1. Introduction

A major pathway for radionuclides released from a radioactive waste repository into the biosphere may occur via groundwater. The migration process and mechanism of radionuclides depend on the physicochemical properties and forms of radionuclides when they migrate through geological media. Particulate migration, especially colloid migration, in geological media is usually very complex, and their behaviors are not well analyzed and understood. Although many studies were performed on the migration processes of various colloids, which were known to accelerate the migration of radionuclides through geological media [1-7], a comprehensive understanding and analytical explanation have not yet been provided. Thus, the effect of colloids on radionuclide migration is not rigorously considered in the safety assessment of radioactive waste repositories, although many scientists recognized and approved the importance of colloids in radionuclide migration [1-4].

To evaluate the roles and importance of colloids in radionuclide migration, the basic physicochemical properties of colloids, such as composition, size, concentration, and stability in the subsurface environment should be characterized preferentially. Particularly, the properties of natural aquatic colloids existing in groundwater are very important, because the groundwater colloids can form radiocolloids by combining with radionuclides that are released from a repository and then migrate through groundwater. Thus far, many studies were performed for the characterization of natural aquatic colloids, while only a few studies were carried out for groundwater colloids in granitic rock systems [8-19].

Generally, colloids are defined as particulate matter with sizes ranging from 1 nm to 1 μm [1,13]. In terms of radioactive waste disposal, colloids can be classified as natural colloids that exist naturally in aquatic solutions and engineering colloids generated from various engineering components of a disposal system. Additionally, radioactive colloids can be classified into true colloids (or homogeneous colloids) formed by hydrolysis, polymerization, and growth of dissolved radionuclides, and pseudo-colloids (or heterogeneous colloids) formed by the association of dissolved radionuclides with natural and engineering colloids [1,6].

It is known that groundwater colloids exist ubiquitously in all groundwater systems. Generally, the concentration and size of groundwater colloids vary with depth depending on hydrological, geological, and geochemical conditions. Organic colloids such as humic substances are abundant in surface water at shallow depths, while inorganic colloids such as metal oxides and clay particles exist in all types of subsurface waters [10]. Determining the properties of groundwater colloids is very difficult because of their sensitive behavior under different geochemical conditions and cautious handling and treatment processes. For example, we should be very careful in the sampling process of groundwater for the characterization of groundwater colloids because the properties of groundwater colloids can be overestimated for many reasons [10].

Colloidal particles should be separated from groundwater to analyze the properties of groundwater colloids and may be concentrated for the convenience of various analyses. One of the most common methods for separation and concentration of colloidal particles is the ultrafiltration method, and tangential flow ultrafiltration (TFUF) is considered as a highly effective ultrafiltration method [20,21]. Earlier studies show that ultrafiltration methods using ultrafilters with different pore sizes were used to analyze the size of colloidal particles [14,17,22]. However, this method has a few drawbacks, such as careful treatment and sample contamination. Thus, nondestructive measurement methods, such as Coulter counter [8], scanning electron microscopy/ transmission electron microscopy (SEM/TEM) [23], atomic force microscopy (AFM) [24], and dynamic light scattering (DLS) [25] are widely used for measuring the size of colloidal particles.

The DLS method using light scattering phenomena of particles is considered as a very effective method for measuring the size and concentration of colloidal particles suspended in the solution. However, this method has a detection limitation when the size and concentration of colloidal particles are small (< 10 nm) and low (< 100 μg·L⁻¹), respectively. A new method called laser-induced breakdown detection (LIBD) was developed [26] and has been successfully used to measure the size and concentration of colloidal particles in aquatic solutions without preconcentration of the colloids [15,16,27]. Recently the asymmetric flow field-flow fractionation (AF4) method has been used for separating and monitoring groundwater colloids by combining various detection systems such as UV-Vis (Ultraviolet-Visible) detector [18], ICP-MS (Inductively Coupled-Plasma Mass Spectrometry) [15,19,27] or LIBD [15,27].

It is also very important to determine whether colloidal particles are stable or unstable in a given groundwater condition when colloids migrate in the groundwater. If colloids are unstable, they will not migrate further because they would coagulate and precipitate. Generally, the stability of colloids in groundwater depends on various properties and conditions such as: 1) the size, density, and surface charge of the colloids, 2) solution chemistry of the groundwater considered, and 3) hydrodynamic flow conditions of the groundwater system considered [28,29]. Techniques used to determine the stability of colloidal particles are diverse, and various methods were used [28]. Among these methods, the most easy and reliable method is the zeta-potential measurement of a colloid surface using the dynamic light scattering technique [29].

Although various advanced techniques have been developed, it may be impossible to characterize groundwater colloids using only one or two analytical methods. An approach using multiple analytical methods is needed to characterize groundwater colloids reliably. Therefore, in this study, the properties of groundwater colloids such as composition, size, concentration, and stability sampled from different depths of a granitic site were investigated by using the multiple analytical approaches after separating and concentrating the groundwater colloids using the TFUF method. Furthermore, the roles of groundwater colloids in the mobility of radionuclides in a fractured granite system are also discussed.

2. Experiment and Analyses

2.1 Groundwater Sampling

The groundwater samples used in this study were obtained from the DB-1 borehole of the KAERI (Korea Atomic Energy Research Institute) Underground Research Tunnel (KURT) site, where a multi-packer system was installed. KURT is a small underground research tunnel located within the KAERI research area in northern Daejeon, Republic of Korea (see Fig. 1). Geological and geochemical information about the KURT site, which includes the base rock, can be found in earlier works [30-32]. Groundwater samples can be taken at eight different locations from the DB-1 borehole using the installed multi-packer system. These eight locations were chosen by considering the presence of water-conducting fractures within the borehole region.

Fig. 1

Location, geological formation, and facility outline of the KAERI Underground Research Tunnel (KURT) where the groundwater samples were taken.

In this study, three different groundwater samples (named I5, I6, and I7) taken at relatively deeper locations of the DB-1 borehole were used for the characterization of groundwater colloids because shallow groundwater samples can be affected by an inflow of surface water and unknown variations from other in situ experiments. Table 1 shows the geochemical parameters and elemental compositions of the three groundwater samples used for the characterization of the groundwater colloids. The values of pH and electric conductivity (EC) of the groundwater samples were almost constant regardless of depths, and ranged 8.3–8.5 and 170–174 μS·cm⁻¹, respectively. The contents of dissolved oxygen (DO) also show constant and low values (0.02 mg·L⁻¹) for the three groundwater samples. Despite the low DO values, the redox conditions of the groundwater samples show weakly oxidizing conditions (Eh = 11–108 mV) depending on the depth.

Table 1

Geochemical properties and elemental composition of the KURT (KAERI Underground Research Tunnel) groundwater samples from DB-1 borehole

A carboy was used for sampling and preservation of the groundwater samples, comprising a custom-made 40 L aluminum (Al) container, and coated on the inside with Teflon to prevent the dissolution of Al by contacting groundwater for a long time. A specially designed device was also installed in the cover of the carboy to prevent groundwater from contacting with air and maintain inside pressure of the carboy. Fig. 2 shows a photograph of the groundwater sampling work from the DB-1 borehole and the specially designed carboy.

Fig. 2

Groundwater sampling from a DB-1 borehole where a multi-packer system is installed at the KURT.

Groundwater was sampled using natural confining pressure from the DB-1 borehole without pumping. The sampling rate was controlled to be less than 400 mL·min⁻¹ with a valve to prevent generation of undesirable particulate matter when the sampling rate was too high in the sampling process [13]. All groundwater samples were filtered using a 450 nm prefilter (Millipak 40, Millipore, MA, USA) to remove particulate matter larger than 450 nm in diameter before performing ultrafiltration.

2.2 Separation and Concentration of Groundwater Colloids

A TFUF system consists of a peristaltic pump (Masterflex, Cole-Parmer), a filter holder (Pellicon Mini, Millipore) including tubing, and an ultrafilter (regenerated cellulose ultrafilter, Millipore). The pore size of the ultrafilter was 10,000 NMWC (nominal molecular weight cut off), which is equivalent to a pore size of approximately 1 nm which defines the boundary between colloid and dissolved species. Thus, most of the groundwater colloids cannot pass through the ultrafilter, and groundwater colloids can be concentrated when they contact tangentially with the ultrafilter by a repeated circulating process. The TFUF was carried out in an anaerobic globe box filled with N₂ gas (99.999%) to preserve the properties of groundwater colloids and minimize air contamination (see Fig. 3).

Fig. 3

Schematic diagram (upper) for the tangential flow ultrafiltration system (TFUF) used in the separation and concentration of groundwater colloids. Photographs show the TFUF system (lower left) and the same system installed in an anaerobic glove box (lower right).

Before conducting TFUF, the TFUF system was cleaned several times using 0.1 N NaOH solution and high purity (18.2 MΩ·cm) deionized water (Milli-Q^®, Millipore). The input flow rate of groundwater (W), flow rate of filtered water (F), and flow pressure of recovered water after ultrafiltration (R) were controlled appropriately to increase the efficiency of the TFUF system. From a preliminary experiment for the TFUF system, W was set to 400 mL∙min⁻¹ and R was set to 12 psi, however, F was not controlled manually. Consequently, 20 L of groundwater was concentrated into 200 mL for I7 and 100 mL for I5 and I6 groundwater samples, using the TFUF system. Thus, the concentration factor defined as CF (CF = V_i/V_f, V_i = initial volume of solution, and V_f = final volume of water after concentration) [21], can be estimated as 100 for I7 and 200 for I5 and I6. Concentrated samples of groundwater colloids were preserved in a refrigerator until the groundwater colloids were characterized using various analytical methods.

2.3 Analytical Methods

The composition of the groundwater colloids was determined by the difference in elemental composition between the concentrated and the original groundwater samples. The compositions of the groundwater samples were analyzed using ICP-MS (Ultramass 700, Varian). For each groundwater sample, 20 μL of HNO₃ (Merck, 70%) was added to approximately 10 mL of the sample to make the sample pH approximately 1.0, and then preserved in a refrigerator before ICP-MS measurements. The major elements analyzed by ICP-MS to determine the composition of groundwater colloids are Na, Ca, Fe, K, Mg, Si, Li, Al, Mo, and Ba.

The average size and size distribution of the groundwater colloids was determined using the DLS technique (Zetasizer Nano-ZS, Malvern Instruments) using concentrated groundwater samples without a pretreatment process. Additionally, the stability of groundwater colloids can also be determined using a special cell prepared for electrophoretic mobility (or zeta-potential) measurements using the same DLS instrument [28].

The size information of the groundwater colloids in the concentrated groundwater samples was also microscopically analyzed by field emission-transmission electron microscopy (FE-TEM, JEM-2100F, JEOL) as a comparative analysis to the DLS method. The composition of the groundwater colloids was analyzed using an energy dispersive X-ray spectroscopy (EDS) device (INCA Energy, Oxford) attached to the FE-TEM. For the FE-TEM/ EDS measurements, a small amount of the concentrated colloid solution was dropped on a copper-grid sample plate of 200 mesh and then dried on an absorbing paper. Thereafter, the copper-grid sample plate was preserved in a glove box to avoid contact with air before measurements. Measurements using the FE-TEM/EDS were carried out under an accelerated voltage of 200 V and a current of 15 mA.

The total dissolved organic carbon (DOC) of the groundwater samples were directly analyzed with a total organic carbon (TOC) analyzer (TOC-V CSH, Shimadzu). Liquid chromatography coupled with an organic carbon detector (LC-OCD, DOC-Labor Dr. Huber) was used to characterize dissolved organic matter (DOM) in the groundwater samples. LC-OCD is a very useful method for characterizing DOM in various aquatic systems [33-35]. Fig. 4 shows the schematic diagram of the LC-OCD system used in this study.

Fig. 4

Schematic diagram for the conventional liquid chromatography-organic carbon detector (LC-OCD) system used in the study.

In the LC-OCD system, the mobile phase was precleaned by ultraviolet (UV) oxidation in the DOCOX (dissolved organic carbon oxidation)-reactor and first pumped to an autosampler and then to the chromatographic column. The solution enters the UV-detector (UVD) and then the organic carbon detector. A side stream after UVD enters into the special capillary UV-lamp and then the second UVD measured at 220 nm. Here, dissolved organic nitrogen (DON) and ammonium were determined after conversion to nitrate in the DONOX (dissolved organic nitrogen oxidation)-reactor. The main components of the OCD are the so-called “Graentzel” thin-film reactor and non-dispersive IR-detector. The specifications of the systems are listed in Table 2.

Table 2

Specifications of the LC-OCD (Liquid Chromatography-Organic Carbon Detector) system used in the study

3. Results and Discussion

3.1 Composition and Concentration of Groundwater Colloids

The composition and concentration of groundwater colloids were analyzed by measuring the concentrations of the major elements considered using ICP-MS before and after the TFUF. The concentration of the groundwater colloids was calculated using the difference in elemental concentrations of the concentrated groundwater sample from the original groundwater sample using the concentration factor (CF) of the TFUF as follows:

Toal concentration of groundwater colloids

where

Table 3 shows the calculated results for the composition and concentration of the groundwater colloids determined by ICP-MS measurements. Major elements consisting of groundwater colloids are commonly Ca, Na, Si, and small amount of Fe regardless of the sampling depth. The concentrations of the groundwater colloids ranged about 12–46 μg·L⁻¹ with increasing sampling depths. A maximum concentration of groundwater colloids is observed to be about 46 μg·L⁻¹ for the deepest I7 groundwater sample.

Table 3

Composition and concentration of the groundwater colloids from the KURT site analyzed by inductively coupled plasma-mass spectrometry

Representative electro-microscopic images observed by FE-TEM are shown in Fig. 5 for the groundwater colloids of the concentrated samples from the groundwater samples I5, I6, and I7 at different magnification scales. The elemental composition of the groundwater colloids analyzed by EDS is summarized in Table 4 for some points in the concentrated groundwater samples. It was observed that the major elements consisting of the groundwater colloids were Al, Si, P, S, Ca, and Fe, regardless of sampling depth. The elemental composition of the groundwater colloids observed by EDS is similar to the result from ICP-MS measurements except P and S. Furthermore, as shown in Fig. 6, small crystalline colloids (diameter of about 5 nm) and their agglomerates (approximately 100 nm) composed of Fe were found from FE-TEM/EDS measurements of the I7 groundwater sample.

Fig. 5

Field emission-transmission electron microscopy images of the groundwater colloids sampled from KURT. groundwater samples I5, I6, and I7, concentrated by a tangential flow ultrafiltration at different magnification scales, respectively: (a) and (b) for I5; (c) and (d) for I6, (e) and (f) for I7.

Table 4

Summary of composition analysis results for the groundwater colloids from the KURT site using field emission-transmission electron microscopy/ energy dispersive x-ray spectroscopy

Fig. 6

Field emission-transmission electron microscopy images for the KURT groundwater sample I7 concentrated by a tangential flow ultrafiltration system: (a) small crystalline Fe colloids of about 5 nm; (b) a colloid of about 100 nm existing as an agglomerate of the small colloids. Red circles show the area analyzed by an energy dispersive X-ray spectroscopy.

Thus, the ICP-MS and FE-TEM/EDS measurements indicate that the groundwater colloids are mainly composed of aluminosilicate minerals, calcite, Fe-oxides, and organics. Aluminosilicate minerals are clay minerals that contain mainly Al and Si. Calcite (CaCO₃) may mainly contribute to the Ca content of groundwater colloids. The Fe content found in the I7 sample of Table 3 may have originated from crystalline Fe-oxides. Additionally, groundwater colloids containing P and S elements were found in the EDS measurements, which are presumed to originate from DOM in groundwater. Details of the organic colloids are discussed in the following section. The compositions of groundwater colloids from the KURT groundwater samples show small differences between the groundwater samples. This may be due to the small differences in elemental concentrations of Al, Si, P, S, Ca, and Fe in the original groundwaters, which mainly consist of groundwater colloids, between groundwater samples (Table 1).

3.2 Size of Groundwater Colloids

The average size and size distribution of groundwater colloids sampled from the DB-1 borehole of the KURT site were analyzed using the DLS method and the results are shown in Fig. 7. The size of the groundwater colloids from the I5, I6, and I7 groundwater samples ranged 50–400 nm and were most abundant in the size range 50–100 nm. The average size and standard deviation of the groundwater colloids for the I5, I6, and I7 groundwater samples were estimated to be 255.0 ± 2 .1, 248.9 ± 0.7, and 241.9 ± 4.5 nm, respectively. This result reveals that the average size of groundwater colloids is similar to each other regardless of sampling depth and groundwater properties.

Fig. 7

The results of size measurements using a dynamic light scattering (DLS) method for the groundwater samples concentrated by a tangential flow ultrafiltration system.

As shown in Fig. 5, the size of the groundwater colloids from the I5, I6, and I7 groundwater samples was also measured microscopically using FE-TEM and compared with the results of the DLS measurements. It is difficult to determine the average size of groundwater colloids in the photographs shown in Fig. 5. However, the following observations are highlighted from Fig. 5:

3.3 Organic Colloids

Natural organic matter (NOM) in groundwater that can be identified by the LC-OCD are usually bio-polymers, humic substances, building blocks, neutrals, and organic acids [34]. Table 5 shows the analysis result of DOM for groundwater samples I5, I6, and I7 using the LC-OCD.

Table 5

Characteristics of dissolved organic matters (DOMs) analyzed by liquid chromatography-organic carbon detector depending on the size of DOMs represented by nominal molecular weight cut-off (NMWL)

The DOC contents measured by the TOC analyzer were 2,223, 2,397, and 1,828 μg·L⁻¹ for groundwater samples I5, I6, and I7, respectively (see Table 1). However, as shown in Table 5, the contents of DOC measured by the LC-OCD for the groundwater samples I5, I6, and I7 were 1,130, 1,743, and 384 μg·L⁻¹, respectively. The DOC contents measured by the LC-OCD were much lower than those measured by the TOC analyzer. The low DOC content by the LC-OCD may be due to the inherent chromatographic characteristics of DOM in separating organic matter.

The DOC content measured by the LC-OCD for the I7 groundwater sample was 384 μg·L⁻¹, which is a typical value found in common groundwater. However, the DOC contents for I5 and I6 were 1,130 and 1,743 μg·L⁻¹, respectively, which are relatively higher values compared to that of I7 but are still within the range of typical groundwater. This difference in DOC contents between groundwater samples may have originated from different sources of groundwater. Else, the I5 and I6 groundwaters could be affected by surface water through a water-conducting fracture network in that area considered.

Organic matter in the size range of colloids analyzed by the LC-OCD is a biopolymer and humic substance [34,35]. The concentrations of biopolymer and humic substance for groundwater samples were 1–11 and 8–18 μg·L⁻¹, respectively, as presented in Table 5. The concentrations of organic colloids (sum of biopolymer and humic substance) were estimated to be 15, 29, and 18 μg·L⁻¹ for the I5, I6, and I7 groundwater samples, respectively. Among the groundwater samples, I6 showed the highest concentration of organic colloids, and this trend was very similar to that of DOC content. However, it is observed that the major organic matter analyzed by the LC-OCD are building blocks and neutrals (aldehydes, ketones, etc.) that can be formed by decomposition or breakdown of humic substances.

Thus, it is concluded that the concentrations of organic colloids in the KURT groundwater samples are relatively low, which is a typical trend observed in granitic groundwater systems. When smaller DOMs such as building blocks and neutrals are associated with dissolved uranium species, the mobility of uranium can be greatly affected, because DOMs can strongly combine with uranium in the groundwater.

3.4 Stability of Groundwater Colloids

Colloidal stability in groundwater conditions is very important for evaluating the mobility of colloids when colloids migrate through groundwater systems. In this study, the stability of groundwater colloids from the KURT site was analyzed by measuring the electrodynamic zeta-potentials of groundwater colloids concentrated by TFUF. The zeta-potentials of groundwater colloids were determined by transforming electrophoretic mobility (μ_e) measured by the DLS method into zeta-potential (ζ) by the Helmholtz- Smoluchowski equation [28]:

where μ_e and ζ are the electrophoretic mobility and zetapotential of groundwater colloids, respectively, η_l is the viscosity of the groundwater solution, and ε_r and ε_o are the dielectric constants of the solution medium and free space, respectively.

Table 6 shows the measured μ_e and calculated ζ of the groundwater colloids concentrated from the KURT groundwater samples. Generally, colloidal particles are classified into highly unstable, relatively stable, moderately stable, and highly stable when ζ values are ± 0–10 mV, ± 10–20 mV, ± 20–30 mV, and over ± 30 mV, respectively [36]. Thus, the stability of colloidal particles increases as ζ value is far from 0 mV. Based on the results in Table 6, it can be concluded that the groundwater colloids from the KURT groundwater samples I5, I6, and I7 were moderately stable in the given groundwater conditions, although the groundwater colloids were concentrated using the TFUF from the original groundwater samples. Consequently, the groundwater colloids in KURT can be mobile through granitic groundwater because they are stable in the given groundwater conditions without any coagulation or precipitation.

Table 6

Electrophoretic mobility (μ_e) and zeta-potential (ζ) of the KURT groundwater colloids

3.5 Comparison of Properties for Granitic Groundwater Colloids

Based on the results of the ICP-MS and FE-TEM/EDS measurements, the groundwater colloids were found to be mainly composed of clay minerals, calcite, Fe-oxides, and some organics. As previously explained in Table 7, the average sizes of the groundwater colloids determined by the DLS method were approximately 255, 249, and 242 nm for the I5, I6, and I7 KURT groundwater samples, respectively. The concentrations of organic and inorganic colloids are summarized in Fig. 8. The concentrations of inorganic colloids determined by ICP-MS were approximately 18, 12, and 47 μg·L⁻¹, for the I5, I6, and I7 groundwater samples, respectively. Contrarily, the concentrations of organic colloids determined by LC-OCD were approximately 15, 29, and 18 μg·L⁻¹ for the I5, I6, and I7 groundwater samples, respectively. Thus, the total concentrations of groundwater colloids in the I5, I6, and I7 groundwater samples were 33, 41, and 64 μg·L⁻¹, respectively.

Table 7

Comparison of the characteristics of groundwater colloids from granitic groundwater samples

Fig. 8

Summary of the concentrations of groundwater colloids from the KURT groundwater samples, I5, I6, and I7.

The characteristics of groundwater colloids measured in the present study are compared with those from other studies at granitic rock formations in Table 7. According to the comparative analysis of Table 7, the composition of groundwater colloids in granitic groundwaters is relatively similar to each other, regardless of sampling sites. The concentrations of the groundwater colloids from the KURT site are relatively smaller than those at other sites. This may be due to the fact that the groundwaters of the KURT site have low ionic strength and contain less organic matter.

As shown in Table 7, the size distribution and composition of groundwater colloids obtained at a depth of 300 m of MIU (Mizunami Underground Research Laboratory) of granite in Japan using a microfiltration/ultrafiltration method combined with SEM/EDS showed that most of the inorganic colloids were in the size range of 50–450 nm and composed of clay minerals and iron (hydr)oxides [17]. They also reported that organic matter might associate with inorganic colloids. In particular, the analytical study for the KURT groundwater colloids (from the groundwater sample I5) using the AF4 technique combined with a UV-Vis detector, as shown in Table 7, showed that the size distribution ranged from a few nm to 40 nm [18]. They also reported the size distribution of groundwater colloids measured by the conventional DLS method after preconcentration by ultrafiltration with a main population in a few hundreds of nanometers, which is very similar to our result measured by the DLS technique. The discrepancy between the size distributions obtained by DLS and AF4 methods can be explained with physicochemical change of colloids during ultrafiltration, different sensitivity of detection systems, and limitations of DLS method for analysis of natural samples [18].

3.6 Roles of Groundwater Colloids in the Mobility of Radionuclides

As mentioned earlier, many laboratory and field studies [1-27] were performed to investigate the characteristics of natural aquatic colloids in subsurface systems. The main results of these studies demonstrated the importance of natural colloids in the mobility of radionuclides, especially of strongly sorbing radionuclides such as actinides. The roles of groundwater colloids in the mobility of radionuclides can be considered in terms of two aspects: (1) groundwater colloids (both inorganic and organic colloids) will facilitate the migration of strongly sorbing radionuclides such as Pu [2,3]; (2) surface coating by organic groundwater colloids such as humic substances will contribute to the stabilization of inorganic colloids [37-39].

Once sorbed on natural colloids forming pseudo-colloids, the migration characteristics of radionuclides become dominated by those of the colloids. Hence, the retardation of a radionuclide is also dependent upon the properties of the host colloid (i.e., natural groundwater colloids). Additionally, as colloids are excluded from the micropores of the rock matrix, the effectiveness of rock matrix diffusion and sorption as retardation mechanisms will be lowered. Consequently, the migration velocity of radio-colloid is generally greater than that of the dissolved radionuclides [6].

In terms of colloidal stability, the importance of NOMs, such as humic substances, has long been recognized to prevent aggregation of inorganic colloidal particles through the alteration of colloidal surface charges, because sorbed NOMs can dominate the surface properties of colloidal particles [38]. As NOMs contain many functional groups, their sorption onto colloidal particles produces a uniformly negative surface charge on the surface of the colloidal particles. This charge could increase the stability of colloidal particles by producing repulsive electrostatic double-layer forces, at least at low ionic strengths. Additionally, some NOMs have high a molecular mass, which could impart stability through repulsive steric forces. Many studies show that inorganic colloids exhibit higher colloidal stability in the presence of humic substances [39].

Generally, the significance of colloids in the mobility of radionuclides in a radioactive waste repository can be estimated using the “colloid ladder” as shown in Fig. 9 [40]. It is already well accepted that five requirements must be fulfilled to prove that colloid-facilitated transport of radionuclides in a repository host rock may be of significance to the long-term performance of a waste repository. The five requirements are as follows:

Fig. 9

The “colloid ladder” concept for evaluating the significance of colloids in the mobility of radionuclides in a radioactive waste repository.

Consequently, based on the colloid ladder concept, the groundwater colloids found in the fractured granite system of the KURT site can form stable radio-colloids and increase the mobility of radionuclides when they associate irreversibly with radionuclides released from a radioactive waste repository.

4. Conclusions

In this study, groundwater colloids existing in granitic groundwater were taken from the DB-1 borehole of the KURT site where a multi-packer system is installed. The groundwater colloids were separated and concentrated using a TFUF system and then characterized using various analytical methods. Based on the results of ICP-MS and FE-TEM/EDS measurements, the groundwater colloids are mainly composed of clay minerals, calcite, Fe-oxides, and some organics. The average size of the groundwater colloids determined by the DLS method was mainly 242–255 nm while some smaller (10–50 nm) colloids of Fe oxides were also observed in the FE-TEM/EDS measurements. The concentrations of inorganic colloids determined by ICP-MS were approximately 12–47 μg·L⁻¹ and the concentrations of organic colloids determined by LC-OCD were approximately 15–29 μg·L⁻¹. Thus, the total concentrations of groundwater colloids were approximately 33–64 μg·L⁻¹ for KURT groundwater samples. From the results of zetapotential measurements, it was found that the groundwater colloids from the KURT groundwater samples were moderately stable under the given groundwater conditions. These characteristics of the KURT groundwater colloids appear to be very similar to those of other granitic systems.

Thus far, various methods were developed and proposed to analyze natural aquatic colloids, while their characterization was not achieved successfully using only one method. Therefore, in this study, several different analytical methods were used to obtain reliable data on the characteristics of groundwater colloids, such as composition, concentration, size, and stability. The results of this study will provide the basic properties of groundwater colloids in the granitic groundwater system of the KURT site, which is necessary for understanding and evaluating the mobility of colloids and their effects on radionuclide migration through fractured granite rock.