Organic mass fraction

FIGURE 9.60 Excess water in particles with diameters 0.1 ¡xm near the Grand Canyon in 1992 as a function of organic mass fraction (adapted from Saxena et al., 1995).

ent reaction mechanisms and products for the gas phase compared to the condensed phase.

In short, understanding the gas-particle partitioning of SOC is important in a number of areas in atmospheric chemistry as well as related aspects of toxicology.

The earliest work in this area assumed that particles in the atmosphere were solid and that the uptake of SOC involved adsorption to a solid or solid-like surface. It was subsequently recognized that many atmospheric particles are liquid or have liquid-like outer layers, and hence the uptake of gases could be treated as absorption into a liquid. These approaches are summarized in the following. It should be noted that these treat the equilibria between the gas- and condensed-phase species; i.e., it is assumed that thermodynamics rather than kinetics controls the distribution between the phases. The implications of this assumption are discussed later.

Given these caveats, we first treat the case of adsorption to a solid surface, and then absorption into a liquid particle or liquid surface layer on a particle. As we shall see, the distribution of SOC between the gas and condensed phases can be used to infer the nature of the sorbent sites.

1. Adsorption on Solid Particles

A gas-particle partition coefficient, Kp, is commonly used to describe the distribution of a SOC between the gas and particle phase, ft is defined as the ratio of the SOC in particles (in units such as ng /xg 1) to that in the gas phase (in units such as ng m 3). In essence, it is the fraction of the mass of total suspended particulate matter (TSP) that is the SOC of interest divided by the SOC gas-phase concentration. This gas-particle partition coefficient, which has units of m3 p.g 1, can be calculated from the following:

In Eq. (MM), F (from /liter-associated material) is the concentration of the SOC in air that is in the particle phase, in ng m 3, A (from the use of an adsorbent to collect the gas) is the gas-phase concentration in ng m 3, and TSP is the concentration of iotal suspended particles, in ¡xg m 3. (Note that this is the definition in common usage today; however, some earlier papers (e.g., Yamasaki et al., 1982; Pankow, 1987) defined Kp as the inverse, i.e., as equal to ^4(TSP) F.) Because measurements of F, TSP, and A are difficult to make in an artifact-free manner and because equilibrium may not always hold in the atmosphere (see later), the quantity (F TSP) A is often referred to as the measured partition coefficient, in contrast to the true, ther modynamic partioning coefficient, Kp. As shown in Box 9.f, Eq. (MM) can be shown to be equivalent to assuming that adsorption of the gas on the particle surface follows a Langmuir isotherm (Yamasaki et al., 1982; Pankow, 1987).

Pankow (1987) showed that this assumption of a linear Langmuir isotherm is consistent with the pioneering work of Junge (1977) on the adsorption of SOCs on particles. The fraction of the total SOC in the atmosphere adsorbed on aerosol particles, denoted was hypothesized by Junge to be related to the surface area (>ST, cm2 per cm3 of air) of the TSP and the saturation vapor pressure of the SOC by cf> cST (pL cST). (TT)

in this case pL is the saturation vapor pressure of the adsorbing gas at that temperature and c is a constant characteristic of the compound and the temperature.

As discussed in detail by Pankow (1987) and Pankow and Bidleman (1992), the Junge approach can be reduced to an expression of the form

where Ns is the number of moles of surface adsorption sites per cm2, ^4TSI. is the specific surface area (in cm2 ¡xg ') of the TSP, A Hd is the enthalpy of desorption for the SOC directly from the surface, A//vap is the enthalpy of vaporization of the (subcooled) liquid, T is the temperature (in K), and R is the gas constant. In many cases, the compound is a solid at that temperature, rather than a liquid, in which case pL (Torr) is the vapor pressure of the subcooled liquid; the reason for choosing the pressure of the subcooled liquid is the assumption that the adsorbed (or absorbed; see later) states resemble disordered liquids more than highly ordered crystalline solids (Pankow, 1994b).

Thus the relationship between log Kp and log pL should be of the form logiCp mr log pL br, (VV)

where mY and bY are constants. Using a somewhat different approach, Mackay et al. (1986) predicted a similar relationship for partitioning of semivolatile compounds in the atmosphere. From Eq. (UU), mr should be equal to 1 and bT should be equal to \og{(NsAtsi,T 16)exp(AHd A//vap) RT}.

Values of pL for a series of PAH determined experimentally at 25°C by Yamasaki et al. (1984), and subsequently temperature-corrected to 20°C by Pankow and Bidleman (1992), are given in Table 9.24, along with the values for the parameters c and d which describe

Was this article helpful?

0 0

Post a comment