Environmental Fate and Transport of Solvent-Stabilizer Compounds - Environmental Investigation and Remediation - page 143

For hydrophobic compounds, K d and K ow are highly correlated for all sorbents except pure
mineral substrates, in which micropores may play a role. For nonpolar organic compounds such as
chlorinated solvents, the correlation between the logarithms of K oc and K ow values was determined
to follow the relationship (Schwarzenbach and Westall, 1981)

log K oc = 0.72 ( K ow + 0.49), R 2 = 0.95. (3.34)

Another commonly used relationship (Karickhoff, 1981) is

K oc = 0.41 K ow . (3.35)

These two relationships are not reliable for estimating partition coefi cients of polar organic
compounds such as alcohols and aliphatic ethers. A regression analysis of about two dozen experi-
mentally determined K oc values produced the following correlation (Nguyen et al., 2005):

log K oc = 0.73 (log K ow + 0.52), R 2 = 0.83. (3.36)

Equation 3.36 has a rather low correlation coefi cient ( R 2 ). The compounds used to derive Equation
3.36 do not include any ether compounds. A more dynamic and reliable approach to estimating K oc
uses linear free-energy relationships (LFERs) * to correlate compound structure and sorption to soil
organic matter (Niederer et al., 2006b).
Numerous other relationships have been developed for different classes of compounds. In gen-
eral, compounds with octanol/water partition coefi cients less than 1000 will migrate easily through
the subsurface without appreciable retardation due to sorption (McCarty and Reinhard, 1981).
K ow values in the literature are prone to considerable uncertainty, sometimes approaching two
orders of magnitude or more. Experimental determination of K ow values may incorporate method
errors; for example, small quantities of emulsii ed octanol in the aqueous phase can carry high con-
centrations of the organic compound being tested, leading to erroneously high determinations of the
concentration in water (Mackay et al., 1993). K ow sorption models may be unreliable for polar com-
pounds even if experimental K ow values are used for input data (Niederer et al., 2006b).
Using physicochemical data without carefully assessing its reliability can lead to signii cant
error in estimating contaminant fate. Fate and transport experts address poor data quality of K ow
values in the literature by using structure-activity estimation methods to compute K ow , by using
carefully selected data, or by applying probabilistic assessments to fate and transport modeling
(Renner, 2002).

3.3.2.6 Effect of Surface Oxides on Sorption
Increased amounts of oxygen on granular activated-carbon surfaces can decrease the afi nity of
organic compounds for sorption to carbon. Oxidation of carbon by acids increases the presence of
oxygen surface functional groups and decreases the adsorption capacity of organic compounds
(Snoeyink, 1999). Decreased adsorptive capacity was observed in activated carbon for several sta-
bilizer compounds (phenol, nitromethane, methyl ethyl ketone, n- butanol, and 1,4-dioxane) as a
result of an increase in acidic surface oxides (Vidic and Suidan, 1991).

3.3.2.7 Competitive Sorption
Solvent-stabilizer compounds are usually released as mixtures composed primarily of chlorinated
solvents and oily wastes. The rate of migration of individual components of the mixture will depend

* LFERs, also called the linear Gibbs energy relation, describe a linear correlation between the logarithm of a rate constant
or equilibrium constant for one series of reactions and the logarithm of the rate constant or equilibrium constant for a
related series of reactions.