Integrated Laboratory and Field Characterization
of Organic Carbon in PM2.5 Formed through

Chemical Reactions

John Offenberg1, Tadeusz Kleindienst1, Edward Edney1, Michael Lewandowski1, Mohammed Jaoui2

!U.S EPA/ORD/NERL'HEASD. 109 T.W. Alexander Dr., RTP, NC27711; 5 Alion Science and Technology, P.O. Box 12313, RTP, NC 27709

Abstract

An integrated laboratory and field research program is underway at the
National Exposure Research Laboratory (NERL) to characterize
organic carbon in PM25 (particulate matter) formed through chemical
reactions. Information from this study will provide critical data needed
to improve the treatment of secondary organic aerosol (SOA) formation
in the Community Multiscale Air Quality (CMAQ) model. In the
laboratory portion, SOA-producing hydrocarbon precursors are
irradiated in a photochemical reaction chamber in the presence of
nitrogen oxide (NOx) and sulfur dioxide (S02). In these reactions,
identifiable organic compounds indicative of the precursor gases are
formed. These same compounds are then measured in field study
samples to ensure that relevant chemical systems are being studied. In
collaboration with the California Institute of Technology and the
University of Antwerp, analytical methods and instruments are used to
identify the products. Efforts are also underway with the Atmospheric
Modeling Division (AMD) of NERL to incorporate findings from the
field and laboratory measurements to improve the treatment of SOA
within CMAQ. The project results should provide the Office of Air
Quality Planning and Standards with critical data on important
regulatory issues, among them (1) contributions of each SOA precursor
to the PM2.5 concentration, (2) relative contributions of anthropogenic
and biogenic hydrocarbons to ambient SOA concentrations, and (3)
impacts of S02 reductions on SOA formation. This information will
improve the treatment of SOA in the CMAQ model and help states
evaluate control strategies for reducing ambient PM25. These results
will support effective regulations and provide information that improves
public health and reduces ecological impacts.

Goals and Objectives

Approach for Studies

• Conduct field studies to measure the organic fraction of
ambient PM2 5. Identify tracer compounds in the ambient
samples, such as those that would be collected from the
atmosphere depicted above.

• Conduct laboratory experiments to identify reaction systems
responsible for the observed tracer compounds. Use the
NERL smog chamber to generate these atmospheres.
Establish reaction mechanisms for SOA formation.

• Conduct modeling studies to predict the formation and
partitioning of SOA within PM2 5.

• Collaborate with the AMD in NERL to incorporate the
findings in CMAQ.

Schematic of Approach for Implementation Studies

Smog Chamber

Results and Conclusions

S02



Outputs

SOA Tracers
Effects of Pollutants

Outcomes
Atmospheric Models

Types of Experiments Conducted

Lab Irradiation Experiments

Field Studies

Toluene/NOx/S02

RTP, NC 2000 summer

a-Pinene/NOx /Air + S02

RTP, NC 2003

/3-Pinene/NOx/Air

Baltimore, MD2001 summer

d-1 i m on e n e/NOx/Ai r

Philadelphia, PA 2001 summer

lsoprene/NOx/Air + SO?

New York City, NY 2001 summer

Toluene/a-Pinene/NOx + S02

Detroit, Ml 2004 summer

a- Pi n en e//3- P i n e n e/d- li mon en e/N Oy



lsoprene/a-Pinene/NOy



Isoprene/a-Pinene/Toluene/NOx + S02



•	Identify the major SOA precursors important in PM25

•	Identify tracer compounds for the major SOA precursors

•	Determine reaction mechanisms for SOA formation

•	Aid AMD in NERL to improve treatment of SOA in CMAQ

•	Use the NERL smog chamber to generate atmospherically
relevant air mixtures for exposure studies

Experimental Methods

•	Irradiate individual aromatic and biogenic hydrocarbons in the
presence of NOx and S02 in the NERL smog chamber and
measure the SOA masses produced by the photooxidations.

•	Analyze chamber SOA samples using LC/MS, derivative-based
GC/MS, Ion Trap MS, and MALDI methods to identify SOA tracer
compounds.

•	Compare chamber composition and concentration data with model
predictions whose formation mechanisms include contributions
from gas-aerosol partitioning, acid catalyzed reactions, and
polymer formation, and others.

•	Assess whether tracer concentrations can be used to determine
contributions of SOA precursors to ambient PM2 5.

2003 RTP PM2.5 Concentrations Data

•	Smog chamber irradiations of biogenic hydrocarbons (emitted from
trees and other vegetation) and aromatic hydrocarbons (emitted
mainly from cars) show that these compound can be converted to
SOA by chemical reactions.

•	For compounds such as isoprene and a-pinene, the addition of
S02 increases the amount of SOA formed above that obtained in
the absence of S02.

•	Laboratory results suggest that several chemical processes must
be included in a model to explain SOA formation. The types of
processes that appear to be important include (1) exchange of
organic compounds between the gas and particle phases, often
referred to as partitioning, (2) acid catalyzed reactions within the
particle, (3) polymer formation, and possibly (4) cloud water
reactions could be contributing to SOA formation.

•	Toluene, a-pinene, and isoprene SOA tracer compounds detected
in ambient PM2 5 samples collected in the eastern USA indicate
these emitted hydrocarbons are contributing to SOA. Analysis of
field data suggests that SOA in the summer is significant, but
decreases considerably in the colder seasons.

•	The figure below shows how the concentrations of the tracer
compounds change with season. The isoprene tracer is only seen
in the summertime, while the tracer compound for a-pinene is
detected in the spring, summer, and fall. Levoglucosan, a primary
product from wood combustion, is detected throughout the year but
mainly during the winter and spring.

Outputs, Outcomes and Future Directions

•	Continue comparing chamber concentrations and compositions of
SOA formed with atmospherically relevant individual and mixtures
of hydrocarbons irradiated in the presence of NOx and SO2 with
model results for proposed SOA formation mechanisms.

•	Assess whether SOA yields in complex hydrocarbon mixtures are
additive.

•	Work with AMD modelers to develop the CMAQ version of the PM
chemistry model.

•	Results of the laboratory and field studies will used by AMD in the
CMAQ model that will be available to the RPOs for State
Implementation Plan modeling studies.

•	Some of the laboratory methods, developed under this program,
will be used in EPA-NOAA collaborative research to assess the
impact of N2O5 reactions on PM2.5 nitrate levels.

150 200 250

Julian Day 2003

Disclaimer: Although this work was reviewed by EPA and approved
for publication, it may not necessarily reflect official Agency policy.

epascienceforum

Your Health • Your Environment • Your Future


-------