|
Principal Investigator: B. R. Appel
Department of Health Services, California
January 1987
ARB Contract No. A4-158-32 (Full Report)
ABSTRACT
A field study was performed evaluating methods for the measurement of carbonaceous atmospheric aerosols. The
sampling techniques assessed included a proposed fine particulate carbon sampler operating for 12-hour sampling
periods and two tandem filter samplers operating at face velocities of 47 and 10 cm/set, respectively, for four-
or eight-hour periods. Samples were analyzed for elemental or black carbon with a light absorption technique, and
for total carbon and carbonate carbon with a coulometric detector. Organic carbon was estimated by difference.
Samplers were operated at Citrus College, Glendora, California during the period August 12 - August 20, 1986 in
parallel with samplers operated by other investigators.
The proposed fine particulate carbon sampler consisted of a cyclone, parallel plate diffusion denuder packed with
coarse, activated alumina, and a quartz fiber filter followed by a fluidized bed of activated alumina. The denuder
was intended to prevent sorption of vapor-phase carbonaceous material on the quartz filter, particulate matter
thereon, or the fluidized bed. The fluidized bed was intended to trap carbonaceous material lost from the
filter by volatilization following collection in the particulate state.
Based on QA studies as well as atmospheric particulate sample results, the alumina denuder was judged ineffective
in eliminating the positive error in organic aerosol sampling with quartz filters. The results indicate that
compounds not retained efficiently on alumina are being sorbed by quartz filters to produce a positive artifact
in carbonaceous aerosol sampling. The contrast in denuder effectiveness between the present trials and our earlier
studies is discussed. The failure of the denuder in the present work also prevents interpretation of carbon
recoveries on the fluidized bed of Al2O3 regarding the significance of volatilization of particulate carbon
from a quartz filter sampler.
Comparing particulate organic carbon results as a function of prefilter face velocity, measured airborne concentrations
were about 50% higher at 10 compared to 47 cm/set. With tandem filter samplers, after-filter C was about 20% of
the total recovered carbon at both 10 and 47 cm/set. Expressed in µg/m3, after-filter results at 10 cm/set
were about twice the level at 47 cm/set, consistent with greater efficiency for sorption of vapor-phase organic
compounds at lower face velocity. Subtracting the after-filter carbon results from the organic carbon (i.e. particle
plus sorbed vapor phase carbon) measured on the prefilter did not greatly improve the agreement in organic
carbon concentrations measured with the prefilters of the samplers at the two face velocities. Thus, neither measurement
of particulate organic carbon can be considered accurate.
AIHL carbon analyses of atmospheric particulate samples were compared to those by S. Cadle, General Motors Research
Laboratory. The total carbon and organic carbon data sets for the two laboratories were highly correlated (r>0.94).
AIHL total carbon results averaged 4% lower; organic carbon averaged 15% higher, and elemental carbon averaged
17% lower than those by GM.
Back to the Main Menu
|
