Journal of Contaminant Hydrology 172 (2015) 24–32

Contents lists available at ScienceDirect

Journal of Contaminant Hydrology journal homepage: www.elsevier.com/locate/jconhyd

Colloid-associated plutonium aged at room temperature: evaluating its transport velocity in saturated coarse-grained granites Jinchuan Xie ⁎, Jianfeng Lin, Yu Wang, Mei Li, Jihong Zhang, Xiaohua Zhou, Yifeng He Northwest Institute of Nuclear Technology, P.O. Box 69–14, Xi’an City, Shanxi Province 710024, PR China

a r t i c l e

i n f o

Article history: Received 26 June 2014 Received in revised form 8 October 2014 Accepted 23 October 2014 Available online 4 November 2014 Keywords: Plutonium Colloid Transport velocity Size exclusion Charge exclusion Granite

a b s t r a c t The fate and transport of colloidal contaminants in natural media are complicated by physicochemical properties of the contaminants and heterogeneous characteristics of the media. Size and charge exclusion are two key microscopic mechanisms dominating macroscopic transport velocities. Faster velocities of colloid-associated actinides than that of 3H2O were consistently indicated in many studies. However, dissociation/dissolution of these sorbed actinides (e.g., Pu and Np), caused by their redox reactions on mineral surfaces, possibly occurred under certain chemical conditions. How this dissolution is related to transport velocities remains unanswered. In this study, aging of the colloid-associated Pu (pseudo-colloid) at room temperature and transport through the saturated coarse-grained granites were performed to study whether Pu could exhibit slower velocity than that of 3H2O (UPu/UT b1). The results show that oxidative dissolution of Pu(IV) associated with the surfaces of colloidal granite particles took place during the aging period. The relative velocity of UPu/UT declined from 1.06 (unaged) to 0.745 (135 d) over time. Size exclusion limited to the uncharged nano-sized particles could not explain such observed UPu/UT b1. Therefore, the decline in UPu/UT was ascribed to the presence of electrostatic attraction between the negatively charged wall of granite pore channels and the Pu(V)O2+, as evidenced by increasing Pu(V)O2+ concentrations in the suspensions aged in sealed vessels. As a result of this attraction, Pu(V)O2+ was excluded from the domain closer to the centerline of pore channels. This reveals that charge exclusion played a more important role in dominating UPu than the size exclusion under the specific conditions, where oxidative dissolution of colloid-associated Pu(IV) was observed in the aged suspensions. © 2014 Elsevier B.V. All rights reserved.

1. Introduction The transport velocity is an important quantity describing how fast contaminants transport through subsurface media. Low-solubility metal and organic contaminants often existed as colloidal species in natural aquatic environments (Plathe et al., 2013; Zänker and Hennig, 2014). In field experiments of the aquifers, model colloids were observed to travel much faster

⁎ Corresponding author. Tel.: +86 29 84767789; fax: +86 29 83366333. E-mail address: [email protected] (J. Xie).

http://dx.doi.org/10.1016/j.jconhyd.2014.10.011 0169-7722/© 2014 Elsevier B.V. All rights reserved.

than solute tracers, e.g., bacteria travelled 1.25–2.5 times faster than Br− (Harvey et al., 1989; McKay et al., 1993a,b) and 1–2 times faster than the groundwater (Sinton et al., 2000). Also, faster velocities of colloid-associated actinides such as Pu, Am, and Np than that of 3H2O were widely reported in lab-scale experiments (Artinger et al., 1998, 2000; Delos et al., 2008; Schäfer et al., 2003, 2004). To accurately assess environmental risk of the contaminants, especially the colloidal actinides with long half-life and high radiotoxicity, insights into their transport velocity are needed. Transport velocities of colloidal contaminants are impacted by their size and surface charge. Nano-sized contaminants can

J. Xie et al. / Journal of Contaminant Hydrology 172 (2015) 24–32

travel shorter pathways in pore channels than the bulk water does and thus exhibit faster velocities (i.e., an effect of size exclusion) (Keller and Auset, 2007; Sinton et al., 2010). It is well known that the natural media such as granites and soils are generally characterized by negatively charged surfaces under environmentally relevant pH conditions (Drelich and Wang, 2011). For the contaminants having the negatively charged surfaces, they are therefore concentrated around the channel centerline where the local water velocity is at its maximum, due to the electrostatic repulsion exerted by the channel wall. In this case, their transport velocities can be enhanced (i.e., an effect of charge exclusion). On the contrary, electrostatic attraction between the positively charged contaminants and the wall may decrease their velocities because this attractive interaction makes the contaminants closer to the pore wall where the water velocity is at its minimum. Size exclusion was usually invoked to explain the velocity-enhanced phenomenon (Keller et al., 2004; Small, 1974). However, this mechanism, which does not involve charge effect on the transport of charged contaminants, is limited to the uncharged nano-sized particles and thus cannot account for the decrease in transport velocity. Accordingly, it seems likely that electrostatic interaction (charge exclusion) may sometimes play an important role, to which attention is paid in this study. Here, the charge exclusion describes the case that the positively charged contaminants are excluded from the domain closer to the pore centerline. This contrasts with the other case that the negatively charged contaminants are excluded from the domain closer to the pore wall. In our recent study, an expression for the relative transport velocity of colloid-associated Pu and 3H2O (UPu/UT) was established based on a newly developed model of electrostatic interactions coupled with a parabolic water velocity profile (Xie et al., 2014a). The quantitative relationship between UPu and UT was thus determined as UPu/UT =1.1 to 1.5, depending on the water flow rates and ionic strengths. A further study on the effects of aging time on UPu/UT has been now performed. This study was motivated by the following considerations. The Pu(IV) readily form relatively stable colloid-associated species, due to its high affinity for the mineral surfaces (Xie et al., 2012, 2013a). However, dissociation/dissolution of Pu(IV) from the surfaces was expected to occur with aged colloid-associated Pu. Typically, the associated Pu(IV) was partially dissolved in the presence of oxygen and then oxidized to more soluble Pu(V)aq. Whether, and to what extent, this dissolution may affect UPu/UT is not yet clear. In contrast to commonly observed faster velocities of colloidal actinides than that of 3H2O, emphasis is now given to a possibility of UPu/UT b 1 in specific conditions. Both the present and previous studies are considered as a complete report on relative movement of Pu and the bulk water in the pore channels. The mobility of colloid-associated Pu, as affected by its aging, is not studied here. The relationship between Pu species and its mobility (the percent recovery of Pu) was studied in detail in Xie et al., 2013a. The colloidal granite particles with the size smaller than 1 μm were fractioned from the granites collected at Lop Nor in northwestern China. The colloid-associated Pu was formed by sorption of Pu(IV) on these colloidal particles and then aged at room temperature up to 135 d. The coarse-grained granites packed into the columns were used as the stationary phase in transport experiments.

25

2. Material and methods 2.1. Coarse-grained granites The granite drill cores returned from Lop Nor were crushed and sieved with stainless steel sieves (0.7 to 1.0 mm in pore diameter). Their elemental and mineralogical compositions were analyzed by X-ray fluorescence spectroscopy (Axios) and X-ray diffraction (D/MAX-2500), respectively. The fine particles attached onto the surfaces of the coarse-grained granites having 0.7 to 1.0 mm in size were completely removed with pure water (18.2 MΩ, Millipore) in order to avoid introducing unknown amounts of mineral colloidal particles in transport experiments. After gradually oven-drying at 50 °C, these granite grains whose specific surface areas (BET) were determined by ASAP 2020 (Micromeritics) were used as the stationary media packed into the columns. 2.2. Colloidal granite particles The granite particles of b1 μm in size were fractionated from crushed granites of b30 μm in size, as the following: The crushed granites of ~80 g were added to a glass beaker (4 L pure water), and then the colloidal suspension with a calculated Stoke's diameter of b1 μm was siphoned from the upper suspension into a polypropylene vessel. The prepared suspension was stored in a refrigerator (4 °C) and used as the colloidal source materials. To determine the mass concentration of the parent suspension, six 25 mL aliquots of the suspension were transferred separately to concave Teflon membranes and then dried at 50 °C by infrared light. The mass concentration was determined as 195.2 ± 5.9 mg/L. The suspension was diluted to obtain the desired particle concentrations. Electrokinetic potential (ζ) of the colloidal particles, as a function of the pH from 1.8 to 11.2, was measured using Nano ZS (Malvern). According to Smoluchowski's equation, electrophoretic mobility was converted to the ζ-potential in this study. This experiment was performed to examine surface charge properties of the granites under natural alkalin conditions. The specific surface areas of the dried colloidal particles were also determined by the ASAP 2020. 2.3. Colloid-associated Pu The prepared suspension was diluted with tritiated water to 35 mg/L (the concentration of colloidal granite particles) in a Teflon bottle. A small amount of 239Pu stock solution (~ 1.6 μg/g in 0.5 mol/L HNO3) was drop by drop added to the suspension stirred continually with a magnetic stirrer, resulting in final Pu concentration of about 10− 9 mol/L. The atom ratios in the 239Pu stock solution were 240Pu/239Pu = 0.0346, 241Pu/239Pu = 0.000355 and 242 Pu/239Pu = 0.0000323. After 10 min, the pH was adjusted to 8.0 by 1.0 mol/L NaOH. The pH was measured by Orion 720A meter after calibrated by pH 4.00, 6.86, and 9.18 standard buffer solutions. The Na+ concentration caused by the NaOH solution was ~0.002 mol/L. This Pu suspension consisted of colloid-associated Pu and dissolved Pu (i.e., aqueous Pu). Four 65 mL aliquots of the suspension were pipetted separately into 0.l L Teflon bottles. The suspensions in the first three bottles were sealed with lutes and then aged at room

26

J. Xie et al. / Journal of Contaminant Hydrology 172 (2015) 24–32

0.1 mol/L sodium acetate/acetic acid buffer (pH = 4.5), or 0.6 mol/L HClO4 (pH = 0.5). These two tubes with 2 ml of 0.5 mol/L TTA in xylene were shaken for 5 minutes. A 1-mL aliquot of the aqueous phase at the bottom of either tube was drawn into a syringe and then transferred to a 5 ml polythene test tube (for subsequent removal of U and matrix elements, described in Section 2.6). The buffered aqueous phase contained only the Pu(V)aq, whereas the acidified aqueous phase contained Pu(V)aq and Pu(VI)aq. The Pu sorbed on the colloidal particle surfaces (i.e. colloidassociated species) were leached with 0.03 mol/L HCIO4 (pH 1.5). The leachate was filtered through 10 kD membrane, and then the valence states of Pu in the filtrate were examined via the TTA solvent extraction. The Pu that could not be leached out (i.e., the residue fraction) was assumed to be Pu(IV). On the basis of a mass balance, the percentages of Pu(IV), Pu(V), and Pu(VI) in the suspension samples were determined. These experiments were performed in duplicate.

temperature (15, 65, and 135 d). The last one, which was not aged, was immediately used for species analysis and transport experiments. Aliquots of the suspensions either unaged or aged were filtered through 10 kD membranes, so that the percentages of colloid-associated Pu were determined (see Section 2.5). The colloid-associated species were the fraction retained on the membranes. These experiments were performed in duplicate. Plutonium valence states were determined by a solvent extraction method (see Section 2.5). The remainder of each suspension was immediately used for the transport experiments. 2.4. Transport experiments The coarse-grained granites were loaded separately into four polypropylene columns with the dimensions of 2.0 cm in diameter and 20.0 cm in height. Aqueous solution with pH 8.0 and 0.002 mol/L Na+ was fed into the columns from the bottom by a peristaltic pump having a fixed inflow rate of 2.2 mL/min, until the saturated water contents monitored by weighing the columns were achieved. Subsequently, 40 mL of the Pu suspension (~1.3 pore volumes), either unaged or aged, was fed into the columns to displace the aqueous solution existing in the granite systems, followed by re-feeding the aqueous solution of three pore volumes. The column effluent was continually collected by a fraction collector at the column outlet. Both 239Pu concentration and tritium counting were determined in each effluent sample. Some transport parameters were the pore volumes (Vp =29.8 cm3), the effective porosity (ε =0.499), and the saturated water content (θ =0.460 cm3/cm3). The chemistry of the Pu suspensions fed into the columns was shown in Table 1. In addition, the percentages of colloid-associated Pu and Pu valence states in selected effluent samples were determined.

2.6. Plutonium and 3H2O concentration measurement Plutonium-239 concentrations of all samples were measured by inductively coupled plasma mass spectrometry (ICPMS, Element) using the isotope dilution method. Plutonium242 was used as a spike, and the atom ratio of 239Pu/242Pu in the spike was 0.00115. The 238U, which interfered with the detection of mass 239, were removed with Dowex1 × 2 resins through the acid washing. The purification procedure was described in detail in our previous studies (Xie et al., 2013b, c). The atom ratios of 239Pu/242Pu reported by the ICP-MS were calibrated both with the 239Pu/242Pu (0.00115) in the spike and with the intensity of m/z 238 in the samples where the interference of protonated uranium (UH+/H+) at m/z of 239 had a ratio of 4 × 10−5. The 3H2O activities in the column effluent were measured by liquid scintillation counting (Wallac 1414). A portion of each sample (1.0 mL) was mixed with scintillation cocktail (10 mL, PerkinElmer Optiphase HisSafe 3) and counted for 35 min.

2.5. Plutonium valence state examination Plutonium valence state distribution both in solution and on the colloidal surfaces was determined by a combined ultrafiltration and TTA (thenoyltrifluoroacetone, Sigma-Aldrich Co.) solvent extraction method (Fig. 1), which was initially developed by Bertrand and Choppin (1982) and Keeney-Kennicutt and Morse (1985). A 1-mL aliquot of the Pu suspension was withdrawn to determine the total concentration of Pu, which was composed of colloid-associated and dissolved Pu. Then, either concentration was obtained after a 3 mL aliquot of the suspension was filtered through the 10 kD membrane (~1.5 nm, Amicon Ultra-4, Millipore). A 1-mL aliquot of the 10 kD filtrate was used to determine the total concentration of dissolved Pu consisting of Pu(IV)aq, Pu(V)aq, and Pu(VI)aq. Two 1-mL aliquots of the filtrate were then added to 5 ml polythene test tubes having either 1 mL of

2.7. A method for determining transport velocity relationship between charged particles and 3H2O (UPu/UT) Relative velocities of colloidal contaminants and solute tracers such as 3H2O were more often determined by two single data in their breakthrough curves (BTCs) (Delos et al., 2008; Harter et al., 2000; Kurosawa et al., 2006; Levy et al., 2007; Sinton et al., 2010). Yet, it was much difficult to judge which two data were available for reflecting their relative movement. Hence, large deviations from the true velocities may result. A challenge is to develop a model involving size and charge exclusion mechanisms, and then derive a relative velocity

Table 1 The chemistry of the Pu suspensions fed into the columns. Colloidal granite particle (mg/L) 35

Pu (mol/L)a ~10

−9

3

H2O (Bq/g)

620.2

NO3− (mol/L)b ~10

−7

Na+ (mol/L)b

Electrolytic conductivity (mS/cm)c

pHc

~0.002

0.35

8.0

The concentration of total Pu, i.e., Pu both in solution and on the colloidal surfaces. For the aged Pu suspensions, this concentration (10−9 mol/L) could be apparently decreased (256.1 ± 16.1 (unaged), 96.5 ± 17.2 (15 d), 31.4 ± 2.5 (65 d), and 30.2 ± 5.0 pg/g (135 d)) because of their instability. These final concentrations of Pu were not an essential parameter used to determine UPu/UT. b The NO3− and Na+ concentrations were produced while the Pu suspension was prepared and its pH was adjusted. c The electrolytic conductivity and pH of the freshly prepared Pu suspension. a

J. Xie et al. / Journal of Contaminant Hydrology 172 (2015) 24–32

27

Fig. 1. The analytic flow diagram for determining the distribution of Pu oxidation state (IV, V, and VI) in the suspension containing colloid-associated Pu (pseudocolloid) and dissolved Pu.

expression in which all breakthrough data instead of only two single data should be utilized. Evidently, charge exclusion is related to the distance of charged contaminants from pore channel wall/centerline. For the colloidal contaminants, their shorter transport pathways correspond to the domain closer to the channel centerline. Therefore, size exclusion is allowed to result from their shorter distance from the centerline. Such is an essential way by which the velocity relationship can be abstracted from the complicated mechanisms. We recently developed a model of electrostatic interaction coupled with a parabolic water velocity profile (Xie et al., 2014a). The expressions for relative velocities of the negatively charged colloid-associated Pu, uncharged 3H2O, and positive ion Sr2+ (UPu/UT, UPu/U2+ Sr ) were established according to this model. As stated above, one distribution of colloidal contaminants in the cross section of a channel is allowed to be contributed by electrostatic interaction as well as by size exclusion. The same expressions can be established consequently if size exclusion is now involved in the model. Here, we report the Eqs. 1 and 2 to determine the relative velocity of Pu and 3H2O. For more detailed derivations, the reader can refer to our study (Xie et al., 2014a) (see its Supplementary Information).

U Pu =U T ¼ 1 þ

θex‐Pu ; V PuðCp¼50%Þ ≤V TðCp¼50%Þ θ

ð1Þ

U Pu =U T ¼ 1−

θex‐Pu ; V PuðCp¼50%Þ NV TðCp¼50%Þ θ

ð2Þ

where θex-Pu (cm3/cm3) is the exclusion water content of Pu, θ (cm3/cm3) is the experimentally measured water content in the column loaded by coarse-grained granites, Cp is the collection percentages of Pu and 3H2O, VPu(Cp=50%) (cm3) and VT(Cp=50%) (cm3) are their effluent volumes at Cp = 50%. The velocity relationship of UPu/UT is determined by Eq. (1) when Pu moves faster than 3H2O. In this case, the Pu concentration fronts appear ahead of 3H2O, i.e., VPu(Cp= 50%) b VT(Cp=50%). Otherwise, the UPu/UT is determined by Eq. 2. The variables in Eqs. 1 and 2 are determined as θex‐Pu ¼

VS π  a2  H

ð3Þ

where a (cm) and H (cm) are the radius of the column and the height of the loaded granites. V S ¼ V T ðCp¼50%Þ −V Pu ðCp¼50%Þ ; V PuðCp¼50%Þ ≤V TðCp¼50%Þ

ð4Þ

V S ¼ V Pu ðCp¼50%Þ −V T ðCp¼50%Þ ; V PuðCp¼50%Þ NV TðCp¼50%Þ

ð5Þ

28

J. Xie et al. / Journal of Contaminant Hydrology 172 (2015) 24–32

Table 2 Specific surface areas and pore volumes of the granites.

Specific surface area As, (m2/g) Pore volumes W, (cm3/g)

Colloidal particle (1 nm to 1 μm)

Grain (0.7 to 1 mm)

42.7 0.122

2.09 0.00221

where VS (cm3) is the separation volumes between Pu and 3 H2O at Cp = 50%, as experimentally determined via utilizing their effluent volumes (V) shown in the BTCs. Cp ¼

Cumulative effluent mass of the contaminant at one effluent volume  100% Total effluent mass of the contaminant

and effluent were almost equal in electrolytic conductivity (0.35 mS/cm) and pH (8.0). Thus, the ζ-potential of the column effluent as a function of pH was consistent with the result in our recent study (Xie et al., 2014b). Such further indicates that the trace level Pu(IV) (b10−9 mol/L) sorbed on the colloidal granite particles had no effects on the ζ-potential of the particles. On the basis of N2 adsorption isotherms, both the specific surface areas (As) and the pore volumes (W) of the granite grains and particles were calculated in Table 2. The As of the particles (42.7 m2/g) was much larger than that of the grains (2.09 m2/g), leading to a larger amount of Pu sorbed on the particle surfaces.

ð6Þ where the contaminant refers to the Pu and 3H2O. Plutonium exclusion volumes Vex-Pu, i.e., the separation volumes per unit mass of the media, are determined as V ex‐Pu

V ¼ S m

ð7Þ

where m is the mass of coarse-grained granites packed into the column. It is noteworthy that the initial concentrations of Pu and 3 H2O (C0) are not needed when we determine their collection percentages (Cp) by Eq. 6. This is different from the case of the percent recovery. Therefore, both the cumulative effluent mass and the total effluent mass of Pu and 3H2O in Eq. 6 may be also expressed by their effluent intensities in cpm/cps, or by the others. In this study, pg/g for Pu and cpm/g for 3H2O in their BTCs were presented with respect to the corresponding effluent volumes. 3. Results 3.1. Surface characterization of the granites The elemental composition of the granites was 65.2% SiO2, 15.6% Al2O3, and 3.7% FeO (Fe2O3 0.6%); the major mineralogical composition was 61% quartz, 13% anorthose, and 7% chlorite. The isoelectric point (pHIEP) of the colloidal granite particles was determined as pHIEP = 2.75. Detailed results of these two compositions and the ζ-potential functioned by pH were given in our recent studies (Xie et al., 2014b, c). This pHIEP indicates that the granites had the negatively charged surfaces under the environmentally relevant pH 8.0 conditions in this study. The relatively short travel time of the influent suspensions (14.3 min, as calculated based on the inflow rate and the column dimensions) through the granite systems could not lead to detectable dissolution of the granites. This was supported by the observation that both the column influent

3.2. Responses of UPu/UT to the changes in aging time of Pu suspensions Plutonium species and valence state fraction in the suspensions unaged and aged for 15 d at room temperature are reported in Table 3. Plutonium(IV), as the dominating valence state composed of dissolved Pu(IV) and the Pu(IV) complexed/associated with ≡ X-OH sites (X = Al, Fe, and Mn) on the particle surfaces, accounted for about 90% in the two suspension samples. Because of strong affinity of Pu(IV)aq for the mineral surfaces, percentage of dissolved Pu was reduced from 60.3% to 3.33% at 15 d. In these suspensions, Pu(VI)aq was considered absent in terms of their negative values caused by the uncertainty of the TTA solvent extraction method. For the suspensions aged for 65 and 135 d, only Pu(V)aq concentrations were determined to examine whether the Pu(IV) was oxidized during the aging periods. The BTCs for Pu and 3H2O through the saturated coarsegrained granites are presented in Fig. 2. It is evident that concentration fronts of the unaged Pu appeared ahead of 3H2O, and that Pu exhibited faster transport velocity (UPu/UT N 1 determined by Eq. 1). In contrast to the traditional method based on the two single data for Pu and 3H2O breakthrough, our model using all breakthrough data was highly sensitive to the relative location of breakthrough curves for Pu and 3H2O. Small differences both in their breakthrough locations and thus in UPu/UT could be easily discerned. This observed transport phenomenon could have resulted from two microscopic mechanisms: size and charge exclusion. On the one hand, the colloidal granite particles (carriers of Pu), which had lower dispersion and thus limited extent of pathway diversity in the granite systems (i.e., size exclusion), travelled shorter pathways in pore channels and hence transported faster than 3H2O molecules. On the other hand, electrostatic repulsion between the two negatively charged surfaces of stationary granite grains and mobile particles made the colloid-associated Pu closer to the channel centerline where

Table 3 Percentages of colloid-associated and dissolved species, and Pu valence state in the suspensions. Aging time (d)

Colloid-associated Pu (%)

Dissolved Pu (%)

0 (unaged) 15

39.7 ± 1.2 96.7 ± 0.1

60.3 ± 1.2 3.33 ± 0.1

a

The valence state distribution involved the Pu, both in solution and on the colloidal surfaces.

Pu valence state (%) a IV

V

VI

90.1 ± 9.4 89.4 ± 0.57

18.4 ± 2.7 14.2 ± 0.12

−8.50 ± 12.0 −3.60 ± 0.45

J. Xie et al. / Journal of Contaminant Hydrology 172 (2015) 24–32

29

20 6000

3 100

6000

Pu

Pu

H2O

3 H2O

15

t = 15 d

4000

4000

3

t = 0 (unaged)

Effluent intensity of H2O, CT (cpm/g)

Effluent concentrtion of Pu, CPu (pg/g)

150

10 50

2000

2000 5

0

0

0

30

60

90

0

120

0

0

30

60

90

120

6000

6000 15

Pu

Pu

3 H2O

3 H2O 10

Effluent intensity of H2O, CT (cpm/g)

t = 65 d

t = 135 d

4000 10

4000

3

Effluent concentrtion of Pu, CPu (pg/g)

15

5

2000

2000 5

0

0

0

30

60

90

120

0

0

0

30

Effuent volume,V (mL)

60

90

120

Effuent volume,V (mL)

Fig. 2. Breakthrough curves for Pu and 3H2O in the saturated coarse-grained granites. The dash lines indicate the effluent volumes where the collection percentages (Cp) of Pu (black) and 3H2O (blue) were overall equal to 50%. The difference in their effluent volumes was the separation volumes (VS). The collection percentage used here did not involve the initial concentrations of Pu and 3H2O and hence was different from the percent recovery. Prior to transport experiments, the Pu suspensions (pH 8.0) containing colloid-associated Pu and dissolved Pu were sealed in Teflon vessels and then aged at room temperature up to t = 135 d.

responsible for such UPu/UT b1. This could be evidenced by the dependence of UPu/UT on the increased concentrations of dissolved Pu(V) in the suspensions, as shown in Fig. 3. The increase in Pu(V)aq concentrations with the aging time, which principally arose from oxidation of Pu(IV) by oxygen in the sealed vessels, indicates oxidative dissolution of the Pu(IV) -10

9.0x10

1.1 UPu/UT

-10

6.0x10

0.9 -10

3.0x10

Pu(V)aq (mol/L)

Pu(V)aq 1.0

UPu/UT

the local water flow had the maximum velocity. Therefore, the charge exclusion made a further contribution to the observed UPu/UT N 1 of the unaged sample. Interestingly, the bilateral symmetry of BTCs for Pu was broken with increasing aging time (15, 65, and 135 d). In such cases, the time for Pu arriving at its collection percentage (Cp = 50%) was later than that for 3H2O arriving at its Cp =50%, i.e., VPu(Cp=50%) N VT(Cp=50%). The slower velocities of Pu than those of 3H2O (UPu/UT b1 determined by Eq. 2) are shown in Fig. 3, e.g., UPu/UT was declined to 0.745 at 135 d. The unexpected slower velocity of Pu was not observed in our previous studies, in which the Pu suspensions were used for the transport experiments after about 5 d of contact time between Pu and the soil colloids (Xie et al., 2014a). The increases in size of aged mineral colloids, as caused by diminished repulsive energies, were commonly observed (Czigány et al., 2005; Missana et al., 2003). Regarding decreased extent of pathway diversity of the colloidal particles with increasing particle size, the UPu should become faster with the aging time. On the contrary, UPu/UT was obviously decreased. This reveals that the change in particle size (size exclusion) did not play a decisive role in affecting UPu/UT. The electrostatic attraction between cationic PuO2+ and negatively charged surfaces of the grains prevailed over the size exclusion acting on the colloid-associated Pu and therefore was

0.8

0.0 0

20

40

60

80

100

120

140

Aging time, t (d) Fig. 3. Responses of both the relative transport velocity of Pu and 3H2O in the saturated coarse-grained granites and the Pu(V)aq concentration in the suspensions to different aging time. Such Pu(V)aq did not include the Pu(V) weakly sorbed on the surfaces of colloidal granite particles. Error bars denote ± one standard deviation.

30

J. Xie et al. / Journal of Contaminant Hydrology 172 (2015) 24–32

Table 4 Effluent samples from the coarse-grained granite columns: percentages of colloid-associated and dissolved species and valence state distribution of Pu. Aging time (d)

65 135 Average

a

Sample number b

Colloid-associated Pu (%)

Dissolved Pu (%)

11 12 11 12

97.9 98.8 95.4 94.6 96.7 ± 2.00

2.09 1.20 4.57 5.40 3.32 ± 1.99

Pu valence state (%) c IV

V

VI

90.2 84.6 87.2 84.6 86.7 ± 2.67

13.1 10.9 12.5 10.6 11.8 ± 1.21

−3.27 4.50 0.300 4.80 1.58 ± 3.83

a

Prior to transport experiments, the Pu suspensions sealed in Teflon vessels were aged at room temperature. Effluent samples were successively collected, and then the sample 11 and 12, having relative high Pu concentrations, were used to examine Pu species and valence states. c The valence state distribution involved the Pu, both in solution and on the colloidal surfaces. b

sorbed on the colloidal particle surfaces. As a result, these Pu(V)aq, having attraction interaction with the surfaces of stationary granite grains, were distributed in the slower transport domain near the channel walls. We examined both species and valence states of Pu in some effluent samples, and the results are reported in Table 4. The relatively small percentages of Pu(V) demonstrate that the dissolved Pu(V) was weakly sorbed on the grain surfaces via the electrostatic attraction. The dissolved Pu, accounting for only ~3.3% in the effluent samples, further indicates that the effluent Pu existed mainly as the species sorbed on the particle surfaces, instead of as the dissolved species. For the freshly prepared suspension (the colloidal granite particle of 35 mg/L), the absorbance measured by an ultraviolet–visible spectrophotometer (HP, VECTA-XM) at 400 nm wavelength was 0.0496 ± 0.002. However, the absorbance of the column effluent was decreased to 0.0132 ± 0.001. Their difference caused by retention of the relatively large particles in the granite systems indicates that colloid filtration did take place. The strong affinity of Pu(IV) with mineral surfaces resulted in the increase in colloidal fraction of Pu from 39.7 ± 1.2% (unaged) to 96.7 ± 0.1 (15 d), as also observed in our previous study (Xie et al., 2014b). The increasing concentration of Pu(V)aq with the aging time was ascribed to oxidation of Pu(IV) by the oxygen in the sealed vessels and by Fe(III) (0.6% Fe2O3) in the granites. This indicates that the aged Pu suspensions were characterized by chemical instability. Also, the great decrease in final concentrations of Pu in the aged suspensions reveals that aggregation of colloidal granite particles occurred. The aggregates settling out of suspension were mainly responsible for the change in the final concentrations of Pu from 256.1 ± 16.1 (unaged) to 96.5 ± 17.2 (15 d), 31.4 ± 2.5 (65 d), and 30.2 ± 5.0 pg/g (135 d). In this case, the mobility of Pu (percent recovery of Pu) could not be effectively evaluated because of the too wide span of these concentrations. 4. Discussion Separation volumes between the unaged Pu and 3H2O were determined by Eq. 4, in which case VPu(Cp=50%) was smaller than VT(Cp=50%). Otherwise, separation volumes for the aged Pu were determined by Eq. 5. Such volumes illustrated with dash lines in Fig. 2 reveal the exclusion domain that was totally inaccessible to the Pu by either size or charge exclusion. However, uncharged 3H2O could diffuse into the entire domain in pore channels, and therefore its transport velocity could represent the bulk water velocity. Exclusion volumes and

exclusion water content were determined by Eqs. 7 and 3, respectively. These parameters reported in Table 5 exhibited nonmonotonic dependences on the aging time, e.g., the decrease in VS from 0 to 15 d, and then the increase from 15 to 135 d, indicating the change in mechanisms dominating UPu. According to results in Table 5, a conceptual diagram for colloid-associated Pu and PuO2+ transport in the negatively charged granite media is provided in Fig. 4, which is helpful on understanding the complicated transport mechanisms. The colloid-associated Pu was concentrated around the centerline (r1 domain) with the highest local water velocity, caused by the size exclusion and the electrostatic repulsion exerted by the channel wall. The UPu/UT N1 was thus observed. In contrast, PuO2+ was distributed in the domain (R – r2) near the channel wall with the slowest water velocity, due to the electrostatic attraction. This led to UPu/UT b1. The Pu tailings in its BTCs (Fig. 2) were attributed to desorption (detachment) of the PuO2+ attached onto the channel walls. Then these desorbed PuO2+ slowly moved along the walls, which further supported the observed UPu/UT b 1. The tailings in BTCs for Pu became pronounced with the aging time, as consistent with the increasing PuO2+ concentration in the aged suspensions (Fig. 3). It is noted that the colloid-associated Pu was excluded from the (R – r1) domain. On the contrary, the PuO2+ was excluded from the r2 domain. Thus, these two species of Pu were distributed with the opposite direction (r1, R – r2). This accounted for the nonmonotonic dependence of VS, Vex–Pu, and θex-Pu on the aging time (see Table 5). 5. Conclusions The colloid-associated Pu was aged in the sealed vessels at room temperature, and then their transport experiments through the coarse-grained granites were performed in order to study whether Pu transported slower than the bulk water (3H2O) in specific conditions. The results show that Pu(V)aq concentration was greatly increased from 3.2 × 10−11

Table 5 Separation volumes (VS), exclusion volumes (Vex–Pu), exclusion water content (θex-Pu), and relative transport velocity of Pu and 3H2O (UPu/UT). Aging time (d)

VS (cm3)

Vex–Pu (cm3/g)

θex-Pu (cm3/cm3)

UPu/UT

0 (unaged) 15 65 135

1.66 1.37 3.68 6.96

0.0206 0.0167 0.0449 0.0848

0.0279 0.0229 0.0617 0.117

1.06 0.950 0.864 0.745

J. Xie et al. / Journal of Contaminant Hydrology 172 (2015) 24–32

31

Fig. 4. Microscopic distribution of colloid-associated Pu and Pu(V)O2+, and their transport in pore channels. The local water velocities in the capillary channel were characterized by a nearly parabolic distribution across the section. The pore channels include both the pore spaces formed adjacent to grain–grain junctions and the intra-grain pores larger than the colloid size.

(unaged) to 8.2 × 10−10 mol/L (135 d) with the aging time. This indicates oxidative dissolution of the Pu (IV) associated with the surfaces of colloidal granite particles. The Pu(V)O2+ was thus excluded from the domain near the channel centerline, due to the electrostatic attractive interaction with the negatively charged channel walls. As a result, a continuous decline in relative transport velocities of UPu/UT from 1.06 to 0.745 was observed in this study. Charge exclusion rather than size exclusion was responsible for such observed UPu/UT b 1, because size exclusion was exclusively suitable for evaluating relative movement of the uncharged particles and solute tracers. Acknowledgments We thank Quanlin Shi, Zhiming Li and Haijun Dang for their valuable contributions to this research. This research is supported by the National Natural Science Foundation of China (No. 21477097). References: Artinger, R., Kienzler, B., Schüßler, W., Kim, J.I., 1998. Effects of humic substances on the Am-241 migration in a sandy aquifer: column experiments with Gorleben groundwater/sediment systems. J. Contam. Hydrol. 35, 261–275. Artinger, R., Marquardt, C.M., Kim, J.I., Seibert, A., Trautmann, N., Kratz, J.V., 2000. Humic colloid-borne Np migration: influence of the oxidation state. Radiochim. Acta 88, 609–612. Bertrand, P.A., Choppin, G.R., 1982. Separation of actinides in different oxidation states by solvent extraction. Radiochim. Acta 31, 135–137. Czigány, S., Flury, M., Harsh, J.B., 2005. Colloid stability in vadose zone Hanford sediments. Environ. Sci. Technol. 36, 1506–1512. Delos, A., Walther, C., Schäfer, T., Büchner, S., 2008. Size dispersion and colloid mediated radionuclide transport in a synthetic porous media. J. Colloid Interface Sci. 324, 212–215. Drelich, J., Wang, Y.U., 2011. Charge heterogeneity of surfaces: mapping and effects on surface forces. Adv. Colloid Interf. 165, 91–101. Harter, T., Wagner, S., Atwill, E.R., 2000. Colloid transport and filtration of Cryptosporidium parvum in sandy soils and aquifer sediments. Environ. Sci. Technol. 34, 62–70.

Harvey, R.W., George, L.H., Smith, R.L., LeBlanc, D.R., 1989. Transport of microspheres and indigenous bacteria through a sandy aquifer: results of natural- and forced-gradient tracer experiments. Environ. Sci. Technol. 23, 51–56. Keeney-Kennicutt, W.L., Morse, J.W., 1985. The redox chemistry of Pu(V)O2+ interaction with common mineral surfaces in dilute solutions and seawater. Geochim. Cosmochim. Acta 49, 2577–2588. Keller, A.A., Auset, M., 2007. A review of visualization techniques of biocolloid transport processes and the pore scale under saturated and unsaturated conditions. Adv. Water Resour. 30, 1392–1407. Keller, A.A., Sirivithayapakorn, S., Chrysikopoulos, C.V., 2004. Early breakthrough of colloids and bacteriophage MS2 in a water-saturated sand column. Water Resour. Res. 40. http://dx.doi.org/10.1029/2003WR002676 W08304. Kurosawa, S., James, S.C., Yui, M., Ibaraki, M., 2006. Model analysis of the colloid and radionuclide retardation experiment at the Grimsel Test Site. J. Colloid Interface Sci. 298, 467–475. Levy, J., Sun, K., Findlay, R.H., Farruggia, F.T., Porter, J., Mumy, K.L., Tomaras, J., Tomaras, A., 2007. Transport of Escherichia coli bacteria through laboratory columns of glacial-outwash sediments: estimating model parameter values based on sediment characteristics. J. Contam. Hydrol. 89, 71–106. McKay, L.D., Cherry, J.A., Bales, R.C., Yahya, M.T., Gerba, C.P., 1993a. A field example of bacteriophage as tracers of fracture flow. Environ. Sci. Technol. 27, 1075–1079. Mckay, L.D., Gillham, R.W., Cherry, J.A., 1993b. Field experiments in a fractured clay till 2. Solute and colloid transport. Water Resour. Res. 29, 3879–3890. Missana, T., Alonso, Ú., Turrero, M.J., 2003. Generation and stability of bentonite colloids at the bentonite/granite interface of a deep geological radioactive waste repository. J. Contam. Hydrol. 61, 17–31. Plathe, K.L., von der Kammer, F., Hassellöv, M., Moore, J.N., Murayama, M., Hofmann Jr., T., M.F.H., 2013. The role of nanominerals and mineral nanoparticles in the transport of toxic trace metals: field-flow fractionation and analytical TEM analyses after nanoparticle isolation and density separation. Geochim. Cosmochim. Acta 102, 213–225. Schäfer, T., Artinger, R., Dardenne, K., Bauer, A., Schuessler, W., Kim, J.I., 2003. Colloid-borne americium migration in Gorleben groundwater: significance of iron secondary phase transformation. Environ. Sci. Technol. 37, 1528–1534. Schäfer, T., Geckeis, H., Bouby, M., Fanghänel, T., 2004. U, Th, Eu and colloid mobility in a granite fracture under near-natural flow conditions. Radiochim. Acta 92, 731–737. Sinton, L.W., Noonan, M.J., Finlay, R.K., Pang, L., Close, M.E., 2000. Transport and attenuation of bacteria and bacteriophages in an alluvial gravel aquifer. N. Z. J. Mar. Freshw. 34, 175–186. Sinton, L.W., Mackenzie, M.L., Karki, N., Braithwaite, R.R., Hall, C.H., Flintoft, M.J., 2010. Transport of Escherichia coli and F-RNA bacteriophages in a 5 m column of saturated pea. J. Contam. Hydrol. 117, 71–81.

32

J. Xie et al. / Journal of Contaminant Hydrology 172 (2015) 24–32

Small, H., 1974. Hdrodynamic chromatography: a technique for size analysis of colloidal particles. J. Colloid Interface Sci. 48, 147–161. Xie, J.C., Lu, J.C., Zhou, X.H., Wang, X.H., Li, M., Du, L.L., Zhou, G.Q., 2012. The kinetic stability of colloid-associated plutonium: settling characteristics and species transformation. Chemosphere 87, 925–931. Xie, J.C., Lu, J.C., Zhou, X.H., Wang, X.H., Li, M., Du, L.L., Liu, Y.H., Zhou, G.Q., 2013a. Plutonium-239 sorption and transport on/in unsaturated sediments: comparison of batch and column experiments for determining sorption coefficients. J. Radioanal. Nucl. Chem. 296, 1169–1177. Xie, J.C., Wang, X.H., Lu, J.C., Zhou, X.H., Lin, J.F., Li, M., Xu, Q.C., Du, L.L., Liu, Y.H., Zhou, G.Q., 2013b. Colloid-associated plutonium transport in the vadose zone sediments at Lop Nor. J. Environ. Radioact. 116, 76–83. Xie, J.C., Lu, J.C., Lin, J.F., Zhou, X.H., Li, M., Zhou, G.Q., Zhang, J.H., 2013c. The dynamic role of natural colloids in enhancing plutonium transport through porous media. Chem. Geol. 360/361, 134–141.

Xie, J.C., Lu, J.C., Lin, J.F., Zhou, X.H., Xu, Q.C., Li, M., Zhang, J.H., 2014a. Insights into transport velocity of colloid-associated plutonium relative to tritium in porous media. Sci. Rep. 4, 5037. http://dx.doi.org/10.1038/srep05037. Xie, J.C., Lin, J.F., Zhou, X.H., Li, M., Zhou, G.Q., 2014b. Plutonium partitioning in three-phase systems with water, colloidal particles, and granites: new insights into distribution coefficients. Chemosphere 99, 125–133. Xie, J.C., Lin, J.F., Zhou, X.H., Li, M., Zhou, G.Q., 2014c. Plutonium partitioning in three-phase systems with water, granite grains, and different colloids. Environ. Sci. Pollut. Res. 21, 7219–7226. Zänker, H., Hennig, C., 2014. Colloid-borne forms of tetravalent actinides: a brief review. J. Contam. Hydrol. 157, 87–105.

Colloid-associated plutonium aged at room temperature: evaluating its transport velocity in saturated coarse-grained granites.

The fate and transport of colloidal contaminants in natural media are complicated by physicochemical properties of the contaminants and heterogeneous ...
897KB Sizes 0 Downloads 6 Views