3.1. Geological material, tracer standards, chemicals, and water phase

Na-bentonite was obtained from the Beijing Research Institute of Uranium Geology without further processing. The true density (ρ_t) of the GMZ bentonite was 2.7 g/cm³. The aqueous phase used in the laboratory tests was synthetic groundwater (SGW). Table 1 lists the average values of anionic and cationic concentrations of underground water sampled from six boreholes at a site with crystalline host rock.

The ionic strength of groundwater in this area is approximately 2.35–2.93 mM. Tritiated water (HTO) was used as a conservative tracer, due to the fact that it is representative of all chemical elements that do not show retention in the solid phase. The standard solution used in this study could be traceable to Packard Spec-ChecTM and NIST SRM 4288A for HTO and ⁹⁹TcO₄⁻, respectively. The samples were measured using a liquid scintillation analyzer (LSA, Perkin Elmer Tri-Carb 3170 TR/SL) with counting efficiency of 91%–92% of ⁹⁹Tc. 10 mL of the sample was combined with 10 mL scintillation cocktail (Ultima Gold^TM LLT and Ultima Gold^TM AB, Perkin Elmer, USA for HTO and ⁹⁹Tc, respectively) in a 20 mL polyethylene counting vial. All chemicals used in the experiments were of analytical purity, and deionized water (DIW) was used throughout the experiments.

3.2. Characterization

The specific surface area of the GMZ bentonite sample was determined using the N2-BET instrument (Micromeritics ASAP 2020). The chemical composition of GMZ bentonite was analyzed using a wavelength dispersive X-ray fluorescence (WD-XRF) spectrometer (Kxios, PANalytical Inc). Analysis of the Na-bentonite crystals was conducted using a Bruker D8 Advance diffractometer operated at 40 kV and 40 mA with a ceramic Cu X-ray (1.5418 Å) tube. Scanning was performed in the range of 5–80 2θ at a speed of 1°/min. A LynxEYE compound silicon strip detector with lower background (< 0.1 cps) was employed. Patterns were identified via comparison with JCPD standards. The Na-bentonite sample was prepared in pressed potassium bromide (KBr) pellets and characterized using Fourier transform infrared (FT-IR) spectrometer (Horiba FT-730 spectrometer, Kyoto, Japan) over a frequency range from 4000 to 400 cm⁻¹ at a resolution of 4 cm⁻¹.

3.2. Through-diffusion experiment

Samples of bentonite clay of various dry densities (1.2, 1.4, 1.6, 1.8 and 2.0 g/cm³) were statically compacted using a lifting jack (providing uniaxial compressive force) into thin slabs of 3 mm in thickness with a cross-sectional area of 19.6 cm². The slabs were placed between two chambers filled with SGW (100 mL), one of which contained the diffusing radionuclide. The sample discs were sandwiched between the source and collection (sampling) reservoirs made of chemically inert polypropylene (PP) with excellent mechanical properties. Overall, the diffusion cell comprised three parts: a source reservoir, a compacted bentonite sample holder, and a collection (sampling) reservoir. Fig. 1 is a schematic illustration of the column units used in the through-diffusion experiments. The sample was fixed using two Teflon O-ring washers (5 cm outer diameter and 3.6 cm inner diameter, 2.9 cm thickness) and two porous filters (Φ5 cm×0.7 cm) with reinforced PP material at both ends of the inlet and outlet reservoirs. The sample holder was firmly fastened by screw-threads to the chamber units using a torque wrench to reduce the deformation of the cells, minimize diffusion resistance and avoid the swelling pressure induced by bentonite clay after being saturated with water. A commercially available micro-porous PVC-SiO₂ membrane (separator) with 40–50 nm pore size (Europe Amer-Sil^TM) was used to prevent the generation of colloids (average size of 300 nm) from the bentonite. Molecules of HTO or H₂O with a size of 2.75 Å, and ⁹⁹TcO₄⁻ with a size of 5.04 Å, were able to penetrate the membrane easily.

Fig.1.Full-screen View

Schematic drawing of diffusion chamber (in mm)..

Diffusion experiments were conducted as follows:

(1) Before being poured into the source reservoir, the SGW underwent equilibration with the solid phase of compacted bentonite over a period of 12 days, wherein an air pump was used to create suction pressure of 2–3 kg/cm² to enable the injection of a large volume of water into the source reservoir in order to accelerate saturation;

(2) Following water saturation, a very small volume of inactive tracer solution and HTO spiked SGW were poured into the source reservoir and maintained at a fixed concentration of approximately 280 dpm/mL;

(3) Samples of 10 ml were obtained from the collection reservoir (initially devoid of radioisotopes) at definite intervals, while the source tank underwent circulation for 30 min using a peristaltic pump;

(4) The HTO concentration was tested using LSA; and

(5) The HTO source reservoir was replaced by SGW spiked with ⁹⁹Tc radionuclide and maintained at a fixed concentration of approximately 1180 dpm/mL, whereupon Steps (3) and (4) were repeated.

The experimental setup included five through-diffusion chambers for dry densities of 1.2, 1.4, 1.6, 1.8 and 2.0 g/cm³ (Fig. 2).

Fig.2.Full-screen View

Experimental setup for 5 through diffusion chambers compacted to different dry densities (1.2, 1.4, 1.6, 1.8 and 2.0 g/cm³).

The experiments were carried out at room temperature. The volume of the inlet reservoir was sufficient to maintain the concentration at a near constant level (within a few percent) throughout the experiments. When the diffusion experiment was started, the solution in the collection reservoir was replaced with fresh solution (SGW) at regular time intervals (daily). The replaced solution was collected and measured using LSA to analyze the concentration of HTO and ⁹⁹Tc, which had diffused out from the bentonite. The quantities of individual radionuclides obtained in each collection and the accumulated concentration ratio CR(t) were plotted against the elapsed time t to yield through-diffusion curves.

4.1. Characterization

The N2-BET surface area of the GMZ bentonite sample was 26 m²/g. XRF analysis results show that the chemical composition was as follows: SiO₂, 63.71%; Al₂O₃, 13.44%; MgO, 2.67%; Fe₂O₃, 2.01%; Na₂O, 1.67%; CaO, 0.90%; K₂O, 0.67%; TiO₂, 0.12%; MnO, 0.04%; P₂O₅, 0.03%; and LOI, 12.91%.

Fig. 3(a) shows XRD patterns of the Na-bentonite, indicating that it is composed primarily of montmorillonite (marked as M). The other peaks indicate impurities corresponding to quartz (Q), albite (A), a type of feldspar, and sodium aluminum silicate (S). Fig. 3(b) shows the FT-IR spectrum of Na-bentonite. The absorption band at 3634 cm⁻¹ indicates the stretching vibrations of structural OH groups (Si-OH) in the GMZ bentonite. Free water molecules in bentonite produce a broad band at 3430 cm⁻¹, corresponding to H₂O-stretching vibrations. The band at 1643 cm⁻¹ is associated with an overtone of the bending vibration of water. The band at 795 cm⁻¹ confirms the presence of quartz and the bands at 536 and 477 cm⁻¹ are attributed to Al-O-Si and Si-O-Si bending vibrations, respectively. The very strong absorption band at 1047 cm⁻¹ is associated with Si-O bending vibrations [18].

Fig.3.Full-screen View

XRD patterns (a) and FT-IR spectrum (b) of the GMZ bentonite.

4.2. Diffusion of HTO and ⁹⁹Tc

The activity concentrations of tritium (HTO) and ⁹⁹Tc were measured. and the accumulative concentration curves are plotted as a function of diffusion time. The diffusion curves yielded by ΣC_t/C₀ versus time agree with the theoretical breakthrough curves (BTCs), as shown in Fig. 4. The BTCs can be divided into two parts: (a) the transient period in which the concentrations build across the sample, and (b) the period of steady-state diffusion in which the concentration in the measurement chamber increases steadily following the asymptote; i.e., a linear increase is observed over time.

Fig.4.Full-screen View

Break-through curves of (a) HTO and (b) ⁹⁹Tc in the compacted bentonite with dry densities of 1.2, 1.4, 1.6, 1.8 and 2.0 g/cm³.

It should be noted that Eq. (4) is valid only for the steady-state condition. As long as the diffusion process remains in the transient stage, the values of D_a and R_f, obtained from Eqs. (3) and (4) deviate. The dimensionless parameter, t_d=D_at_f /L², is introduced here to determine whether the diffusion process reaches equilibrium. Crank [14] claimed that steady state diffusion is achieved when t_d exceeds 0.45. As listed in Tables 2 and 3, all the t_d values for the HTO and ⁹⁹Tc in each column are greater than 0.45, indicating that the experiment time was of sufficient duration to reach steady-state diffusion. The diffusion parameters calculated from Eq. (4) are also listed in Tables 2 and 3, which summarizes the through-diffusion experiments involving compacted bentonite (0.3 cm slabs) of various densities. The apparent and effective diffusion coefficients of HTO ranged from (1.68±0.40)×10⁻¹¹ to (2.80±0.62)×10⁻¹¹ m²/s and from (4.61±1.28)×10⁻¹² to (16.2± 2.50)×10⁻¹² m²/s, respectively. The apparent and effective diffusion coefficients of ⁹⁹TcO₄⁻ range from (5.26±0.16)×10⁻¹² to (7.78±0.43)×10⁻¹² m²/s and from (1.49±0.002)×10⁻¹² to (4.16±0.07)×10⁻¹² m²/s, respectively. The distribution coefficients of HTO and ⁹⁹TcO₄⁻ ranged from (0.70±0.12)×10⁻² to (1.36±0.53)×10⁻² mL/g and from (1.12±0.06)×10⁻² to (5.79±2.22)× 10⁻²mL/g, respectively. Our results from experiment dealing with diffusion parameters are comparable to those obtained in Ref.[10, 11, 19, 20]. The difference between our values and those in previous work can be attributed to differences in ionic strength and pH value of the liquid phase, the use of different types of bentonite, variations in the selection of steady-state regions on the diffusion break-through graphs, linear regression of the points along the line, and errors due to counting statistics [21].

In this study, HTO was shown to access the total pore volume of the geological materials, which is equivalent to the water content; therefore, this was used as a conservative tracer. As shown in Table 2, the porosity values (θ_exp =D_e/D_a) calculated from the diffusion measurements for HTO were similar to the theoretical value for total porosity (θ_cal). This implies that all the pores in the compacted bentonite were available for the mass transfer of neutral species [22]. ⁹⁹Tc is heptavalent and anionic in an oxidizing environment (as pertechnate ⁹⁹TcO₄⁻). There was very little or no sorption of technetium on bentonite under these experimental conditions, due to exclusion effect of bentonite with a negative surface charge. As shown in Tables 2 and 3, the K_d value decreased with increasing dry density. The differences may be attributed to the fact that HTO and ⁹⁹TcO₄⁻ ions are unable to enter small occluded pores, thereby preventing them access to the entire volume; i.e., all the sorption sites in the compacted bentonite [13].

As the diffusion coefficient depends on the size and shape of molecules and interaction with solvent, the diffusion coefficients of ⁹⁹TcO₄⁻ are significantly smaller than those of HTO due to the larger radius of the hydrated anion (~ 0.35 nm for ⁹⁹TcO₄⁻) [22]. It can be also explained by changes in the species of ⁹⁹Tc during diffusion. The GMZ bentonite used in these experiments was untreated; therefore, it contained Fe-bearing minerals associated with minerals such as magnetite, hematite, goethite, which could significantly slow the mass transport of ⁹⁹Tc by reducing Tc(VII) to Tc(IV) as aqueous (Fe²⁺) or solid [Fe(II)] [20]. Some ⁹⁹TcO₄⁻ could be transformed into insoluble TcO₂·nH₂O [12, 23]. According to XRF data, 2.01% of the Fe₂O₃ in this study was contained in the GMZ bentonite; however, the mechanism underlying the reaction of iron oxide and Tc(VII) has yet to be fully elucidated. Fig. 5 shows that a higher density leads to lower D_e values for HTO and ⁹⁹Tc. Both radionuclides appear to resist migration due to the constrictivity and tortuosity of the bentonite pores, which are directly related to the compaction pressure [21].

Fig.5.Full-screen View

Effective diffusion coefficient of HTO and technetium as a function of dry density.