Sorption equilibrium distribution coefficients

Celorie, J.A., Woods, S.L., Vinson, T.S., and Istok, J.D. A comparison of sorption equilibrium distribution coefficients using... 

In most mathematical analyses used to establish bounds for radionuclide migration rates through the abyssal red clays, the sorption properties of the sediment are generally represented mathematically by the sorption equilibrium distribution coefficients for each of the species involved. These coefficients are usually denoted by Kp. and are defined by... 

Therefore, the preliminary investigation described herein examined several aspects of the behavior of the equilibrium distribution coefficients for the sorption of rubidium, cesium, strontium, barium, silver, cadmium, cerium, promethium, europium, and gadolinium from aqueous sodium chloride solutions. These solutions initially contained one and only one of the nuclides of interest. For the nuclides selected, values of Kp were then... 

Thus, the sorption of chemicals on the surface of the solid matrix may become important even for substances with medium or even small solid-fluid equilibrium distribution coefficients. For the case of strongly sorbing chemicals only a tiny fraction of the chemical actually remains in the fluid. As diffusion on solids is so small that it usually can be neglected, only the chemical in the fluid phase is available for diffusive transport. Thus, the diffusivity of the total (fluid and sorbed) chemical, the effective diffusivity DieS, may be several orders of magnitude smaller than diffusivity of a nonsorbing chemical. We expect that the fraction which is not directly available for diffusion increases with the chemical s affinity to the sorbed phase. Therefore, the effective diffusivity must be inversely related to the solid-fluid distribution coefficient of the chemical and to the concentration of surface sites per fluid volume. 

Table I summarizes some typical distribution coefficients. Sediments become enriched in plutonium with respect to water, usually with a factor of vlO5. Also living organisms enrich plutonium from natural waters, but usually less than sediments a factor of 103 - 101 is common. This indicates that the Kd-value for sediment (and soil) is probably governed by surface sorption phenomena. From the simplest organisms (plankton and plants) to man there is clear evidence of metabolic discrimination against transfer of plutonium. In general, the higher the species is on the trophic level, the smaller is the Kd-value. One may deduce from the Table that the concentration of plutonium accumulated in man in equilibrium with the environment, will not exceed the concentration of plutonium in the ground water, independent of the mode of ingestion.

Therefore, based on available literature, the following sorption results were expected (l) as a result of the smectite minerals, the sorption capacity of the red clay would be primarily due to ion exchange associated with the smectites and would be on the order of 0.8 to I.5 mi Hi equivalents per gram (2) also as a result of the smectite minerals, the distribution coefficients for nuclides such as cesium, strontium, barium, and cerium would be between 10 and 100 ml/gm for solution-phase concentrations on the order of 10"3 mg-atom/ml (3) as a result of the hydrous oxides, the distribution coefficients for nuclides such as strontium, barium, and some transition metals would be on the order of 10 ml/gm or greater for solution-phase concentrations on the order of 10 7 mg-atom/ml and less (U) also as a result of the hydrous oxides, the solution-phase pH would strongly influence the distribution coefficients for most nuclides except the alkali metals (5) as a result of both smectites and hydrous oxides being present, the sorption equilibrium data would probably reflect the influence of multiple sorption mechanisms. As discussed below, the experimental results were indeed similar to those which were expected. 

The distribution coefficients determined for cadmium (at ll C) are given in Figure 3- The coefficients appear somewhat less than the corresponding data for strontium and barium. Such results could be due to either anionic complex formation (22) and/ or a less favorable sorption equilibrium. 

The distribution coefficients evaluated for silver (at C) are also given in Figure 3 The silver coefficients determined in 0.68 N NaCl solutions are somewhat less than the corresponding coefficients for cesium and rubidium, and also for strontium and barium. Such results are probably due to either anionic complex formation (22) and/or a less favorable sorption equilibrium. (FurthermoreJ the experiments done using silver in sodium chloride solutions required equilibration times on the order of 90 days, as opposed to two to four days for most other experiments, and it appears that processes, which may or may not be important, are involved which are not understood.)... 

In all of these cases, the structure of the organic sorbate, the composition of the surface, and the conditions of the vapor or solution exchanging with the solid must be considered. However, it is important to note that with some experience in thinking about the organic chemicals and environmental situation involved, we can usually anticipate which one or two sorption mechanisms will predominate. For example, in Chapter 9 we wrote an expression reflecting several simultaneously active sorption mechanisms, each with their own equilibrium descriptor, to estimate an overall solid-water distribution coefficient for cases of interest (Eq. 9-16) ... 

Geochemical models of sorption and desorption must be developed from this work and incorporated into transport models that predict radionuclide migration. A frequently used, simple sorption (or desorption) model is the empirical distribution coefficient, Kj. This quantity is simply the equilibrium concentration of sorbed radionuclide divided by the equilibrium concentration of radionuclide in solution. Values of Kd can be used to calculate a retardation factor, R, which is used in solute transport equations to predict radionuclide migration in groundwater. The calculations assume instantaneous sorption, a linear sorption isotherm, and single-valued adsorption-desorption isotherms. These assumptions have been shown to be erroneous for solute sorption in several groundwater-soil systems (1-2). A more accurate description of radionuclide sorption is an isothermal equation such as the Freundlich equation ... 

Each sorption experiment was conducted by adding 5.0 mL of the appropriate traced solution, prepared as described above, to a weighed (-1 g) portion of Hanford sediment. To simulate advancement of a radioelement plume from a failed tank through previously waste-wetted sediment, each sediment sample was preequilibrated twice with the relevant untraced solution prior to introduction of the traced solution. Each pre-equilibration lasted at least 2 hr. Following a 7-day equilibration with the traced solution, each sediment-solution mixture was centrifuged, the solution was filtered through an ultrafilter, and the radionuclide solution concentration was determined. Distribution coefficients and fractions of radionuclides sorbed were determined for each sorption experiment. The distribution coefficient, Kd, is the activity per gram of sediment divided by the activity per mL of solution at equilibrium. 

Pilot plant studies (flow rates, 1 cm/s) with the SB-1 anion exchange resin (column diameter, 0.3-0.7 cm) yielded distribution coefficients of the order D = 400 cmVg. The boron sorption process was shown to be film diffusion controlled. The equilibrium values of boron loading were reached in 6-8 hr [280]. Boron elution and resin regeneration were carried out with 0.1 M NaOH. The complete elution of boron required 10 column volumes at 10 BV and yielded concentrates of 100 mg/L. This facilitated the eventual reduction to solid concentrates of alkali metal borates [281]. 

---