Environmental Engineering Reference
In-Depth Information
It has long been recognized that certain primary compounds were semivolatile.
Polycyclic Aromatic Hydrocarbons (PAHs) have received special scrutiny because
of their known toxicity; they range from completely volatile Napthalene to completely
condensed Coronene, with intermediate molecular weight compounds occurring
significantly in both phases. The semi-volatile character of many other ambient
organics (Fraser et al., 1997, 1998) and sources (Schauer et al., 1999) has also
been addressed extensively. However, the overall contribution of semivolatile
material to ambient POA has not been comprehensively addressed until recently.
For more than 20 years, dilution samplers have been used to measure POA
emission factors. Development of these samplers was motivated by the semivolatile
character of primary emissions; specifically upon heating. Initial studies focused
on the effect of temperature on partitioning. Only modest amounts of dilution
(about 100×) are required to reduce the temperature of the exhaust to ambient
levels, which dramatically increases the amount of POA because of the strong
dependence of vapor pressure on temperature. More recent experiments have
demonstrated that POA emission factors decrease when the aerosol is isothermally
diluted (Lipsky et al., 2006). Both of these changes are expected based on partitioning
theory.
Although some primary emissions are clearly semivolatile, most CTMs have
treated them as non-volatile. The implicit assumption is that the partitioning measured
using a dilution sampler is representative over the full range of atmospheric conditions
simulated by the model, and that the semi-volatile primary mass is a small fraction
of the total POA.
It is also reasonable to expect that the organic aerosol compounds will react
with atmospheric oxidants like the OH radical, O
3
, NO
3
, etc. leading to chemical
changes. Limited studies of the heterogeneous chemistry of POA components and
realistic POA have confirmed that these reactions are quite efficient in transforming
POA to other compounds. At the same time, the POA components that evaporate
after dilution as they move away from their sources will react in the gas-phase,
forming products with lower volatility that can condense back in the particulate
phase (Robinson et al., 2007). Both of these pathways oxidize primary emissions,
forming oxygenated organic aerosol (OOA).
At this point the traditional simple framework and corresponding definitions of
POA and SOA have broken down. Is the OA formed from a compound that starts
its atmospheric life in the particulate phase, evaporates, reacts in the gas phase and
then condenses primary or secondary? Further confusion arises because traditionally
'primary' and 'secondary' refer to the aerosol mass, and not the specific chemical
compounds making up that mass.
2. The Volatility Basis Set as a Unifying Framework
The need to treat the volatility of primary OA, the formation of secondary organic
aerosol (SOA), the reactions of both primary and secondary OA components, and
