CONDENSATION SAMPLING OF SOLUBLE TRACE GASES IN THE ATMOSPHERE.

摘要

The potential of a condensate method for sampling atmospheric trace gases is investigated. A theoretical approach to the collection of condensate by a cooled surface is presented, followed by an experimental procedure and verification of the method. Heat and mass transfer problems are solved to give distributions of temperature, velocity and mass concentration. From these distributions, collection rates for water vapor, trace gases, and particulates are derived. It is shown that the collection of particulates is negligible in most cases when compared to the collection of gases in the condensate; highly soluble gases are sampled with the greatest efficiency. Experimentally, a plate, cooled to below the ambient dew point temperature, collected condensate which was subsequently analyzed for various components. Atmospheric concentrations were determined from the corresponding species in the condensate. Several trace gases have been detected by the condensate method. Concentrations of two of these, ammonia and nitric acid, are compared with filter methods, with reasonably good agreement. Comparison is also made with concentrations of ammonia, formaldehyde, nitric acid and nitrous acid found by others under similar conditions. Typical daily variations of trace components in more remote areas surrounding Tucson, Arizona, are presented in an attempt to estimate background Southwest U.S.A. levels; these are as follows: 1.0 ppb for ammonia, 0.5 ppb for formaldehyde, 0.1 ppb for nitrous acid, 0.4 ppb for nitric acid, 1.1 ppb for formic acid, 0.9 ppb for acetic acid, 0.3 ppb for sulfur dioxide, 0.3 ppb for hydrochloric acid and 1 ppb for hydrogen peroxide. The diurnal variation of these gases and relationship between them is discussed. Possible sources of error in the condensate method of sampling are considered. The collection rate of water vapor is measured and compared to that theoretically predicted; explanations are presented for the discrepancy between the two. The sensitivity to errors in the measurement of the collection temperature and dew point temperature is determined.