5.
Comparison of Aerosol Carbon Instruments at a Remote, Background
Site
Roger
L. Tanner and Solomon T. Bairai
Air,
Land, and Water Sciences Department, Tennessee Valley Authority,
Muscle Shoals, AL 35662
Major
problems still remain in the measurement of carbonaceous materials
in ambient aerosols. Specifically, the use of speciation techniques
and organic tracers to identify the sources of primary and secondary
organic aerosols in the atmosphere is hampered by the ambiguities
in aerosol organic carbon measurement. TVA is operating four speciation
instruments at the Look Rock, TN, site as part of the VISTAS RPO
focus site sampling. Instruments include an R&P Model 5400
Aerosol Carbon Analyzer and a Magee Scientific Aethalometer. For
a period of about 5 weeks in early fall, 2003, TVA also operated
a collocated Sunset Labs OC-EC analyzer at this site. Both carbon
instruments were operated on a 1-hour sampling protocol, and the
5-min Aethalometer data were averaged over 1-hr periods for comparison..
Operation of the Model 5400 instrument used the normal protocol
of dual collectors with analysis of 1-hr collected aerosol samples
the following hour with two temperature thresholds, 275ºC
for organic carbon and 750ºC for total carbon, while the
second collector was used for sampling. Quantification of CO2
from combustion of the samples uses a Licor Model 800 sensor.
The operation of the Sunset Labs OC-EC instrument uses a single
in-line quartz collector with a 50-min sample collection period,
followed by a 10-min analysis period with the temperature of the
sample ramped at about 10ºC/sec to 650ºC, and about
6ºC/sec for the 650-850ºC ramps (organic and elemental).
To deal with the adsorbed VOC problem, an in-line denuder is used
with the OC-EC instrument, and the Model 5400 is operated with
metal collectors, but no in-line VOC denuder. An optical BC-type
signal is obtained from the OC-EC instrument as well as a TOT-like
signal for elemental carbon.
A
comparison of the data indicates that the organic carbon data
are systematically higher than for the Model 5400 using a 275ºC
temperature plateau. The optical EC data from the OCEC instrument
agrees well with Aethalometer BC data, even though the BC levels
at the Look Rock site were frequently at or below the Aethalometer
limits of detection. Ambiguity still exists over the OC quantification
with the instruments tested in this study, although there are
preliminary indications that, when positive and negative artifact
are accounted for, the difference between the Sunset Labs total
carbon quantity and the Aethalometer BC value likely represents
the current best estimate of organic carbon
|