Page 166 - Partition & Adsorption of Organic Contaminants in Environmental Systems
P. 166

SORPTION FROM WATER SOLUTION     157

            We consider first the probable source of the sorption nonlinearity for nonpo-
            lar solutes, where the nonlinear capacity is relatively small and approaches
            apparent saturation at very low C e/S w (0.010 to 0.015). On peat soil, the non-
            linear capacities of about 0.18mg/g for EDB and 0.15mg/g for TCE are well
            within the allowed monolayer adsorption capacity of the soil based on its BET-
                                2
            N 2 surface area of 1.4m /g. The same is true for EDB and LND on Woodburn
            soil, where the nonlinear capacities of EDB and LND are <0.008mg/g and the
                                        2
            surface area of the soil is 11.2m /g. Since there is little tendency for nonpolar
            solutes to engage in specific interaction with SOM and since the peat soil has
            a very low mineral content, these features might be ascribed to strong solute
            adsorption at low C e/S w values on a small amount of HSACM (e.g., charcoal-
            like materials) in soil (Chiou, 1995), on which water exhibits a weak compet-
            itive adsorption (see Chapter 6, section 6.5, on the adsorption of water on
            activated carbon). At moderate to high C e/S w values, this adsorption is largely
            saturated and the partition in SOM predominates to make the isotherm essen-
            tially linear.
              The HSACM hypothesis is consistent with the characteristics of solute
            adsorption on activated carbon. For adsorbates with a density of about 1g/mL
                                          2
            on a typical activated carbon, 1m /g of carbon surface area corresponds to
            about 0.25mg/g for the adsorbate monolayer capacity, the saturation capacity
            is about twice as high. Adsorption on activated carbon rises sharply at low
            C e/S w values (see the related discussion in Chapter 6). At C e/S w = 0.01 to 0.02,
            the adsorbed capacity is about 40 to 50% of the saturation capacity (Manes
            and Hofer, 1969; Chiou and Manes, 1974). From this point up to C e/S w = 1,
            adsorption approaches full saturation more gradually. With a small quantity
            of assumed HSACM and a significant amount of SOM in soil, the isotherms
            for nonpolar solutes at C e/S w > 0.01 to 0.02 would therefore become relatively
            linear as the (linear) partition into SOM outweighs the adsorption on
            HSACM. On peat soil, the nonlinear EDB and  TCE capacities (0.15 to
            0.18mg/g) are consistent with the HSACM hypothesis, if most of the soil
                            2
            surface area (1.4m /g) comes from a small amount of HSACM. For Woodburn
            soil, the surface area is much higher and the nonlinear capacity for nonpolar
            EDB is much smaller. The large surface area results presumably from mineral
            surfaces, on which the strong interaction with water minimizes the solute
            adsorption (Chiou and Shoup, 1985; Chiou et al., 1985). The lower nonlinear
            capacity of EDB on Woodburn soil may be attributed to a trace amount of
            HSACM in the soil. Since adsorption on activated carbon occurs primarily by
            London forces as discussed in Chapter 4, and the same is expected for
            HSACM, the solute polarity would not be relevant for competitive adsorp-
            tion. This is corroborated by the nonspecific suppression of the EDB sorption
            by both polar and nonpolar co-solutes on peat.
              The nonlinear sorption characteristics of nonpolar solutes on peat are inter-
            nally consistent with the N 2 adsorption data on this sample. The N 2 adsorption
            exhibits a similar sharp rise and a downward concavity at low relative pres-
            sures (P/P°) (£0.02) with a monolayer capacity of 0.36mg/g (Chiou et al., 1993;
   161   162   163   164   165   166   167   168   169   170   171