Factors controlling mobility of 127I and 129I species in an acidic groundwater plume at the Savannah River Site

Science of the Total Environment 409 (2011) 3857–3865

Contents lists available at ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r.com/ locate /sc i totenv

Factors controlling mobility of 127I and 129I species in an acidic groundwater plumeat the Savannah River Site

Shigeyoshi Otosaka a,b,⁎, Kathleen A. Schwehr a, Daniel I. Kaplan c, Kimberly A. Roberts c, Saijin Zhang a,Chen Xu a, Hsiu-Ping Li a, Yi-Fang Ho a, Robin Brinkmeyer a, Chris M. Yeager c, Peter H. Santschi a

a Laboratory for Oceanographic and Environmental Research, Department of Marine Sciences, Texas A&MUniversity, Building 3029, 200 Seawolf Parkway, Galveston, TX 77553, United Statesb Research Group for Environmental Science, Japan Atomic Energy Agency, Tokai-Mura, Ibaraki 319-1195, Japanc Savannah River National Laboratory, Aiken, SC 29808, United States

⁎ Corresponding author at: Research Group for EnviroEnergy Agency, Tokai-Mura, Ibaraki 319-1195, Japan. Te29 282 6760.

E-mail address: [email protected] (S. Ot

0048-9697/$ – see front matter © 2011 Elsevier B.V. Aldoi:10.1016/j.scitotenv.2011.05.018

a b s t r a c t

Article history:Received 15 February 2011Received in revised form 10 May 2011Accepted 10 May 2011Available online 8 June 2011

Keywords:Savannah River SiteIodine-129Iodine-127GroundwaterSpeciationIodideIodateOrgano-iodine

In order to quantify changes in iodine speciation and to assess factors controlling the distribution andmobilityof iodine at an iodine-129 (129I) contaminated site located at the U.S. Department of Energy's Savannah RiverSite (SRS), spatial distributions and transformation of 129I and stable iodine (127I) species in groundwaterwere investigated along a gradient in redox potential (654 to 360 mV), organic carbon concentration (5 to60 μmol L−1), and pH (pH 3.2 to 6.8). Total 129I concentration in groundwater was 8.6±2.8 Bq L−1

immediately downstream of a former waste seepage basin (well FSB-95DR), and decreased with distancefrom the seepage basin. 127I concentration decreased similarly to that of 129I. Elevated concentrations of 127I or129I were not detected in groundwater collected from wells located outside of the mixed waste plume of thisarea. At FSB-95DR, the majority (55–86%) of iodine existed as iodide for both 127I and 129I. Then, as the iodidemove down gradient, some of it transformed into iodate and organo-iodine. Considering that iodate has ahigher Kd value than iodide, we hypothesize that the production of iodate in groundwater resulted in theremoval of iodine from the groundwater and consequently decreased concentrations of 127I and 129I indownstream areas. Significant amounts of organo-iodine species (30–82% of the total iodine) were alsoobserved at upstream wells, including those outside the mixed waste plume. Concentrations of groundwateriodide decreased at a faster rate than organo-iodine along the transect from the seepage basin. We concludedthat removal of iodine from the groundwater through the formation of high molecular weight organo-iodinespecies is complicated by the release of other more mobile organo-iodine species in the groundwater.

nmental Science, Japan Atomicl.: +81 29 282 5171; fax: +81

osaka).

l rights reserved.

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

From 1955 to 1988, the F-area facilities at the U.S. Department ofEnergy's Savannah River Site (SRS F-area) produced radionuclides,including tritium (3H) and plutonium-239 (239Pu) for nuclearweapons, medical, industrial and scientific purposes. Seepage basinsof F-area (Fig. 1) routinely received acidic (pH=1–4) wastewatercontaining low levels of radionuclides and chemicals from the F-areaseparations facilities. In addition to 3H and Pu isotopes, fissionproducts (e.g., 90Sr, 95Zr, 106Ru, 129I and 137Cs), heavy metals (e.g., Cu,Hg and Pb), and synthetic organic compounds are reported as themain contaminants in the wastewater (Ryan, 1984; Kaplan et al.,1994; SRNS, 2008). The wastes were from nitric acid recovery unitsand evaporators used to concentrate uranium nitrate solutions(WSRC, 2001). The 2.7×104 m2 seepage basins were originally

unlined, 10-m deep ditches that received the acidic wastewaters(Killian et al., 1987). The F-area seepage basins are located about600 m upgradient from Fourmile Branch, which is a secondarytributary that flows about 6.4 km to drain into the Savannah River.

Seepage basins at the SRS F-area were closed in 1988 by addinglimestone and blast furnace slag to the basin bottom and then coveredwith a low permeability engineered barrier system to reducegroundwater infiltration. Limestone was used to promote theprecipitation of trivalent, tetravalent, pentavalent and hexavalentmetals, and blast slag was used to reduce radionuclides to immobilespecies such as Pu(IV) and Tc(IV). Since then, the pH values ingroundwater near the seepage basins have increased at a rate of0.02 units year−1, while concentrations of most radionuclides signif-icantly decreased (WSRC, 2006). One exception has been 129I, whereremobilization from soil has been detected and attributed to rising pHvalues of the groundwater (Kaplan et al., 2011) and flushing theseepage basin with large volumes with background, pH 5.5,groundwater (Denham and Vangelas, 2008).

In 2004, barrier walls and base injection gates (Fig. 1) wereinstalled 300–500 m downgradient of the primary seepage basin to

Fig. 1. Locations of the SRS F-area seepage basin, barrier walls, base injection gates and sampling wells in this study.

Fig. 2. Relationships between pH and oxidation-reduction potential (ORP) in the SRSF-area groundwater (closed symbols) and from the upper aquifer in the SRS E-area,adjacent to the F-area. The inscribed areawas adopted fromBaas-Becking et al. (1960)and is representative of most sediments. Lines representing O2/H2O and the H2/H2Ocouples provide limits of oxidation-reduction potentials in aqueous systems (organo-iodine was not considered in these calculations).

3858 S. Otosaka et al. / Science of the Total Environment 409 (2011) 3857–3865

adjust groundwater pH for better management of 3H and metalreleases to the seep line. Since July 2005 a total of 4×107 l of basesolution (pH=10) has been added through the gates. After theinstallation of barrier walls, it was found that the water tableupstream from the walls was elevated and that in the downstreamwater table was lowered (WSRC, 2006). This was presumed to be dueto a “flattening” effect by blocking of the groundwater stream.

Iodine, a biophilic elements, has a stable isotope with massnumber 127 (127I) and 23 radioisotopes with mass number rangingfrom 125 to 135 that are mainly produced by atomic fission ofuranium and trans-uranium isotopes. Because of its long half-life,excessive inventory and toxicity, iodine-129 (129I:half-life15.7×106 years) is among the top three risk drivers for wastedisposal at the Yucca Mountain, Idaho National Laboratory andSavannah River Sites (DOE, 2002, 2011; Santschi and Schwehr,2004), and is also regarded as one of themost important radionuclidesthat should be assessed for its biogeochemical cycling on a global scale(UNSCEAR, 2000; Hou et al., 2008; Hu et al., 2009).

Iodine in aquatic systems on the earth's surface can be groupedinto three species; iodide, iodate and organo-iodine. Since thespeciation of iodine is affected by various parameters of waterchemistry, such as temperature, pH, redox potential and organicmatter productivity, it is necessary to understand the factors thataffect its speciation to address the behavior of 129I in the environment.With regard to inorganic iodine, iodine should theoretically exist asiodide in groundwater and anoxic surface waters (Fig. 2). Recentstudies on SRS groundwater, however, reported that a significantproportion of 129I existed as iodate (Schwehr et al., 2009; Zhang et al.,2010). The processes that produce 129I-iodate under these conditionshave not been quantitatively assessed.

Several studies have demonstrated that significant proportions ofiodine in groundwater and surface water exist as organo-iodine(Oktay et al., 2001; Schwehr et al., 2005, 2009). Some laboratoryexperiments have reported that the primary product of iodideoxidation (e.g., hypoiodous acid) is immediately incorporated intoorganic matter in soils (Bichsel and von Gunten, 1999; Warner et al.,2000). Significant amounts of iodine are eventually incorporated intosoils on time scales ranging from a few days to several months

(Ashworth et al., 2003; Hou et al., 2008; Shimamoto and Takahashi,2008; Schwehr et al., 2009; Hu et al., 2009; Yamaguchi et al., 2010).Furthermore, Amachi et al. (2001) have shown that a wide range ofbacteria are capable of producing volatile methyl iodide in soils,which, however, are orders of magnitude lower in concentration thaniodide, iodate or non-volatile organo-iodine species.

The degree of incorporation of dissolved iodine into soil dependson the species. Distribution coefficients (Kd) of iodide (~1 Lkg−1) aremuch lower than those of iodate and organo-iodine (103 Lkg−1)(Fukui et al., 1996; Kaplan et al., 2000; Schwehr et al., 2009). It is alsoknown that concentrations of iodine in soils are more than threeorders of magnitude higher than that in the base rocks, which is

3859S. Otosaka et al. / Science of the Total Environment 409 (2011) 3857–3865

because iodine supplied to the land surface by wet and dry depositionis selectively accumulated in soil organic matters (Muramatsu et al.,1990; Yoshida et al., 1992). Several natural organic matter (NOM)fractions, such as humic substances play an important role insequestering iodine (Readlinger and Heumann, 2000; Warner, et al.,2000). While immobilization of iodine is controlled by abiotic andbiotic processes, both of these processes are strongly pH dependent(Steinberg et al., 2008; Li et al., 2011).

In early studies on the interactions of iodine with soil organicmatter it was suggested that iodination of NOM, i.e., iodineincorporation via covalent C–I binding to aromatic moieties, isdominated by high molecular-NOM (e.g., Wachsmuth, 1967). How-ever, recent studies have reported that fulvic acid and waterextractable colloids, which have relatively low molecular weight,also comprise a significant fraction of the iodinated organic matter; inaddition, the importance of re-mobilization of such organo-iodinespecies as carriers of iodine in aquatic systems has been demonstrated(Xu et al., 2011). Furthermore, Schwehr et al. (2009) havedemonstrated that studies of iodination and re-mobilization of theorgano-iodine in groundwater need to be carried out at near-ambientconditions with regard to iodine concentration (~100 nmol L−1).They reported that sorption studies conducted at elevated iodineconcentrations masked the importance of organo iodine speciationand underestimated iodine Kd values. This experiment underscoresthe importance of studying iodine speciation at ambient concentra-tions. Schwehr et al. (2009) and Zhang et al. (2010) also reported thatsignificant amounts of iodate can be produced as a consequence ofiodination.

Speciation of iodine and the subsequent removal of iodine fromthe groundwater were both controlled by in-situ reactions of iodine inthe aqueous phase, as well as soil-NOM-water interactions. Thus, aproper understanding of the biogeochemical characteristics of a givengroundwater system is required to trace the behavior of iodine/radioiodine species. The aims of this paper are thus to (1) investigatespatial changes in concentrations of iodine and radioiodine species ingroundwater at the SRS F-area as a function of distance from thesource region, and (2) to infer factors controlling the distribution andmobility of iodine/radioiodine. Our approach was to measure 129I and127I speciation (iodide, iodate, and organo-iodine) in a dozen wellsdown gradient from the seepage basin. Additionally, a number ofancillary aqueous chemistry parameters were measured (includingpH, oxidation-reduction potential, dissolved organic carbon, andtritium) to provide insight into the aqueous geochemical conditionsacting on iodine to influence its speciation. For this field study weassessed iodine speciation under a wide range of ambient conditions,without artificially influencing these conditions, to evaluate the

Table 1Hydrographical and chemical properties of SRS F-area groundwater.

Samplinglocation

Wellcategorya

Samplingdate

Distance from seepagebasin (m)

T(

FSB-95DR 1 8/1/2009 21 N8/3/2010 21 2

FSB-126D 1 8/3/2010 241 2FSB-117D 1 7/12/2010 269 2FSB-79 2 7/9/2010 379 2FPZ-6A 2 1/6/2010 532 N

2/8/2010 532 N7/21/2010 532 1

FSB-104D 3 7/12/2010 432 2FSB-138D 3 8/24/2010 521 2FPZ-2A 3 7/21/2010 538 1FPZ-3A 3 7/21/2010 630 1FSB-109D 4 7/9/2010 60 2FSB-118D 4 7/8/2010 290 1FSB-120D 4 7/13/2010 303 2

See Fig. 1 for the positional relation.a Sampling wells are categorized into four regional groups: (1) Above barrier wall/gate,

degree to which both isotopes were equilibrated by comparing theiodine speciation of 127I with that of 129I. Finally we evaluated the datausing ternary diagrams and thermodynamic calculations.

2. Methods

2.1. Field sampling

Groundwater samples were collected from 12 wells in the SRSF-area (Table 1; Fig. 1). Groundwater near the seepage basins atthe F-area flows in the southwestward direction (Killian et al.,1987; WSRC, 2006). Considering the direction of groundwater flowand the geometry of the barrier walls and base injection gatesdownstream of the basins, representative sampling wells wereselected and classified into four regional groups: (1) above barrierwall/gate, (2) below barrier wall, (3) below base injection gate,and (4) outside of the mixed waste plume. Sampling was carriedout between July and August, 2010. At wells FSB-95DR and FPZ-6A,groundwater samples were also collected in August 2009, andJanuary–February 2010, respectively. Groundwater was purged andsampled from the wells using a variable speed submersible pump(Grundfos Redi-flo 2, Olathe, KS) or a battery-operated peristalticpump at a flow rate of approximately 1 L min−1 during sampling.The purged water was passed through a YSI flow through cell(Model 6820 and 6920, Yellow Springs, OH) in which temperature,pH, oxidation-reduction potential (ORP) and conductivity weremonitored. Once roughly 2 well volumes were passed through andthe pH and conductivity readings were stable, the well wassampled. Surface water from the Fourmile Branch was alsocollected using a plastic pump. Surface water and groundwatersamples were filtered through a 0.45 μm cartridge filter, filled in anamber container, and stored in the refrigerator until analysis.

2.2. Chemical analysis

Analyses of 127I and 129I were carried out using the method ofZhang et al. (2010). For iodide analysis, a 5 mL aliquot of eachgroundwater sample was prepared by adding 1% acetic acid andphosphate buffer to near neutral pH, followed by addition of N,N-dimethylaniline solution and 2-iodosobenzoate solutions. To extractthe iodinated dimethylaniline, 0.5 mL of cyclohexane was added tothe container and shaken for 1 min. A sub-sample (1 μL) of thecyclohexane phase was injected into a gas chromatography massspectrometer (GC-MS; Finnigan Trace GC and Polaris Q EI-MS withTR-5MS capillary column), and iodinated dimethylaniline (mass

emp pH ORP Tritium DOC°C) (SHE; mV) (kBq L−1) (μmol L−1)

o data No data No data No data No data1.9 3.20 654 23.3±2.4 60.5±2.22.9 3.42 491 70.7±7.1 29.9±1.30.7 3.60 494 27.1±2.7 8.3±0.01.8 5.20 428 25.3±2.6 16.2±0.4o data No data No data No data No datao data No data No data No data No data8.0 3.81 391 95.1±9.5 13.8±1.12.0 6.66 402 0.12±0.03 22.6±1.00.9 4.18 384 8.47±0.87 15.5±0.58.3 6.77 360 2.70±0.29 5.5±0.39.5 4.08 487 5.29±0.55 33.4±2.23.8 5.53 381 1.99±0.22 35.6±1.79.2 4.97 392 0.40±0.06 16.6±1.51.7 5.20 550 1.41±0.16 13.6±0.8

(2) Below barrier wall, (3) Below base injection gate, and (4) Outside of waste plume.

3860 S. Otosaka et al. / Science of the Total Environment 409 (2011) 3857–3865

number 247 for derivatives of 127I and 249 for 129I, respectively) wasmeasured as described in Zhang et al. (2010).

To determine the total inorganic iodine (TII: iodate plus iodide)aqueous samples, iodate was first reduced to iodidewith 0.01 mol L−1

sodium meta-bisulfite (pHb2, prepared with 1 mol L−1 hydrochloricacid) and iodide was measured as described above. Iodate was thencalculated by: TII — iodide = iodate.

Organo-iodine was determined by calculating the differencebetween total iodine and TII. Total iodine was determined by thecombustion of water samples. An aliquot (1–2 mL) of the watersample was evaporated and combusted under oxygen gas over avanadium pentoxide catalyst at 900 °C for 10 min. Combusted iodine

Fig. 3. Levels of (a) pH, (b) ORP, (c) 3H, (d) DOC, (e) total 127I, (f) total 129I, and (g) the 129I/Data obtained in July–August 2010 are shown in the figures.

gas was trapped in ultrapure water, and iodate in the receiving waterwas measured according to the procedures described above. Recoveryyield of the combustion of iodinewasmonitored by liquid scintillationcounting using 125I as the tracer.

Groundwater samples for iodide and iodate measurements werepurified through Strata SAX SPE columns (anion exchange column;Phenomenex, Torrance, CA) before above mentioned processes toeliminate interferences from inorganic ions and charged organiccompounds in samples. A 2,4,6-tribromoaniline solution was used asan internal standard for the GC-MS measurement. Measurementswere taken in duplicate or triplicate for all iodine species and for allsamples.

127I ratio in groundwater at SRS F-area as a function of distance from the seepage basin.

3861S. Otosaka et al. / Science of the Total Environment 409 (2011) 3857–3865

pH and ORP were measured on purge water using a flow throughcell (closed system) with portable water quality sensors (YSI 6820and 6920, respectively). 3H concentrations were determined by liquidscintillation spectrometry (Perkin Elmer Tri-Carb). Dissolved organiccarbon (DOC) was determined using a Shimadzu TOC-5000 analyzerafter removing inorganic carbon in the sample using 1 mol L−1

hydrochloric acid. The relative standard deviation was within 7% forall samples.

3. Results

3.1. Spatial distribution of hydrographical and chemical properties

Hydrographical and chemical properties of SRS F-area groundwa-ter are listed in Table 1. The horizontal distance between the seepagebasin and each well, based on GIS data, is also shown in Table 1. Therelationship between groundwater pH values and distance from theseepage basin is shown in Fig. 3(a). A pH value immediatelydowngradient of the seepage basin (FSB-95DR) is 3.2, which issignificantly lower than that measured in groundwater collected fromwells near the seepage basin but outside of the mixed waste plume(i.e., FSB-109D, FSB-118D and FSB-120D: pH=5.0–5.5). At a wellimmediately downgradient of the barrier wall (FSB-79), groundwaterpH was similar to values at “background” sites (pH=5.0–5.5), exceptat a well near the seepline, downgradient of the wall (FPZ-6A), wherethe groundwater pH was 3.8. Not surprisingly, the highest ground-water pH value (6.7) was measured at a well immediately down-gradient of the base injection gate (FSB-104D).

ORP in groundwater was highest at FSB-95DR (654 mV) anddecreased with distance from the basin toward the barrier wall(402 mV in well FSB-104) (Fig. 3b).

3H concentrations were highest in wells near the source, i.e., thebasins (23–27 kBq L−1) and lower at wells below the base injectiongate (0.1–8 kBq L−1) and outside of mixed waste plume (0.4–2 kBq L−1) (Fig. 3c). However, 3H concentrations did not decreaseas a function of distance from the seepage basins, and locally high 3Hconcentrations were observed at wells near the barrier wall, such as atFSB-126D (71 kBq L−1) and FPZ-6A (95 kBq L−1). It must be kept inmind that 3H originally placed in the basins in 1955 has had time totravel the full length of the study site and reached Fourmile Branch;the travel time is ~12 years (Killian et al., 1987). As such, 3H is notmoving on the site as a front, rather the entire length of the site isconsume by a 3H plume.

Table 2Concentrations of 127I and 129I species in groundwater collected from SRS F-area.

Samplinglocation

Samplingdate

127I concentration (nmol L−1)

Iodide Iodate

FSB-95DR 2009/8/1 172⁎ 0.0⁎

2010/8/3 189±13 70.2±24.6FSB-126D 2010/8/3 80.6±4.5 12.0±10.4FSB-117D 2010/7/12 75.6±4.3 60.6±16.0FSB-79 2010/7/9 56.1±1.9 27.8±5.1FPZ-6A 2010/1/6 55.3⁎ 20.8⁎

2010/2/8 95.7⁎ 53.4⁎

2010/7/21 72.9±4.2 34.9±13.6FSB-104D 2010/7/12 13.6±1.6 75.0±4.3FSB-138D 2010/8/24 12.9±0.2 35.1±3.0FPZ-2A 2010/7/21 0.8±1.8 11.0±2.3FPZ-3A 2010/7/21 5.1±2.3 7.5±2.5FSB-118D 2010/7/8 3.7±1.7 27.0±5.2FSB-109D 2010/7/9 3.0±3.0 4.6±3.2FSB-120D 2010/7/13 16.2±3.4 21.7±10.0Fourmile branch 2010/7/13 39.4±3.5 18.6±4.6

BDL: Below detection limit.Errors indicate 1-sigma statistics of duplicate analysis.For conversion of 129I concentration, 1 Bq L−1 equals to 1.19 nmol L−1.⁎ Indicate data where n=1.

DOC concentrations in groundwater were higher at FSB-95DR, andgenerally decreased with distance from the seepage basin (Fig. 3d).However, the DOC gradient did not show any significant correlationwith pH or 3H levels of the groundwater. Surface water collectedfromFourmile Branch had 7–10 times higher DOC concentration thanF-area groundwaters.

3.2. Spatial distribution of 127I and 129I concentrations

Relationships between 127I and 129I concentrations and horizontaldistance from the seepage basins observed in July–August 2010 areshown in Fig. 3(e) and (f), respectively. 129I concentrations ingroundwater were 10.6±0.4 Bq L−1 at FSB-95DR and decreasedwith distance downstream from the basin. Except for the highconcentration of both 129I and 3H at the seepage well (FPZ-6A), therewas no correlation between the distributions of these two radionu-clides. Concentrations of both 129I and 127I in groundwater sampleswere higher in the wells adjacent to the seepage basin. Thedistribution of 127I (Fig. 3e) was similar to that of 129I (Fig. 3f), andsignificantly high 127I groundwater concentrations were not detectedfrom wells located outside of the mixed waste plume (i.e., FSB-109D,FSB-118D and FSB-120D). Isotopic ratios of 129I/127I ranged between0.04 and 0.01, and showed a gradient from the seepage basin in thedownstream direction, except for FPZ-6A where high 3H and 129Iconcentrations were detected (Fig. 3g).

Although iodide is generally considered the dominant species ofiodine in groundwater based on thermodynamic principles (Fig. 2),significant concentrations of iodate were detected in F-area ground-waters (Table 2 and Fig. 4). At FSB-95DR, themajority (51–69%) of 127Iexisted as iodide. However, iodide contributed only 10% downstreamof well FSB-104D, a well located near the injection wall. From the datashown in Fig. 4(a) and (b), the rates of decrease of 127I concentrationsas a function of distance from the seepage basin (whereby theconcentration at FSB-95DR was considered as 100%) were 0.15%m−1

for iodide, 0.10%m−1 for iodate, and 0.12%m−1 for organo-iodine.With regard to 129I (Fig. 4c and d), the decreasing rates were0.14%m−1 for iodide, 0.10%m−1 for iodate, and 0.11%m−1 for organo-iodine. Rates of decreaseof 127I-iodidewere slightlyhigher than thoseof127I-iodate and 127I-organo-iodine, and were in the same range for 129I.These decreases of species concentrations as a function of distance canbe attributed to: (1) loss from the aqueous phase due to sorption to theaquifer sediment, (2) dilution, and (3) transformation to another iodinespecies.

129I concentration (Bq L−1)

Organo I Iodide Iodate Organo I

77.2⁎ 5.65⁎ 0.10⁎ 0.83⁎

110±28 5.81±0.68 1.52±0.72 3.23±0.8216.8±11.0 1.15±0.18 0.91±0.34 BDL73.6±17.9 1.29±0.45 0.97±0.46 2.52±0.7456.1±23.0 1.29±0.28 −0.10±0.29 0.59±0.3032.6⁎ 0.51⁎ 1.17⁎ 0.55⁎

46.8⁎ 2.07⁎ 1.33⁎ 1.01⁎

32.4±14.8 1.41±0.16 1.09±0.30 2.23±0.3448.8±6.5 BDL 0.73±0.18 BDL19.9±5.0 BDL 0.34±0.10 0.81±0.276.1±2.1 BDL BDL BDL24.1±2.7 BDL BDL BDL2.7±5.4 BDL BDL BDL35.5±3.0 BDL BDL BDL25.0±10.1 BDL BDL BDL82.6±7.9 BDL BDL BDL

Fig. 4.Horizontal distributions of 127I and 129I species in groundwater at the SRS F-area obtained in July–August 2010. (a) and (c) show data obtained from a transect crossing barrierwall (FSB-95DR, FSB-126D, FSB-79, FPZ-6A, and Fourmile Branch). (b) and (d) show data from a transect crossing the base injection gate (FSB-95DR, FSB-104D, FSB-138D, FPZ-2A,FPZ-3A and Fourmile Branch).

3862 S. Otosaka et al. / Science of the Total Environment 409 (2011) 3857–3865

4. Discussion

4.1. Source of iodine/radioiodine in the groundwater at SRS F-area

From the horizontal distribution of 129I shown in Fig. 3(f) and(g) we can conclude that there is a single source of 129I at the studysite. Furthermore, a ternary diagram of iodine species (Fig. 5),indicates that iodine was supplied as iodide, which is the most mobile

Fig. 5. Ternary diagram for iodide, iodate and organo-iodine species showing contr

species. Although 3H is known as a conservative tracer that originatesfrom the seepage basins and has a low Kd value of essentially zeroL kg−1, there was little similarity between distributions of 3H andiodide (Figs. 3c and 4c–d). This inconsistency between distributions of3H and 129I indicates that different processes are controlling thetransport of these two radionuclides.

Kaplan et al. (2011) reported that between 1993 and 2009 129I(inorganic forms only) concentrations in groundwater at FSB-95DR

ibutions of (a) 127I and (b) 129I in groundwater collected at the F-area of SRS.

3863S. Otosaka et al. / Science of the Total Environment 409 (2011) 3857–3865

increased on average rate of 1.4 Bq L−1year−1. In contrast, 3Hconcentrations in FSB-95DR decreased with an effective half-life ofabout 2 years (Savannah River National Laboratory, 2010), which ismuch shorter than the radiological half-life of 12 years. These resultsindicate that 3H groundwater concentrations in this well decreased byradioactive decay as well as mixing with uncontaminated groundwa-ter at a rate of 0.3 year−1 (equivalent to a turnover time of 3.4 years).Denham and Vangelas (2008) reported that reduction of the fluxof vadose zone water resulted in an increase of groundwater pHbeneath the seepage basin. This result indicates that a relativeproportion of uncontaminated groundwater increased after cappingof the basin. Therefore, we can assume that current supply ofradionuclides originating from the seepage basin to the groundwaterpathway is much smaller than that of uncontaminated groundwater.The higher concentrations of 129I in the groundwater beneath theseepage basin could indicate that the primary source of 129I is notgroundwater emanating from the seepage basin itself. More likely,however, is that 129I concentrations have resulted from theremobilization of solid-phase-bound 129I.

Surprisingly, the concentrations of stable iodine (127I) alsodecreased sharply with distance away from the basin (Fig. 3e), andsignificantly higher 127I concentrations (370 nmol L−1) were notobserved even in the second well down gradient from the basin(FSB-126D), which had an 127I concentration (109 nmol L−1) thatapproach those of background levels and Fourmile Branch (Fig. 3c).During the 1960s, stable iodine was used as a booster in the SRSfacilities to reduce Pu(III) in nitric acid solution to metallic Pu (Orth,1963). By this procedure, 0.6 mol of iodine was theoretically used toproduce 1 mol of Pu. Concentrations of 239+240Pu in present-daygroundwater near the seepage basin are reported to be 200 atomsL−1 (Dai et al., 2002; Buesseler et al., 2009), which is much lowerthan that of 127I (1015–1017 atoms L−1). If ~1% of Pu was contained inthe waste, the ratio of Pu : I should be close to 1/100 : 0.6, which ismuch higher than observed Pu : I ratio=200 : 1015. The reason isthat themajority of Pu in thewaste is reduced and precipitated in thebasin sediments or the added iron sulfide slag amendments. Thus,chemical wastes disposed by this procedure are not regarded as theprimary source of the high concentrations of 127I in the present-daygroundwater at this site. One other possibility is that the 127I mayhave unintentionally been added as an impurity with the enormousamounts of nitrate and nitrite salts; about 2.8×109 g nitrate wereadded to the basins between 1961 and 1983 (Killian et al., 1987).

Under environmental conditions, the speciation of inorganiciodine is generally controlled by pH and ORP. As shown in Fig. 3(a),pH values in groundwater increased with distance from the seepagebasin, ranging from 3.2 to 6.8 within the study site. Theoretically, anincrease of pH enhances the proportion of groundwater iodate(Fig. 2), thus it is reasonable to expect significant contributions ofiodate in groundwater in downstream wells: as was measured in theF-area plume (Figs. 4 and 5). On the other hand, ORP decreases withdistance from the basins, ranging from 654 to 360 mV, and isinconsistent with the observed spatial trend of iodine speciation.This range of ORP values suggests that the aquifer ranges from oxic tosub-oxic conditions and that as the plume moved toward the seeplineit became progressively more reduced. As the iodine plume becamemore reduced, its pH also increased. These results indicate thatspeciation of inorganic iodine speciation in the F-area groundwatersystem is more controlled by pH than ORP.

The pH in groundwater at FSB-95DR, the well immediatelydowngradient of the seepage basin, increased by 0.7 units in the last17 years (Kaplan et al., 2011). From a series of batch experimentsusing a simulated soil-groundwater system, it has also been suggestedthat the desorption Kd value of iodine from soil to groundwater(indicating the potential soil retardation of iodine) decreases withincreasing pH and the tendency is remarkable in the low pH ranges(pH=3–4) (Kaplan et al., 2011). Considering the Kd-pH correlation

and the temporal change in the pH of F-area groundwater, it could beestimated that the desorption Kd values in the basin sediment/precipitants decreased from 50 Lkg−1 to 10 Lkg−1 in the 16 yearsbetween 1993 and 2009. The iodine concentration in soil at SRS isabout 1 mmol kg−1-dry weight (Hu et al., 2009), which is higher thanthat of groundwater. Based on the iodine content in soil and thedesorption Kd value, it can be estimated that 80 μmol of iodine wasreleased per unit volume (1 L) of contacting groundwater, and therelease rate of iodine from the soil was 5 μmol L−1year−1. Assumingthat a physical turnover rate of upstream groundwater is 0.3 year−1,as described above, the amount of iodine released into groundwaterbeneath the seepage basin over 16 years is estimated to have been12 μmol L−1. This amount is N30 times higher than the 127Iconcentration in groundwater at FSB-95DR (370 nmol L−1: Fig. 3e).Although the estimation is based on a laboratory experiment, thepotential release of iodine from soil due to increasing pH values wouldcertainly be sufficient to raise 127I and 129I concentrations in groundwaters downgradient of the basin.

4.2. Factors controlling the speciation of iodine

Considering the range of pH and ORP values in F-area groundwatersamples, iodide should be the dominant iodine species (Fig. 2).However, in groundwater in regions adjacent to the F-area seepagebasin, the dominant iodine species is iodide only in upstream areas,iodate or organo-iodine is dominant in downstream areas. Schwehret al. (2009) carried out a contact experiment of sediment andgroundwater taken from SRS, and found that (1) groundwater iodideis significantly retarded by SRS sediment, (2) the retardation couldonly be measured at near-ambient iodide concentration(~100 nmol L−1), and (3) groundwater after contacting the sedimentcontained appreciable amounts of organo-iodine. As shown in Fig. 4(a) and (b), iodide concentrations in F-area groundwater aresufficiently low to result in retardation. If the iodine retardation iscontrolled by hydrohalogenation at sediment surfaces, the expectedtime scale is days to weeks (Schwehr et al., 2009; Yamaguchi et al.,2010) to 100 days (Ashworth et al., 2003). Since the flow rate ofgroundwater at F-area is 0.15 m year−1 (Darcy flux: Killian et al.,1987;WSRC, 2006), it is not surprising that iodide can be transformedto organo-iodine in the upstream regions. Assuming that iodine ingroundwater at F-area was supplied as iodide (i.e., iodide : iodate=100 : 0) from the seepage basin, the contribution of 127I and 129I in anend-member groundwater would be estimated to iodide : iodate :organo-iodine=80 : 0 : 20 from upgradient data in Fig. 5(a) and (b),indicating the existence of organo-iodine in the end-membergroundwater. Both iodide and organo-iodine could be supplied orproduced in the groundwater at the seepage basin or immediatelydowngradient (0–200 m), and it is the differential rate of removal ofthese species as they travel through the aquifer that dictates theirproportion in groundwater below the barrier or gate (N300 m fromthe seepage basin). The rate of decrease of organo-iodine withdistance from the basin (0.11–0.12%m−1) is lower than that of iodide(0.14–0.15%m−1), either favoring iodide removal from the aqueousphase over organo-iodine or iodide transformation to (iodate or)organo-iodine.

A higher content of DOC is observed in groundwater near theseepage basin (Fig. 3d). This is not unexpected as several xenobioticcompounds, such as tributylphosphate (MW=266) may have beenpresent in wastewater in the seepage basins. Considering that themajority (40–90%) of organo-iodine is found in the b1 kDa fraction(Zhang et al., 2010), iodination of such xenobiotic compounds couldcontribute to elevated concentrations of organo-iodine in groundwa-ter near the seepage basin. However, there was no significantcorrelation between DOC and organo-iodine concentration (r=0.30,n=6), indicating that iodination of the NOM is partially controlled bythe composition of the organic matter, such as aromatic content (Xu

3864 S. Otosaka et al. / Science of the Total Environment 409 (2011) 3857–3865

et al., 2011). Additionally, the concentrations of such xenobioticcompounds reported in groundwater (~0.6 μmol L−1) are muchlower than that of DOC (~60 μmol L−1) that we measured. Finally,at wells FPZ-3A and FSB-109D, where higher DOC concentrationswere observed (33 and 36 μmol L−1, respectively), more than 60%of the total iodine existed as organo-iodine. Since groundwater atFSB-109D is not contaminated with 3H, 129I, or xenobiotics, we canconclude that the production of organo-iodine in groundwater iscontrolled by ambient natural organic matter (NOM): 127I specia-tion in FSB-109D was composed of 7% iodide, 11% iodate and 82%organo-iodine (Table 2).

It is well known that humic substances play an important role inuptake of iodine by NOM in soils (Readlinger and Heumann, 2000;Warner, et al., 2000). As demonstrated by Xu et al. (2011), a part of theiodinated NOM can be released from the soil and remobilized as“dissolved” organo-iodine. The kinetics of the transfer of iodinebetween soil and groundwater is pH dependent (e.g., Steinberg et al.,2008), as are the chemical properties of NOM (e.g., Sparks, 2003).Furthermore, Li et al. (2011) also reported that iodination of bacteria,isolated from F-area aquifer, occurred more efficiently at lower pHvalues. Thus, we can hypothesize that the accumulation andremobilization processes of organo-iodine in F-area groundwateroccur as follows: (1) high DOC, high iodide and low pH values inupstream groundwater enhanced the incorporation of iodide into soilorganic matter, and (2) organo-iodine in soil was remobilized bygroundwater flow.

Groundwaters in wells downstream from the barrier walls hadhigh 3H and 129I concentrations and lower pH values (pH=3.8–5.2)(Fig. 3b, d, and f). Also, at seepline well, FPZ-6A, the highest 129I/127Iratio was observed (Fig. 3g). As can be seen from Fig. 5(a) and (b),these characteristics are similar to those observed immediatelydowngradient of the seepage basin. It appears that the groundwaterentering this FPZ-6A can travel the 530 m from the basins and bypassthe remediation system and resurface in the seeplinewith a pH of 3.81(WSRC, 2006). Alternatively, the old water mass that was contami-nated before capping of the seepage basin might have accumulatedin this region due to the installation of the barrier. With regard toFSB-79, a well located ~100 m down gradient from the barrier wall,the pH increased from 4 to 6 in August 2004 when the barrier wallwas installed, and has since been gradually decreasing with time. Asbase was injected in October 2005, the high pH value at this wellwould not have been caused by base injection but more likely byhydrographical changes due to the installation of the wall.

F-area groundwater obtained from wells downstream from thebase injection gate had lower iodine contents and higher pro-portions of iodate on three of four wells (Fig. 3e and 5: See wellsFSB-104, FSB-138 and FPZ-6A in Table 2). Such characteristics ofiodine speciation cannot be explained strictly from thermodynam-ics (Fig. 2) using ambient pH and ORP values (Table 1). However,with the base injection consisted of a pH 10, sodium hydroxide andsodium bicarbonate solution, and iodine would be expected to existprimarily as iodate at such high pH conditions. Thus, it is notsurprising that iodide in groundwater was converted to iodate bycontact with the base solution. Assuming that iodate has a ~1000times higher Kd values than iodide (Fukui et al., 1996), and that theproduction of iodate in groundwater enhanced the removal ofiodine from the system, it would make sense that the concentra-tions of total iodine in groundwater were considerably lower atwells below the base injection gate. At FSB-118D, a well locatedoutside of the mixed waste plume, a high percentage of iodate wasmeasured (Table 2). Of the three wells located outside of the mixedwaste plume, only FSB-118D has experienced a rising pH value afterbase injection (WSRC, 2006). This result supports our contentionthat groundwaters in contact with higher pH water showed higheriodate fractions. Furthermore, there is a positive correlation be-tween the organic matter content of soil and the potential for

iodate uptake (Shimamoto and Takahashi, 2008; Hu et al., 2009;Yamaguchi et al., 2010). If iodate, taken up or transformed by soilorganic matter, would be released to groundwater as a conse-quence of decomposition and/or transformation of the organicmatter, this could result in a higher contribution of iodate ingroundwater in downstream wells.

5. Conclusions

From the distributions of iodine species (iodide, iodate andorgano-iodine) and the geochemical properties of groundwater inthe vicinity of the F-area at SRS, the following transformation andtransport processes of 129I and 127I can be inferred.

1) 129I in groundwater originates primarily from a single source, fromthe seepage basins, and most of the present groundwater waspresumably bound to sediment beneath the basin. Both 127I and 129Iwere supplied as iodide to the groundwater.

2) In the upper reaches of the transect, near the seepage basins,significant amounts of iodine were transferred to organic matterby iodination. Continuous accumulation and remobilization ofiodinated organic matter would support a relatively high mobilityof organo-iodine in groundwater. A “background” groundwatersample collected outside of the mixed waste plume with a higherDOC concentration also had higher organo-iodine content, and thiscould be taken as evidence that NOM is the primary carrier of theorgano-iodine.

3) In groundwater downstream from the base injection gate, rapidchange in pH created an environment that was conductive to thetransformation of iodide to iodate. The transformation to iodatepromoted increased iodine sediment sorption and lower total 127Iand 129I concentrations. However, there was evidence that somegroundwater was bypassing the remediation system and resurfa-cing in the seepline with very high 129I concentrations and low pHlevels, characteristics of the source term.

4) Along the groundwater pathway in the F-area of SRS, 129I-iodidesupplied from the seepage basins was transformed to 129I-iodateand/or 129I-organo-iodine, and subsequently transported todownstream areas and through the soil-groundwater interface.Migration processes of 129I are thought to be highly pH dependent.

Acknowledgments

This work was funded by the Department of Energy's Subsur-face Biogeochemical Research Program within the Office of Science(DE-PS02-07ER07-18). S.Z. was partially supported by Welch GrantBD0046. Laura Bagwell (SRNL) helped with GIS assistance and JayNoonkester (SRNL) helped coordinate the field work. The workwas conducted by the Savannah River National Laboratory underthe U.S. Department of Energy (DE-AC09-96SR18500).

References

Amachi S, Kamagata Y, Kanagawa T, Muramatsu Y. Bacteria mediate methylation ofiodine in marine and terrestrial environments. Appl Environ Microbiol 2001;67:2718–22.

Ashworth DJ, Shaw G, Butler AP, Ciciani L. Soil transport and plant uptake of radio-iodine from near-surface groundwater. J Environ Radioactivity 2003;70:99-114.

Baas-Becking LGM, Kaplan IR, Moore D. Limits of the natural environment in terms ofpH and oxidation-reduction potentials. J Geol 1960;68:243–84.

Bichsel Y, von Gunten U. Determination of iodide and iodate by ion chromatographywith postcolumn reaction and UV/visible detection. Anal Chem 1999;71:34–8.

Buesseler KO, Kaplan DI, Dai M, Pike S. Source-dependent and source independentcontrols on plutonium oxidation state and colloid associations in groundwater.Environ Sci Technol 2009;43:1322–8.

Dai M, Kelley JM, Buesseler KO. Sources and migration of plutonium in groundwater atthe Savannah River Site. Environ Sci Technol 2002;36:3690–8.

Denham M, Vangelas KM. Biogeochemical gradients as a framework for understandingwaste-site evolution. Remediation 2008;18:5-17.

3865S. Otosaka et al. / Science of the Total Environment 409 (2011) 3857–3865

Department of Energy, United States. Yucca mountain site suitability evaluation. DOE/RW-0549. Washington, DC: U.S. Department of Energy Office of CivilianRadioactive Waste Management; 2002.

Department of Energy, United States. Performance assessment for the Idaho NationalLaboratory Remote-Handled low-level waste disposal facility. DOE/ID-11042.Idaho Falls, ID: U.S. Department of Energy Idaho Operations Office; 2011.

Fukui M, Fujikawa Y, Satta N. Factors affecting interaction of radioiodide and iodatespecies with soil. J Environ Radioactivity 1996;31:199–216.

Hou X, Hansen V, Aldahan A, Possnert G, Christian Lind O, Lujaniene G. A review onspeciation of iodine-129 in the environmental and biological samples. Anal ChimActa 2008;632:181–96.

Hu Q, Moran JE, Blackwood V. Geochemical cycling of iodine species in soils. In: PeedyVR, Burrow GN, Watson R, editors. Comprehensive Handbook of Iodine. Oxford:Academic Press; 2009. p. 93-105.

Kaplan DI, Bertsch PM, Adriano DC, Orlandini KA. Actinide association withgroundwater colloids in a coastal plain aquifer. Radiochim Acta 1994;66:181–7.

Kaplan DI, Serne RJ, Parker KE, Kutnyakov IV. Iodide sorption to subsurface sedimentsand illitic minerals. Environ Sci Technol 2000;24:399–405.

Kaplan DI, Roberts KA, Schwehr KA, Lilley MS, Brinkmeyer R, Denham ME, et al.Evaluation of a radioiodine plume increasing in concentration at the SavannahRiver Site. Environ Sci Technol 2011;45:489–95.

Killian TH, Kolb NL, Corbo P, Marine IW. Environmental information document, F-areaseepage basins. DPST-85-704. Aiken, SC: E.I. du Pont de Nenours and Company,Savannah River Laboratory; 1987.

Li H-P, Brinkmeyer R, Jones WL, Zhang S, Xu C, Schwehr KA, et al. Iodide accumulation byaerobic bacteria isolated from subsurface sediments of a 129I-contaminated aquifer atthe Savannah River Site, South Carolina. Appl Environ Microbiol 2011;77:2153–60.

Muramatsu Y, Uchida S, Sriyotha P, Sriyotha K. Some considerations on the sorption anddesorption phenomena of iodide on soil. Water Air Soil Pollut 1990;49:125–38.

Oktay SD, Santschi PH, Moran JE. Sharma P.129I and 127I transport in the MississippiRiver. Environ Sci Technol 2001;35:4470–6.

Orth DA. Plutonium metal from trifluoride. I&EC Proc Des Develop 1963;2:121–7.Readlinger G, Heumann KG. Transformation of iodine in natural and wastewater

systems by fixation on humic substances. Environ Sci Technol 2000;34:3932–6.Ryan JP. Effluent characterization study for the 200-area effluent treatment F-facility.

DSPT-84-511. Aiken, SC: E.I. du Pont de Nemours and Company; 1984.Santschi PH, Schwehr KA. 129I/127I as a new environmental tracer or geochronometer

for biogeochemical or hydrodynamic processes in the hydrosphere and geosphere:the central role of organo-iodine. Sci Total Environ 2004;321:257–71.

Savannah River National Laboratory. Information document for EM-32 applied scienceinitiative “Attenuation-based remedies for metal and radionuclide contaminatedgroundwater” –Feburary 2010. SRNL-TR-2010-00047. Aiken, SC: Savannah RiverNational Laboratory; 2010.

Savannah River Nuclear Solutions, LLC. Savannah river site groundwater strategy andimplementation plan (U). WSRC-RP-2006-4074. Aiken, SC: Savannah River NuclearSolutions, LLC; 2008.

Schwehr KA, Santschi PH, Moran JE, Elmore D. Near-conservative behavior of 129I in theorange county aquifer system. Appl Geochem 2005;20:1461–72.

Schwehr KA, Santschi PH, Kaplan DI, Yeager CM, Brinkmeyer R. Organo-iodineformation in soils and aquifer sediments at ambient concentrations. Environ SciTech 2009;43:7258–64.

Shimamoto YS, Takahashi Y. Superiority of K-edge XANES over LIII-edge XANES in thespeciation for iodine in natural soils. Anal Sci 2008;24:405–9.

Sparks DL. Inorganic soil components. In: Sparks DL, editor. Environmental SoilChemistry. Second Edition. San Diego: Academic Press; 2003. p. 43–74.

Steinberg SM, Schmett GT, Kimble G, Emerson DW, TurnerMF, RudinM. Immobilizationof fission iodine by reaction with insoluble natural organic matter. J Radioanal NuclChem 2008;277:175–83.

United Nations Scientific Committee on the Effects of Atomic Radiation. UNSCEAR-2000Report, vol. 1, Annex A (dose assessment methodologies), V. Globally DispersedRadionuclides, C. Iodine-129, Vienna: United Nations; 2000.

Wachsmuth ED. The iodination of super-reactive and essential tyrosine residues in amicrobial aminopeptidase. Biochem J 1967;102:17c–9c.

Warner JA, Casey WH, Dahlgren RA. Interaction kinetics of I2(aq) with substitutedphenols and humic substances. Environ Sci Technol 2000;34:3180–5.

Westinghouse Savannah River Company LLC. The Savannah River Site's groundwatermonitoring program: second quarter 2001. ESH-EMS-20010585. Aiken, SC:Westinghouse Savannah River Company; 2001.

Westinghouse SavannahRiver CompanyLLC.Annual corrective action report for theF-areahazardous waste management facility, the H-area hazardous waste managementfacility, and the mixed waste management facility (U). WSRC-RP-2006-4011. Aiken,SC: Westinghouse Savannah River Company; 2006.

Xu C, Zhang S, Ho Y-F, Miller EJ, Roberts KA, Li H-P, et al. Is soil natural organic mattersink or source for radioiodine (129I) in a nuclear waste disposal site? GeochimCosmochim Acta 2011. 2011; in revision.

Yamaguchi N, Nakano M, Takamatsu R, Tanida H. Inorganic iodine incorporation intosoil organic matter: evidence from iodine K-edge X-ray absorption near-edgestructure. J Environ Radioactivity 2010;101:451–7.

Yoshida S,Muramatsu Y, Uchida S. Studies on the sorption of I− (iodide) and IO3− (iodate)

onto andesols. Water Air Soil Pollut 1992;63:321–9.Zhang S, Schwehr KA, Ho Y-F, Xu C, Roberts KA, Kaplan DI, et al. A novel approach for the

simultaneous determination of iodine, iodate and organo-iodide for 127I and 129I inenvironmental samples using gas chromatography-mass spectrometry. Environ SciTechnol 2010;44:9042–8.