GAS PARTICLE PARTITIONING OF SEMIVOLATILE ORGANIC-COMPOUNDS TO MODEL ATMOSPHERIC PARTICULATE MATERIALS .1. SORPTION TO GRAPHITE, SODIUM-CHLORIDE, ALUMINA, AND SILICA PARTICLES UNDER LOW HUMIDITY CONDITIONS

The partitioning of seven polycyclic aromatic hydrocarbons (PAHs) to four model atmospheric particulate materials (APM) in dry nitrogen is examined. Literature data involving the development of a photoelectric aerosol sensor for the quantitation of particulate PAHs are reanalysed in an atmospheric gas-particle partitioning context. The PAHs include chrysene and benz[a]anthracene, three benzofluoranthenes, benzo[a]pyrene, and benzo[e]pyrene. The model sorbents include graphitic carbon, sodium chloride, silica, and alumina. The measured partition coefficients are expressed in both a K(p) (m3-mu-g-1) and a K(p,s) (m3 m-2) format, and are compared to corresponding values for urban particulate material (UPM) at ambient relative humidity (r.h). The partition coefficient K(p) is the ratio of the sorbed concentration (ng-mu-g-1) to the gaseous concentration (ng m-3). The partition coefficient K(p,s) is the ratio of the surface-area normalized sorbed concentration (ng m-2) to the gaseous concentration (ng m-3). Estimates of the compound-dependent heats of desorption from the surfaces (Q1, kJ mol-1) are determined from the slopes of plots of measured values of log K(p) (or log K(p,s)) vs 1/T. For graphitic carbon, a surface that might have been expected to sorb compounds in a non-specific manner, the agreement between the measured log K(p,s) vs 1/T lines and extrapolated UPM lines is very reasonable. The agreement was not as good for the other three sorbents. Silica and alumina exhibited significantly stronger sorption than UPM on a K(p,s) basis. However, the Q1 estimates determined for silica and alumina were not found to be as high as might have been expected under these circumstances; the quality of the Q1 data is therefore in question. The general agreement between the log K(p,s) vs 1/T plots for graphitic carbon in dry nitrogen with those for UPM at ambient r.h. adds further evidence to the argument that partitioning to UPM is adsorptive, and non-specific in nature.