Temperature dependence of atmospheric concentrations of semivolatile organic compounds

Reported data on the temperature dependence of atmospheric concentrations of semivolatile organic compounds (SOCs) are compiled and expressed as linear regressions of the logarithm of the partial pressure in air versus reciprocal temperature: In p(A) = m/T + b. Two simple models are introduced to explain the dependence of these air concentrations on temperature. The first assumes equilibrium between the atmosphere and the earth's surface. In the second, air concentrations are established as a result of chemical inflow and outflow in advected air and reversible exchange with a soil or water surface. The model equations are rearranged to express the partial pressure of the chemical as a function of temperature. On the basis of these models, it is shown that only under selected circumstances, namely, if surface contamination is high and atmospheric background concentration low, does the slope m of the In p vs 1/T relationship reflect the thermodynamics of air-surface partitioning. Generally, however, m is a measure of the extent to which air concentrations are controlled by evaporation from surfaces close to the sampling location and by advection of air masses with global background concentrations. A shallow slope or low temperature dependence indicates that long-range transport controls atmospheric levels at a sampling site. Steeper slopes indicate high surface concentrations in the vicinity of the site. This hypothesis is applied to the observed temperature dependence of the compiled atmospheric concentration data and is found to be capable of explaining differences in slope m (i) between chemicals, (ii) between sampling sites, and (iii) at different seasons. Research efforts should be directed toward quantifying by measurements and predicting by models the kinetics of exchange of SOCs between the atmosphere and various surfaces.