax = 0.29 keV) with a half-life of 2.1 105 years. In France, fission products originating from reprocessed spent nuclear fuel are being vitrified and stored on the surface before possible final disposal in a deep geological clay formation (Vernaz, 2002). Despite a multibarrier concept for radioactive waste confinement, migration of radionuclide toward the biosphere cannot be excluded. Therefore, for safety assessment purposes, it is necessary to model
256
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
the mobility of long-lived radionuclides in the geosphere and the biosphere. In deep geological formations, where reducing conditions prevail, Tc is expected to exist as Tc(IV), while at the surface the highly soluble pertechnetate ion (TcO4-) is expected to dominate Tc speciation (Lieser and Bauscher, 1987). In the absence of complexing ligands, Tc(IV) may precipitate as highly immobile oxides and sulphides (Bondietti and Francis, 1979; Hunter and Kosawa, 1985; Wildung et al., 2000) or it may become adsorbed on different mineral surfaces. The French agency for radioactive waste management (Andra) is currently supporting research to study the behaviour of radionuclides in soils around an underground laboratory that is being constructed in the East of France. Within this framework, we investigated the behaviour of Tc in clayey organic matter-rich soils. Tagami and Uchida (1996, 1998) pointed out the role of microorganisms in Tc immobilisation in soils under waterlogged conditions. They related the Tc immobilisation to its reduction and sorption onto soil particles, particularly onto organic matter as a precipitate including TcO2 and TcS2. However, the authors did not give any details on microorganisms isolated from the soils. In addition to reduction and precipitation, biosorption of Tc onto bacteria cells can enhance its retention. Biosorption takes place on the cell wall, which is a rigid layer around the cell and the process has fast kinetics (Gadd, 1993). To simulate redox conditions that may develop in hydromorphic soils used in this work, we selected two types of bacteria, commonly present in soils: denitrifying and sulphate-reducing bacteria (SRB) (Postgate, 1984; Ehrlich, 1996, 1997). These two types can grow under redox conditions, ranging from aerobic to strongly anaerobic. Denitrifiers, which are chemoheterotrophic bacteria, are facultative and capable of growing under surface aerobic and anaerobic conditions using either oxygen or nitrate as electron acceptors. In addition, denitrifying bacteria are abundant in agricultural soils investigated in the present work due to the use of fertilizers. We studied Pseudomonas aeruginosa because it is ubiquitous in agricultural areas (Davet, 1996). Sulphate-reducing bacteria (SRB) grow under anaerobic conditions and are ubiquitous in soils (Postgate, 1984; Lovley et al., 1993; Abdelouas et al., 2000). Furthermore, some
species such as Desulfovibrio vulgaris can tolerate the presence of some oxygen in the system (Ganesh et al., 1997). Hence, the growth of these two types of bacteria is expected to lead to a wide range of redox conditions commonly found in natural soils. In addition, a diverse microbial population exists in soils including metal-reducing bacteria (e.g., Shewanella putrefaciens, Lovley, 1995) or other anaerobic bacteria (e.g., Clostridium sp., Francis et al., 2002) and many fungi and yeast can contribute to regulate the physicochemical parameters balance in the soils. Preliminary microbial work done by Andra confirmed the presence of denitrifying, iron- and sulphate-reducing bacteria (Leclerc-Cessac, 2004). For the purpose of simplification, only denitrifiers and sulphate-reducing bacteria were used; with the assumption that they could induce the optimal range of EH conditions including those of metal reducers (Abdelouas et al., 1998). The objective of this work was to determine whether 99Tc could be immobilised in surface soils via reduction and sorption processes catalysed by the growth of natural microbial populations. We used natural soils without any treatment to simulate better events under environmental conditions. Tc speciation was studied in the aqueous phase. However, Tc concentrations in the experiments were too low to study its speciation in the solid phase. As a result, spectroscopic methods (such as X-ray absorption spectroscopy) were not applied in our study for Tc speciation in soils.
2. Experimental 2.1. Materials and methods Three types of soil (S-3, S-5 and S-10) were sampled around the Andra underground laboratory in order to study safety of the disposal of high activity and long-lived radionuclides. The samples were stored at 4 jC until use. Soil samples for batch and column experiments were processed in a laminar flow hood, weighed in sterilised plastic serum bottles and deoxygenated in a glove box filled with nitrogen. Soils used for mineralogical and chemical analyses and certain batches were air dried and disaggregated before passing through a 2-mm mesh sieve in order
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
to prepare homogenous soil samples, according to procedure by Paul and Clark (1989). The water content of soil was measured by further drying at 105 jC overnight. The average water content of soil was about 40%. The total organic carbon was measured by heating the soil at 550 jC for 5 h. The total volatile inorganic matter was determined by heating the soil at 900 jC. The soils were analysed by X-ray fluorescence, gamma spectrometry and by scanning/transmission electron microscopy. In batch and column experiments water samples were filtered through 0.22-Am filters to remove microorganisms and soil particles prior to analysis. The water samples were analysed by a PQ-Excell VGElemental ICP-MS for cations with a detection limit of 0.001 – 1 Ag/l and by a Waters Quanta 4000 capillary electrophoresis for anions with a detection limit of about 0.5 mg/l. Technetium was measured by liquid scintillation with a Packard 2500/TR/AB analyser with a detection limit of 0.01 Bq/l. The pH was measured with an Orion 520A glass electrode and EH was measured with an Orion platinum electrode combined with an Ag/AgCl reference electrode. The EH electrode was checked against a BS 870 oxidation –reduction buffer solution purchased from Radiometer Analytical company. The EH is reported versus normal hydrogen electrode (NHE). Solid samples were studied with a Philips CM30 transmission electron microscope equipped with a Kevax EDX element analyser. The microscope was operated at 300 keV. A few drops of water-containing soil were deposited onto a carboncoated grid and rinsed with deionised water to remove salts resulting from solution evaporation. The samples were subsequently introduced into the microscope to prevent oxidation. Geochemical modelling of the experimental data for Tc speciation was conducted with the code MEDUSA (Puigdomenech, 2001). Tc speciation depends on physicochemical conditions, including pH and EH. MEDUSA is a Windows interface to the MS-DOS programs INPUT-SED-PREDOM, which perform the calculations needed to create chemical equilibrium diagrams. MEDUSA can display several diagrams simultaneously. MEDUSA can also call Hydrochemical Equilibrium Constants Database (HYDRA) to create diagrams based on equilibrium constants retrieved from an electronic database.
257
2.2. Procedure A mixture of experiments was conducted to determine the role of inorganic and organic soils components on the solubility of Tc. Experiments were conducted with/without amendment to stimulate certain types of microorganisms living in the soils. All experiments, except those with columns, were conducted in duplicate. The sequence of experiments was: (1) batch tests with soils and deionised water to which Tc(VII) was added in the form of TcO4NH4, in order to study Tc sorption on soil components. (2) growth of denitrifying and sulphate-reducing bacteria in deionised water, equilibrated and separated from soils, to which Tc(VII) was added in the form of TcO4NH4. The goal was to study the direct effect of bacteria without interference from soil particles. (3) sorption with Tc(VII) on bacteria cells, in order to study the role of biosorption in Tc immobilisation. (4) batch and column tests with soils and deionised water to which Tc(VII) was added in the form of TcO4NH4 and amended with carbon and phosphate for indigenous bacteria growth. The goal was to study the combined effect of bacteria and soil components. (5) sequential leaching of the soils to determine the distribution of the precipitated Tc(IV) within the soil components. The goal was to study the chemical stability of the precipitated Tc(IV). 2.2.1. Batch experiments The purpose of these experiments was to determine whether the soil components could affect the solubility of Tc(VII) via sorption processes. The tests were performed at 22 jC in 160 ml serum bottles with 10 g of dry and sieved soils and 100 ml of deionised water. TcO4 was added from a stock solution to reach a concentration of about 10 6 M. Following degassing and sealing, half of the bottles were kept in anaerobic conditions in the glove box, while the rest were kept on the bench in contact with the air. Water aliquots were taken with syringes within 3 months and filtered through 0.22-Am filters before Tc analysis.
258
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
2.2.2. Growth of bacteria in deionised water equilibrated with soils Ten grams of dry and sieved soils was mixed with 100 ml of deionised water for 48 h. Three water aliquots were taken within this period to confirm the equilibrium by measuring pH and elements such as Ca, Mg, Na, K etc. At equilibrium, the solution was centrifuged, filtered through a 0.45-Am filter and autoclaved at 121 jC for 20 min. The sterilised solution was then stored at 4 jC. For denitrification, a pure culture of P. aeruginosa (CIP A22), purchased from Institut Pasteur (Paris, France), was cultured. P. aeruginosa was grown overnight in a liquid nutriment broth (pH 7) at 30 jC. The cells were then separated from the medium by centrifugation and washed three times using a sterilised 9 g/l NaCl solution. For sulphate reduction, an indigenous SRB consortium isolated from soil S-5 was cultured. The SRB were isolated as follow: first, the indigenous SRB were cultured at 22 jC in a serum bottle containing soil S-5 and water amended with sulphate, phosphate and lactate. Then, after 3 weeks, an inoculum from the suspension was added to the Desulfovibrio medium no. 63 (catalogue DSM, Braunschweig, Germany) at pH 7.6 and incubated at 30 jC for 2 days. The procedure was repeated three times in a modified medium no. 63 with less iron, to prevent precipitation of SRB cells with iron sulphide. To stimulate the growth of P. aeruginosa and SRB in water equilibrated and separated from soil, the solution was amended with carbon and phosphate sources. Hence, for denitrification, after addition of a suspension of P. aeruginosa (CIP A22), the solution was amended with sodium nitrate (5 mM), sodium succinate (5 mM) and sodium trimetaphosphate (20 mg/l). The pH of the mixture ranged between 6.4 and 7, depending on the nature of the soil used to prepare the equilibrated water. For sulphate reduction in the equilibrated water, the solution was amended with an inoculum of SRB, sodium sulphate (10 mM), sodium lactate (10 mM), and sodium trimetaphosphate (20 mg/l). The pH of the mixture was about 7.6, regardless the nature of the soil used to prepare the equilibrated water. Succinate and lactate were added in excess, compared to the stoichiometric equations of denitrification with succinate and sulphate reduction with lactate, as indicated by Abdelouas et al. (1998,
1999a, 2000). Nitrate, sulphate and Tc reduction reactions are given as follows: 14NO3 +5C4H4O24 +3H+ = 7N2+20HCO3 +2H2O 3SO24 +2C3H5O3 = 6HCO3 +3HS +H+ 4TcO4 +C3H5O3 +18H+ = 4Tc4++3HCO3 +10H2O
Denitrification Sulphate reduction Tc reduction
2.2.3. Tc(VII) sorption on bacteria A suspension of P. aeruginosa in 9 g/l NaCl solution was used to study the role of biosorption in Tc immobilisation. A known concentration of bacteria, determined using plate counting techniques, was added to 20 ml NaCl solution, containing Tc(VII), and the pH was adjusted with HCl (pH 2.5) or NaOH (pH 6.9). Sorption experiments lasted for about a month. Prior to Tc analysis, the solutions were filtered through 0.22-Am filters to remove bacteria cells. Tc speciation was determined by tetraphenylarsonium chloride (TPA) extraction (Omori et al., 1994; Bensaid et al., 1998; Vichot et al., 2002). 2.2.4. Growth of indigenous bacteria in soils/ deionised water batches and columns In batch experiments, 10 g of unsterilised soils were added to 100 ml of deionised water and amendment was added to stimulate either indigenous denitrifying bacteria or SRB. The soil/water batch was amended as described in Section 2.2.2 for denitrifying and SRB. For denitrification, the initial pH of the suspension ranged between 6.4 and 7.0, depending on the nature of the soil. For sulphate reduction, the initial pH of the suspension ranged between 7.2 and 8.3, depending on the nature of the soil. Glass columns were filled with S-5 sieved soil, saturated with sterilised pure water and placed in the glove box filled with N2. For the growth of SRB the water was amended as described in Section 2.2.2. The water flowed into the column by gravity with a flow rate near 0.5 ml/day. 2.2.5. Distribution of Tc among soil components: sequential extraction To determine the distribution of Tc among soil components we conducted sequential extraction of soil samples containing precipitated Tc(IV) with a procedure similar to that of Tessier et al. (1979), but modified to minimize oxidation of Tc(IV) with oxy-
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
gen. Thus, the leaching solutions were degassed with a vacuum pump, purged with N2 and kept in a glove box filled with N2. About 2.5 g of wet soil samples was weighed in centrifuge tubes and leaching solutions were added according to the procedure of Tessier et al. (1979). A 1 M magnesium chloride solution (pH 7) was used to extract exchangeable Tc (Tc-Ex) at 22 jC. To extract Tc together with carbonate, a 1 M sodium acetate solution adjusted to pH 5 with acetic acid was added to the residue from Tc-Ex experiment at 22 jC (Tc-Car). The resulting residue was then leached with 0.04 M hydroxylamine hydrochloride in acetic acid solution at 96 jC to dissolve Fe – Mn oxyhydroxides (Tc-Ox). After reaction, the residue was leached with hydrogen peroxide (25%) adjusted to pH 2– 3 with HNO3, heated to 85 jC, then leached with 3.2 M ammonium acetate in 20% (v/v) HNO3 to oxidize the organic matter (Tc-OM). Finally, a microwave digestion of the remaining residue with HF – HNO3 was conducted to extract the Tc held within the mineral matrix (Tc-Min). The Fe concentration was also analysed in the leaching solutions.
3. Results 3.1. Soil characterisation Chemical analyses of soils with X-ray fluorescence are given in Table 1. Soil S-3 is high in Ca and low in Si and Fe compared to S-5 and S-10. The concentrations of the rest of the elements are similar. Analyses under the scanning electron microscope showed the presence of clay minerals, calcite and quartz, in agreement with the chemical composition. A photograph of clay aggregate from soil S-5 is given in Fig. 1. The clay minerals consisted of Si, Al, Fe and little of Mg and K. Table 1 Chemical composition of soils obtained with X-ray fluorescence
Al Ca Fe K Mg Si
S-3
S-5
S-10
42.8 167.0 18.3 14.1 0.2 158.7
65.2 84.5 42.2 18.8 14.1 206.8
56.4 3.0 57.4 8.9 1.9 308.0
Data are given in mg/g.
259
Fig. 1. Scanning electron microscope photograph of a clay aggregate of soil S-5.
No reducing mineral phases such as iron carbonate or sulphide were detected in the samples. Gamma spectrometry analysis showed that the radioactivity levels of soils are typical of soils with low levels of contamination, with 137Cs specific activity between 7 and 28 Bq/kg (Aslani et al., 2003). Results of heat treatments of the soils are given in Table 2. Soil S-5 is rich in organic matter (18 wt.%), while soil S-3 has lost more than 18% in weight after heat treatment at 900 jC, due to decomposition of calcite. Under air, calcite begins to decompose at 740 – 750 jC and decomposition is terminated at 780 jC (Engler et al., 1988). 3.2. Batch experiments Tc concentrations in deionised water reacted with the three soils at 22 jC are given in Table 3. Sorption kinetics of anaerobic tests is plotted in Fig. 2. Soil S-5 showed the fastest decrease of Tc concentration with a final value of 5.10 8 M (Fig. 2a). Unlike the experiments with soils S-5 and S-3 where the pH was near to 8, sample S-10 showed the slowest reduction of Tc concentration, which is due to the low initial pH of
Table 2 Results of heat treatment of dry soils
% Organic matter % Volatile mineral matter
Heat treatment (jC)
S-3
S-5
S-10
550 900
5.64 18.64
18.00 6.17
3.78 0.58
260
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
Table 3 Tc concentrations in deionised water/soils S-3, S-5 and S-10 batch experiments under anaerobic and aerobic conditions Anaerobic (after 26 days) S-3 Tc initial (M) Tc final (M)
10
Aerobic (after 26 days) S-5
6
5.10
10 8
S-10 6
6.10
10 8
S-3
6
2.10
7
Aerobic (after 74 days)
S-5
S-10
10
6
10
10
6
3.10
5.5 (Fig. 2b). Low pH values inhibit development of environmental microorganisms such as denitrifying and sulphate-reducing bacteria (Maier, 2000). The decrease of Tc concentration correlates well with a drop in EH (from + 400 to + 40 mV) during water/soil interaction (Fig. 2b). To test whether the drop in EH is chemically or biologically controlled, we conducted experiments with sterilised batches. The results are given in Fig. 2c, which clearly shows that in the sterilised batch the Tc concentration remained unchanged while it significantly decreased in unsterilised batches. This result suggests that the decrease in EH is controlled by the growth of microorganisms that metabolise the natural organic matter present in sample S-5 (18%). The correlation of the decrease of Tc concentration with EH drop suggests the reduction of Tc(VII) into Tc(IV) by indigenous microorganisms such as metal-reducing bacteria. The unchanged Tc concentration in the sterilised sample suggests that Tc(VII) is not sorbed onto soil mineral particles. In addition, in the experiment where the Tc(VII) was reduced, the Fe concentration increased from 0.06 to 14.7 ppm at the end of the run, due to reduction of highly insoluble Fe(III) from the soil to Fe(II) with higher solubility. The increase in Fe concentration indicates the growth and activity of iron-reducing bacteria (Lovley, 1995). In the sterilised batches, the Fe concentration remained constant. This is in good agreement with Andra’s microbial work in soils similar to those used in our study showing the presence of a consortium of iron-reducing bacteria (Leclerc-Cessac, 2004). Data from batch experiments under aerobic conditions are given in Table 3. It took 74 days for Tc(VII) to be partially reduced. Hence, except for soil S-5 with high organic matter content, the Tc concentration remained unchanged after 26 days of reaction. In addition, the Tc concentration increased back from
6
7
S-3
S-5
10
6
10
6
10
6
8.10
10 8
S-10 6
6.10
10 7
6
8.10
7
3 10 7 to 6 10 7 M after 74 days of reaction, probably as a result of its reoxidation (Table 3). 3.3. Growth of bacteria in deionised water equilibrated with soils 3.3.1. Denitrifying bacteria The results of batch experiments described in Section 3.2 confirmed the implication of microorganisms in the speciation and immobilisationn of Tc. Thus, we tested the interaction of Tc(VII) with growing microorganisms including denitrifying bacteria and SRB, which are ubiquitous in subsurface soils and groundwaters (Abdelouas et al., 2000). We used only equilibrated water, as indicated in Section 2.2.2 to prevent interference from soil particles. The amendment of equilibrated water with nitrate, succinate and phosphate stimulated the growth of P. aeruginosa added to the water. Traces of certain elements in the equilibrated water were enough to sustain cell development. Results of denitrification at 28 jC showed the complete reduction of nitrate to nitrite within 26 h (Fig. 3). The accumulated nitrite was then reduced to N2 after 45 h. During denitrification the pH increased to pH 8, while the EH remained constant at about + 410 mV. The Tc(VII) concentration remained unchanged and the total amount of Tc was extracted with T.P.A., indicating no change in its speciation. Final cell concentrations, determined using plate-counting techniques, were of the order of 108 cells/ml. No biosorption of Tc on bacteria cells was observed. 3.3.2. Sulphate-reducing bacteria A suspension of SRB enriched from soil S-5 as described in Section 2.2.2 was inoculated at 28 jC in equilibrated water amended with sulphate, lactate, phosphate and Tc(VII). The results of the experiments are plotted in Fig. 4. We notice a reduction of Tc(VII) in
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
261
Fig. 3. Denitrification at 28 jC in deionised water equilibrated and separated from soil S-5. The solution was amended with nitrate, succinate and phosphate.
remained unchanged (7 10 7 M) between 39 and 49 days of reaction despite the low EH. This excludes the precipitation of phases with low solubility such as TcO2 but rather supports the sorption of Tc(IV) on SRB cells. The pH increased from 7.6 to 8.2. Extraction with T.P.A. showed that a significant fraction of soluble Tc is present in the reduced form Tc(IV), which is probably complexed with carbonate and organics released through bacteria metabolism. To study the adherence of Tc on the cells, the filtrate containing bacteria with precipitated Tc(IV) was leached with 1 M H2O2 as a strong oxidant. It took 24 h of reaction for Tc to be completely dissolved, which indicates that the sorbed Tc(IV) is strongly complexed with bacteria. In the autoclaved amended water there was no change in Tc
Fig. 2. (a) Tc concentration in deionised water/soil batch experiments under anaerobic conditions at 22 jC. (b) pH and EH in deionised water/soil batch experiments under anaerobic conditions at 22 jC. (c) Tc concentration in unsterilised and autoclaved (auto) deionised water/soil S-5 batch experiments under anaerobic conditions.
the pristine sample accompanied by a drastic drop in the EH down to 80 mV due to sulphate reduction to sulphide. Analysis of solution showed a drop in sulphate concentration. However, the Tc concentration
Fig. 4. Tc concentration in unsterilised and autoclaved (auto) deionised water equilibrated with S-5 soil at 28 jC. The water was amended with sulphate, lactate and phosphate and inoculated with a SRB consortium isolated from soil S-5.
262
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
concentration, in agreement with the high EH of + 420 mV. In addition, the experimental results show that equilibrated water alone is not able to reduce Tc(VII). 3.4. Biosorption of Tc on P. Aeruginosa Biosorption experiments with Tc(VII) onto P. aeruginosa, harvested at the stationary phase and washed, were conducted in 9 g/l NaCl at pH 2.5 and 6.9. The initial bacteria concentration, determined using plate-counting techniques, was 4.1 109 cells/ml. The experiments with Tc(VII) with initial concentrations of about 10-6 M showed no Tc sorption (Fig. 5), in agreement with literature data for experiments at pH >7 (Henrot, 1989; Abdelouas et al., 2002). In addition, experiments with higher Tc(VII) concentration (f 5.10 4 M) did not show any significant decrease in Tc concentration. It has been concluded that, under aerobic conditions where TcO4- is stable, soil biomass cannot contribute to Tc(VII) retention via biosorption and bioaccumulation. These results are in agreement with those described in Section 3.3.1. 3.5. Growth of indigenous bacteria in soil/deionised water batches and columns 3.5.1. Denitrifying bacteria An example of denitrification experiment with indigenous bacteria in soil S-5 at 25 jC is plotted in Fig. 6a. The results are similar to those obtained with a pure culture of P. aeruginosa but it took longer for denitri-
Fig. 5. Tc concentration in the presence of a suspension of P. aeruginosa at pH 2.5 and 6.9. The bacteria concentration was 4.1 109 cells/ml.
Fig. 6. (a) Denitrification at 25 jC in deionised water/soil S-5 batch amended with nitrate, succinate and phosphate. (b) Decrease of Tc concentration following denitrification at 25 jC in deionised water/ soils batches amended with nitrate, succinate and phosphate. Denitrification occurred during the first 96 h.
fication to be completed (108 h). This is due to the low initial bacteria concentration in the soils compared to the batch experiments with a pure culture of P. aeruginosa (Section 3.3). Autoclaved samples did not show any nitrate reduction because of sterilisation. Fig. 6b shows the Tc concentrations in batch experiments with soils S-3, S-5 and S-10, where denitrification was completed within 96 h. The figure shows that the Tc concentration remained fairly constant during denitrification (within the first 96 h). After denitrification, the Tc concentration decreased to about 10 8 M as a result of the EH decrease of > + 400 mV (at 96 h) to + 40 mV (at 306 h) (Fig. 6b). Sterilised batches did not show any reduction of Tc(VII). Geochemical modelling with MEDUSA code using the experimental data, including EH, pH and Tc concentration, suggested the precipitation of TcO2 at the end of the experiment (Fig. 7). The figure shows
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
263
t = 25°C Fig. 7. EH – pH stability diagram of Tc, including experimental data before and after denitrification in deionised water/soil S-5 batch experiments (see Fig. 5b).
that at the beginning of the reaction the experimental data (circles) fit in the stability domain of TcO4 , while at the end of the experiment the data fit well in the stability domain of TcO2. The drop in EH in experiments described in Fig. 6b correlates well with an increase in Fe concentrations (>10 ppm) as a result of Fe(III) reduction from soil particles to Fe(II) probably by a consortium of anaerobic bacteria. The results of the soil column experiment are similar to those from batch experiments. The Tc concentrations, as a function of time in the outlet
Fig. 8. Tc concentration in the outlet water of soil S-5 column at 25 jC. The column was amended with 10 6 M TcO4 only.
water, are plotted in Fig. 8. After 36 days, nearly one pore volume flowed through the column with a flow rate of 0.5 ml/day. After 36 days, the Tc concentration stabilised at about 2.5 10-8 M, indicating the precipitation of hydrated TcO2. The drop in Tc concentration was accompanied by the decrease of EH from > + 400 mV at the beginning of the experiment to 3 mV at 36 d. Furthermore, the experimental data using columns fit well in the stability domain of TcO2 as predicted by MEDUSA calculations. 3.5.2. Sulphate-reducing bacteria Fig. 9a shows the Tc concentrations in water/soil batch experiments amended with sulphate, lactate and phosphate to stimulate the growth of SRB at 25 jC. The initial Tc(VII) concentration was about 10 6 M. We sampled once the soil colour turned into black at day 10, due to iron sulphide precipitation as indicated by Abdelouas et al. (2000). It took 10 days for the Tc concentration to be reduced by a factor of 40, with final concentrations of the order of 10-8 M. The black precipitate was identified, as the iron sulphide mackinawite (FeS), by X-ray microanalysis under the transmission electron microscope. Mackinawite formed as a result of sulphate (solution) reduction to sulphide and Fe(III) (soil) reduction to Fe(II). In
264
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
Fig. 10. Tc concentration in the outlet water of soil S-5 column at 25 jC. The column was amended with sulphate, phosphate, lactate and TcO4 (10 6 M).
TcO2 formation, sorption and coprecipitation with iron sulphide. Coprecipitation of Tc(IV) with iron sulphide has been indicated by Wharton et al. (2000). Fig. 9. (a) Decrease of Tc concentration during sulphate reduction at 25 jC in deionised water/soil batch experiments amended with sulphate, lactate and phosphate. Auto = autoclaved. (b) A sulphatereducing bacterium covered with iron sulphide and isolated from soil S-5.
addition, mackinawite sometimes contains some reduced Tc on bacteria cells (Fig. 9b). The Fe concentration increase in solution was limited (from 0.07 to 0.77 ppm) due to mackinawite precipitation. Final pH values were about 8.2 and the EH was as low as 50 mV. The sulphate and Tc(VII) in the autoclaved sample were not reduced. Given the high concentration of organic matter in the soils (3.8 – 18%) it is likely that anaerobes other than SRB are also involved in Tc(VII) reduction. The presence of soil in the batch experiments, unlike in those using SRB alone (Section 3.3), enhanced the precipitation of Tc(IV) with final concentrations near the solubility of TcO2 after only 10 days of reaction. MEDUSA calculations using the experimental data of EH, pH and Tc concentration suggest the formation of TcO2. The results of the column experiment are similar to those of the batch experiments and are plotted in Fig. 10. The Tc immobilisation is even higher than in experiment without sulphate. The final Tc concentration is as low as 6.10 9 M, probably due to
3.6. Distribution of Tc among soil components: sequential extraction The sequential leaching was applied to two types of experiments with soil S-5: an experiment with SRB (described in Section 3.5.2) and an experiment without amendments (described in Section 3.2), where Tc(VII) was reduced by natural growth of anaerobic microorganisms. The amounts of Tc held in the different fractions of the soils are given in Table 4 and show good agreement between the nominal amount of Tc precipitated in each experiment and
Table 4 Results of the sequential extraction of 99Tc (Bq) from soil S-5 with (S5_SRB + Tc) or without (S5 + Tc) the growth of sulphate-reducing bacteria S5_SRB + Tc S5 + Tc Tc nominal Tc-Ex Tc-Car Tc-Ox Tc-OM Tc-Min Total % Error
6313 146 51 1359 3950 441 5947 6
5976 161 151 1987 3825 229 6353 6
In both experiments, 10 g of soil were mixed with 100 ml of pure water and TcO4 .
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
265
11b). The rest of the iron is present as oxyhydroxides (17 – 21%) and in organic matter (8– 9%). No effect of the presence of SRB on Tc distribution was observed. This confirms that the Tc is indeed strongly complexed with organic matter.
4. Discussion
Fig. 11. (a) Results of the sequential extraction of 99Tc from soil S-5 under anaerobic (S5 + Tc) and SRB (S5_SRB + Tc) conditions. ‘‘S5_SRB + Tc’’ is soil S-5 amended with deionised water containing sulphate, lactate, phosphate and TcO4 . ‘‘S5 + Tc’’ is soil S-5 amended with deionised water containing TcO4 . (b) Results of the sequential extraction of total Fe from soil S-5 under anaerobic (S5 + Tc) and SRB (S5_SRB + Tc) conditions. ‘‘S5_SRB + Tc’’ is soil S5 amended with deionised water containing sulphate, lactate, phosphate and TcO4 . ‘‘S5 + Tc’’ is soil S5 amended with deionised water containing TcO4 .
the total amount of Tc obtained during leaching. The error did not exceed 6%. The results are plotted in Fig. 11a. In both types of experiment more than 90% of the Tc is released together with Fe-oxyhydroxides (23 – 31%) and organic matter (60 – 66%). Much smaller proportions (3– 5%) are present as an exchangeable form and in carbonate minerals. The remaining 3– 7% is present in the residual fractions. Analysis of Fe concentration in the different fractions of the soils using ICP/MS showed that most of the iron is present in the mineralised residual fraction (60 –74%) (Fig.
This work shows that anaerobic microorganisms such as metal- and sulphate-reducing bacteria play a major role in Tc immobilisation in organic matter-rich subsurface environment if oxygen access is limited. Final Tc concentrations were of the order of 10 8 M, which reflects the solubility of amorphous and hydrated TcO2 (Burnett and Jobe, 1997; Meyer et al., 1991). In the absence of living microorganisms, organic matter was not able to immobilise Tc(VII). In a nitrate-rich environment the reduction of Tc(VII) was not possible, mainly as the EH remained largely positive (> + 400 mV). No significant bioaccumulation and sorption of TcO4 occurred on cells because of electrostatic repulsion at pH 6.9. Also at pH 2.5 Tc(VII) did not sorb onto cells, which underlines the noncomplexing nature of pertechnetate. The lack of TcO4 sorption was true for pure cultures of P. aeruginosa and mixed cultures grown from the soils. This result is in good agreement with published data with pure and mixed cultures of aerobic and denitrifying bacteria (Henrot, 1989; Abdelouas et al., 2002; Andre`s et al., 2001). In addition, inorganic soil particles including silicate (quartz, clays) and calcite did not retain pertechnetate ion. Under environmental conditions, where organic matter essential for microorganisms growth is lacking, pertechnetate is expected to migrate with surface and groundwater. However, the presence of plants, which are able to accumulate and retain TcO4 via biotransformation processes, may play an important role in the environmental fate of this radionuclide (Aarkrog et al., 1997; Uchida et al., 2000; Hattink et al., 2003). The growth of denitrifying bacteria had no effect on Tc speciation and its concentration remained unchanged in solution. This suggests that in nitrate- and organic matter-rich environments such as agricultural areas, denitrifiers may compete with other bacteria, including anaerobic microorganisms, for organic matter and inhibit the reduction and precipitation of Tc.
266
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
Under these conditions, TcO4 may be available for plant uptake or may be transported in flowing surface and possibly groundwaters. In the presence of a consortium of anaerobic bacteria, Tc(VII) was reduced and precipitated as hydrated TcO2. These anaerobes appear to use the soil organic matter to grow and reduce Tc(VII). The general assumption stating that under surface oxidizing conditions TcO4- is stable and Tc is largely mobile (Lieser and Bauscher, 1987), does not hold in our soils under oxygen poor conditions because of the high content in organic matter. Localized degradation of organic matter by indigenous bacteria creates localized anaerobic conditions that help reduction and precipitation of Tc. Biological reduction of Tc(VII) by anaerobic bacteria in particular iron-reducing bacteria including Shewanella putrefaciens and Geobacter metallireducens was shown by several authors (Lloyd and Macaskie, 1996; Lloyd et al., 2000; Wildung et al., 2000). In addition, Lloyd et al. (2000) showed that a mixed microbial culture enriched from radionuclide-contaminated sediment, containing a single Geobacter sp., were able indirectly to reduce and precipitate Tc via biogenic magnetite. They also showed the direct enzymatic reduction of Tc(VII) by the iron reducers. More recently, field work conducted by Istok et al. (2004) evaluated the potential of indigenous microbes to reduce U(VI) and Tc(VII) in the presence of high concentration of nitrate in a shallow aquifer. They showed that in the presence of added carbon sources reduction of nitrate and Tc(VII) was observed. They explained the reduction of Tc(VII) by the in situ growth of metal-reducing microorganisms. In amended soil/water batches SRB were able to grow and Tc(VII) was reduced. Abiotic reduction by Fe(II) is also expected (Lloyd et al., 2000). Similar results on the reduction of Tc(VII) by pure and mixed cultures of SRB were published by several authors (Lloyd et al., 1998, 1999; Abdelouas et al., 2002). Geochemical modelling of the experimental data in batch experiments with/without sulphate suggested the reduction of Tc(VII) to Tc(IV) and formation of TcO2. However, under denitrification conditions, the Tc speciation is dominated by TcO4 and no reduction of Tc(VII) occurred. Modelling confirmed the critical role of EH and to a lesser extent pH in Tc immobilisation.
Sequential leaching of the soils containing precipitated Tc(IV) showed that Tc was released at the same time as Fe-oxyhydroxides and organic matter including bacteria. The behaviour of Tc was comparable under iron- and sulphate-reducing bacteria conditions. Two hypotheses can be suggested to explain the leaching results: (1) dissolution of free TcO2, regardless of the nature of soil components (organic matter, iron sulphide, iron oxyhydroxides, etc.). This hypothesis is not supported by the work of Keith-Roach et al. (2003) who investigated Tc-binding in estuarine sediments from locations that have received 99Tc discharge from Sellafield. They found that, in certain sediments, the reduced Tc was indeed released together with Feoxyhydroxides and organic matter, while in others about 80% of the total Tc was released at the same time as carbonate. (2) Tc(IV) is mostly complexed with Fe-oxyhydroxides and organic matter and is released once these phases are dissolved. In both cases we can consider that the reduced Tc is strongly held in the soils. Under sulphate-rich environmental conditions, the reducing conditions due to sulphide formation are expected to contribute to Tc(VII) reduction and to the stability of Tc(IV)-rich precipitated phases. Furthermore, oxidation of iron sulphide leads to the formation of FeOOH, which strongly holds TcO2 (Wharton et al., 2000). Abdelouas et al. (1999b) showed that the presence of mackinawite together with bioprecipitated UO2 in soil columns ensured the long-term chemical stability of UO2. The authors indicated that 90% of the oxygen, in groundwater entering the column, was consumed by mackinawite oxidation.
5. Conclusion The main conclusions drawn from this work are: (1) In organic matter-rich soils Tc(VII) was reduced and probably precipitated as TcO2 with a final concentration of about 10 8 M. The Tc reduction resulted from the drop in EH under indigenous metal- and sulphate-reducing bacteria conditions. (2) Tc(VII) was not reduced in the presence of denitrifying bacteria. In addition, TcO4 did not sorb onto P. aeruginosa. (3) Organic matter and to a lesser extent iron oxyhydroxides play a crucial role in Tc immobi-
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
lisation in soils. Tc(IV) is strongly complexed with organic matter and bacteria. (4) Under sulphate-reducing conditions iron sulphides do not play a major role in Tc immobilisation but are expected to contribute to Tc(IV) stability by maintaining reducing conditions. The present study suggests that anaerobic conditions may very likely develop in the organic matterrich soils surrounding the Andra site due to the activity of anaerobic microorganisms. This may efficiently contribute to the reduction and immobilisation of Tc, hence reducing its bioavailability. However, fieldwork is needed to assess better the in situ geochemical conditions (EH, pH) of the soils in order to predict the behaviour of Tc.
Acknowledgements We would like to thank Andra for funding this work and for providing the soil samples. In addition, special thanks to A. Barreau for SEM and E. Gautier for TEM work performed at the Institut des Mate´riaux de Nantes Jean Rouxel (France).
References Aarkrog A, Chen QJ, Dahlgaard H, Nielsen SP, Trapeznikov A, Pozolotina V. Evidence of 99Tc in Ural River sediments. J Environ Radioact 1997;37:201 – 13. Abdelouas A, Lu Y, Nuttall E, Lutze W. Reduction of U(VI) to U(IV) by indigenous bacteria in contaminated groundwater. J Contam Hydrol 1998;35:217 – 33. Abdelouas A, Lutze W, Nuttall E. Uranium contamination in the subsurface: characterization and remediation. In: Finch R, Burns PC, editors. Reviews in Mineralogy: Uranium: Mineralogy, Geochemistry and the Environment, vol. 38. Washington, DC: Mineralogical Society of America; 1999a. p. 433 – 73. Abdelouas A, Lutze W, Nuttall E. Oxidative dissolution of uraninite precipitated on Navajo sandstone. J Contam Hydrol 1999b;36: 353 – 75. Abdelouas A, Lutze W, Gong W, Nuttall E, Strietelmeier BA, Travis BJ. Biological reduction of uranium in groundwater and subsurface soil. Sci Total Environ 2000;250:21 – 35. Abdelouas A, Fattahi M, Grambow B, Vichot L, Gautier E. Precipitation of technetium by subsurface sulfate-reducing bacteria. Radiochim Acta 2002;90:773 – 7. Andre`s Y, Redercher S, Gerente C, Thouand G. Contribution of
267
biosorption to the behavior of radionuclides in the environment. J Radioanal Nucl Chem 2001;247:89 – 93. Aslani MAA, Aytas S, Akyil S, Yaprak G, Yener G, Eral M. Activity concentration of caesium-137 in agricultural soils. J Environ Radioact 2003;65:131 – 45. Ben-said K, Fattahi M, Musikas Cl, Abbe´ JCh. A novel approach to the oxidation potential of the Tc(VII)/Tc(IV) couple in hydrochloric acid medium through the reduction of TcO4 by Fe2 + ion. Radiochim Acta 1998;83:195 – 203. Bondietti EA, Francis CW. Geologic migration potentials of Tc-99 and Np-237. Science 1979;203:1337 – 40. Burnett KB, Jobe DJ. Solubility and dissolution of TcO2(cr) in the absence and presence of CO23 (aq). Atomic Energy of Canada Limited (AECL). Final report RC-1664, COG-96-238-I; 1997. Davet P. Vie microbienne du sol et production ve´ge´tale. Paris: INRA Editions, 1996. p. 21–272. Ehrlich HL. Geomicrobiology of sulfur. Geomicrobiology 3rd ed. New York: Marcel Dekker; 1996. p. 508 – 59. Ehrlich HL. Microbes and metals. Appl Microbiol Biotechnol 1997;48:687 – 92. Engler P, Santana MW, Mittleman ML, Balazs D. Non-isothermal in situ XRD analysis of dolomite decomposition. Rigaku J 1988;5:3 – 8. Francis AJ, Dodge CJ, Meiken GE. Biotransformation of pertechnetate by clostridia. Radiochim Acta 2002;90:791 – 7. Gadd GM. Microbial formation and transformation of organometallic and organometaloid compounds. FEMS Microbiol Rev 1993;11:297 – 316. Ganesh R, Robinson KG, Reed GD, Sayler GS. Reduction of hexavalent uranium from organic complexes by sulfate-reducing and iron-reducing bacteria. Appl Environ Microbiol 1997;63: 4385 – 91. Hattink J, Harms AV, Goeij JJM. Uptake, biotransformation, and elimination of 99Tc in duckweed. Sci Total Environ 2003; 312:59 – 65. Henrot J. Bioaccumulation and chemical modification of Tc by soil bacteria. Health Phys 1989;57:239 – 45. Hunter JA, Kosawa A. Manganese, technetium and rhenium. In: Bard JA, Parsons R, Jordan J, editors. Standard potentials in aqueous solutions. New York: Institute of Pure and Applied Chemistry; 1985. p. 212. Istok JD, Senko JM, Krumholz LR, Watson D, Bogle MA, Peacock A, et al. In situ bioremediation of technetium and uranium in a nitrate-contaminated aquifer. Environ Sci Technol 2004;38: 468 – 75. Keith-Roach MJ, Morris K, Dahlgaard H. An investigation into technetium binding in sediments. Mar Chem 2003;81:149 – 62. Leclerc-Cessac E. Etude expe´rimentale du roˆle des microorganismes sur la mobilite´ du Se et Tc: Revue finale interlaboratoires. Report Andra No DS/TR/04-012; 2004. Lieser KH, Bauscher C. Technetium in the hydrosphere and in the geosphere: I. Chemistry of technetium and iron in natural waters and influence of the redox potential on the sorption of technetium. Radiochim Acta 1987;42:205 – 13. Lloyd JR, Macaskie LE. A novel PhosphorImager-based technique for monitoring the microbial reduction of technetium. Appl Environ Microbiol 1996;62:578 – 82.
268
A. Abdelouas et al. / Science of the Total Environment 336 (2005) 255–268
Lloyd JR, Nolting HF, Sole VA, Bosecker K, Macaskie LE. Technetium reduction and precipitation by sulphate-reducing bacteria. Geomicrobiol J 1998;15:43 – 56. Lloyd JR, Ridley J, Khizniak T, Lyalikova NN, Macaskie LE. Reduction of technetium by Desulfovibrio desulfuricans: biocatalyst characterization and use in a flowthrough bioreactor. Appl Environ Microbiol 1999;65:2691 – 6. Lloyd JR, Sole VA, Van Praagh VG, Lovley DR. Direct and Fe(II)mediated reduction of technetium by Fe(III)-reducing bacteria. Appl Environ Microbiol 2000;66:3743 – 9. Lovley DR. Microbial reduction of iron, manganese, and other metals. In: Donald L, Sparks L, editors. Advances in agronomy, vol. 54. New York: Academic Press; 1995. p. 175 – 231. Lovley DR, Roden EE, Phillips EJP, Woodward JC. Enzymatic iron and uranium reduction by sulfate-reducing bacteria. Mar Geol 1993;113:41 – 53. Maier RM. In: Maier RM, Pepper IL, Gerba CP, editors. Environmental microbiology. San Diego, USA: Academic Press; 2000. p. 363 – 400. Meyer RE, Arnold WD, Case FI, O’Kelley GD. Solubilities of Tc(IV) oxides. Radiochim Acta 1991;55:11 – 8. Omori T, Muraoka Y, Suganuma H. Solvent extraction mechanism of pertechnetate with tetraphenylarsonium chloride. J Radioanal Nucl Chem 1994;178:237. Paul EA, Clark FE. Soil microbiology and biochemistry. Academic press, INC, Harcourt Brace Jovanovich, Publishers; 1989. Postgate JR. The sulfate-reducing bacteria. 2nd edn. Cambridge, UK: Cambridge Univ. Press; 1984 (208 pp.). Puigdomenech I. HYDRA MEDUSA; make equilibrium diagrams
using sophisticated algorithms. Stockholm, Sweden: Royal Institute of Technology; 2001. Tagami K, Uchida S. Microbial role in immobilization of technetium in soil under waterlogged conditions. Chemosphere 1996;33:217 – 25. Tagami K, Uchida S. Chemical transformation of technetium in soil during the change of soil water conditions. Chemosphere 1998;5:963 – 71. Tessier A, Campbell PGC, Bisson M. Sequential extraction procedure for the speciation of particulate trace metals. Anal Chem 1979;51:844 – 51. Uchida S, Tagami K, Ru¨hm W, Steiner M, Wirth E. Separation of Tc-99 in soil and plant samples collected around the Chernobyl reactor using a Tc-selective chromatographic resin and determination of the nuclide by ICP-MS. Appl Radiat Isotopes 2000;53:69 – 73. Vernaz EY. Estimating the lifetime of R7T7 glass in various media. C R - Acad Sci., Phys 2002;3:813 – 25. Vichot L, Ouvrard G, Montavan G, Fattahi M, Musikas C, Grambow B. XAS study of technetium(IV) polymer formation in mixed sulphate/chloride media. Radiochim Acta 2002;90:575 – 9. Wharton MJ, Atkins B, Charnock JM, Livens FR, Pattrick RAD, Collison D. An X-ray absorption spectroscopy study of the coprecipitation of Tc and Re with mackinawite (FeS). Appl Geochem 2000;15:347 – 54. Wildung RE, Gorby YA, Krupka KM, Hess NJ, Li SW, Plymale AE, et al. Effect of electron donor and solution chemistry and products of dissimilatory reduction of technetium by Shewanella putrefaciens. Appl Environ Microbiol 2000;66:2451 – 60.