|
Main Topic & Follow Up Abstracts
(Click on name to link to the text, or scroll
down.)
- Robert Cary, President,
Sunset Laboratory
- Lloyd Currie, Emeritus
Fellow, National Institute of Standards and Technology
- Judith Chow, Research Professor,
Environmental Analysis Facility, Desert Research Institute
- Hélène Cachier,
France
- Joellen Lewtas, Senior Research Scientist,
US EPA/Office of Research & Dev’t/Nat’l Exposure
Research Lab
- Kirk Fuller, Research
Scientist, National Space Science and Technology Center, Univ.
of Alabama
- Hans Hansson, Air Pollution
Lab, Inst of Applied Environmental Research and Dept of Meteorology,
Stockholm University, Sweden
- Hans Moosmuller,
Research Professor, Desert Research Institute
TOPIC #1 . . .
What is elemental carbon and how do definitions differ
for different applications? What are the OC and EC properties
that are of importance to human health, visibility, climate, and
source attribution? To what extent can a single analytical method
or protocol meet these different needs?
Robert Cary, President, Sunset Laboratory, Inc.
Abstract
Compared with all the other elements, carbon
has the ability to form an incredible number of different compounds
with a likewise incredible number of characteristics. This is
due to its unique bonding ability. When considering the pure element
itself, this bonding allows for two unique substances. Diamond,
which can be considered as the ultimate polycyclic aliphatic-bonded
form of carbon; and graphitic carbon, which can be considered
the ultimate polycyclic aromatic-bonded form. The later, unfortunately,
seems to be the one found commonly as a component of atmospheric
aerosols, and is the one of interest here when referring to elemental
carbon, EC.
There are a number of physical and chemical properties that define
this graphitic aerosol carbon. And there are a number of applications
where some of these properties are important and others not so
important.
From a chemist point of few, elemental carbon consists of an
extended aromatic-ring structure where the carbon atoms are bonded
by sp2 -bonds and delocalized pi-bonds. As the number of rings
increase, these pi-bonding electron orbitals become degenerate
across the entire structure and end up existing at a level with
little or no gap between the valence bands and conduction bands.
This results in elemental carbon being a fairly good conductor
of electricity when influenced by an electric field and contributes
to many of its chemical and physical properties.
The strengths of these bonds cause elemental carbon to be very
thermally refractory, fairly chemically inert at temperatures
less than several hundred degrees and insoluble in solvents.
From a physical-optics point of few, elemental carbon interacts
with electro-magnetic radiation as a result of these conducting
electrons very similarly to metals. In its bulk form as the mineral
graphite, it in fact appears very “metallic”. As small
aerosol particles, this interaction can be quantitated by its
index of refraction consisting of a real part and imaginary part,
which are measures of the ability to scatter and absorb electro-magnetic
radiation. The values of this index of refraction are a function
of not only the wavelength, but also the geometric “size”
and morphology.
Since there are a fair number of electrons in the conductance
bands with no energy gaps, there are no unique absorption frequencies
and thus elemental carbon can absorb energy throughout the visible
spectrum. This is what causes it to appear “black”
as very small particles. Many applications such as climate and
visibility studies are concerned with this interaction of EC with
electro-magnetic energy.
Many applications are more interested in the chemical properties.
Its inertness is one reason EC is used as a tracer for certain
sources. Nearly all processes that form EC do so in a way that
results in at least some occluded organics, and in many cases
the EC is a small fraction of the aerosol with OC being the major
mass fraction. However, when produced as the major mass fraction,
such as in diesel engines, the large surface area causes EC to
act as a very efficient adsorber of many compounds that diffuse
to its surface. Thus, either alone or as a carrier of other compounds,
there is a great deal of interest in its health effects. And it
is this surface that is of interest as a location for other chemical
process.
Most efforts at quantitating EC are focused on its chemical inertness,
thermal stability or optical properties. The difficulty of properly
speciating EC from other forms of carbon, such as organics, arises
for a number of reasons. And the proper speciation becomes most
difficult when the EC fraction is small compared to the other
forms. Methods that depend upon the insolubility or chemical inertness
of EC must contend with the fact that there are forms of OC that
are also insoluble or inert. Optical methods must deal with optical
properties of other particles as well as the morphology and optical
properties of the EC particle itself. And thermal methods must
deal with the likely possibility that some organic forms will
be pyrolytically transformed into EC. And this is complicated
by the fact that the EC fraction of the aerosol itself was originally
produced by a pyrolitic process.
Topic 2 . . .
What options exist for fundamental and traceable OC and
EC standards?
Lloyd Currie, Emeritus Fellow, National Institute
of Standards and Technology, Gaithersburg, Maryland
Click
Here for Dr. Currie's "draft reference
material document" (in Microsoft Word format),
in which he details his discussion on Topic 2.
Follow Up Abstract
On the Distribution of the Blank
L.A. Currie and J.M. Conny (NIST)
The distribution of the blank sets the ultimate limitation
for low-level measurement. Its dispersion strongly influences
the uncertainty of the net signal, and often it is the prime
factor governing detection and quantification capabilities.
If the blank distribution shows a positive skewness, it is strongly
recommended that observations (sample, blank) be paired. The
influence on the validity of detection decisions will be examined
with examples from analytical blanks for sulfur (TIMS), 14C
(AMS), OC (TOT), and EC (TOT).
Follow Up Abstract (Mr.
Klouda will not be attending, but requests inclusion in the
Research Strategy)
A Filter-Based Fine Particulate Matter Reference Material
(RM) 2784 [1]
George A. Klouda*, Helen J. Parish†, James J.
Filliben* and Judith C. Chow‡
* National Institute of Standards and Technology, Gaithersburg,
MD 20899
† SRI International, Menlo Park, CA 94025
‡ Desert Research Institute, Reno, NV 89512
To support national monitoring of fine particulate matter (PM
2.5) concentrations of organic carbon (OC), elemental carbon
(EC) or black carbon, and total carbon (TC), NIST, in collaboration
with SRI International and the Desert Research Institute, has
produced a Filter-Based Fine Particulate Matter Reference Material
(RM) 2784 using the Washington, D.C. Urban Dust Standard Reference
Material (SRM) 1649a. SRM 1649a was resuspended in a purified
airflow and size-segregated by a series of processes - exposure
to air at sonic velocity, sedimentation and impaction –
designed to pass only particles less than 2.5 mm aerodynamic
diameter. This fine aerosol fraction was then collected on 320
quartz-fiber filters, simultaneously, under matched flow conditions.
A total of 7 batches were obtained to complete production of
2240 filters [2]. The mass of fine PM on each filter was determined
by gravimetric analysis before and after the resuspension and
collection process. Process blanks were prepared for possible
chemical analyses and weighing controls were obtained to estimate
potential bias in the gravimetric measurement.
Each filter has a set of characteristics that provides a unique
identification: filter identification number, PM-mass loading,
batch, chamber, and chamber-position. The distribution of PM-mass
loadings with respect to chamber-position, chamber and batch
appears quite variable. After nine outliers with negative loadings
are removed, the average mass and standard deviation for the
remaining 2231 filters is 1064 ± 414 mg dust per filter
(39% relative standard deviation) with a range of 2763 mg. Gravimetric
analysis of control filters treated similarly as the loaded
filters but without exposure to the dust shows that on average
the difference between the initial weighing and the final weighing
is 46 ± 63 mg (standard deviation, n = 140).
An inter-laboratory and -methods comparison is underway to
obtain value assignments [3] and uncertainties [4, 5] for EC,
OC and TC concentrations (in g/g dust) and for the EC/TC ratio
of RM 2784 filters using two specific thermal-optical analysis
(TOA) methods. The NIOSH (National Institute of Occupational
Safety and Health) Method 5040 [6, 7], with a slightly modified
temperature profile for the Environmental Protection Agency
(EPA) PM 2.5 Chemical Speciation Trends Network (STN) [8], will
be applied to thermal-optical transmission (TOT) instruments.
The IMPROVE (Interagency Monitoring of Protected Visual Environments)
Method will use thermal-optical reflectance (TOR) instruments
[9].
Beyond providing value assignments of the analytes identified
above for RM 2784, the results of the intercomparison are expected
to show whether a mass-interval effect exists for the mass fraction
of each analyte defined by the method and the EC/TC ratio, what
is the within-filter homogeneity, and what is the filter-to-filter
homogeneity. A secondary analysis of the data will examine factors
that may have influenced the production of these filters, e.g.,
batch, sampling chamber, and chamber location, i.e. quadrant,
column and row that are unique to each filter.
[1] ISO, “International Vocabulary
of Basic and General Terms in Metrology (VIM),” 2nd
Edition; BIPM/IEC/IFCC/ISO/IUPAC/IUPAP/OIML, International
Organization for
Standardization (ISO), 1993.
[2] Watson, E.L., Irwin, K., and Parish,
H.J., “ Preparation of SRM 2784 on quartz filters,”
SRI International, Final Report, August 2002.
[3] May, W., Parris, R., Fassett, J., Greenberg,
R., Guenther, F., Wise, S., Gills, T., Colbert, J., Gettings,
R. and MacDonald, B., “Definitions of Terms and Modes
Used at NIST for Value- Assignment of Reference Materials
for Chemical Measurements,” Natl. Inst. Stand. Technol.,
Special Publication 260-136, January 2000.
[4] ISO, “Guide to the Expression of
Uncertainty of Measurement: First edition 1993; ISBN 92-67-
10188-9; International Organization for Standardization (ISO),
1993.
[5] Taylor, B.N., and Kuyatt, C.E., “Guidelines
for Evaluating and Expressing the Uncertainty of NIST Measurement
Results, Natl. Inst. Stand. Technol., Tech. Note 1297, (1994).
[6] Birch, M.E. and Cary, R.A., “Elemental
carbon-based method for monitoring occupational exposures
to particulate diesel exhaust,” Aerosol Sci. Technol.,
Vol. 25, pp. 221-241 (1996).
[7] National Institute of Occupational Safety
and Health (NIOSH), “Elemental Carbon (Diesel Particulate):
Method 5040,” NIOSH Manual of Analytical Methods, 4th
ed. (http://www.cdc.gov/niosh/nmam/pdfs/5040f3.pdf), Cincinnati,
Ohio, 15 January 1997
[8] EPA Strategic Plan, “Development
of the Particulate Matter (PM 2.5) Quality System for the
Chemical Speciation Monitoring Trends Sites,” Environmental
Protection Agency, Office of Air Quality Planning and Standards,
Research Triangle Park, North Carolina (www.epa.gov/ttn/amtic/files/ambient/pm25/spec/strate1.pdf).
[9] Chow, J.C., Watson, J.G., Pritchett, L.C.,
Pierson, W.R., Frazier, C.A., and Purcell, R.G., “The
DRI thermal/optical reflectance carbon analysis system: description,
evaluation and applications in U.S. air quality studies,”
Atmospheric Environment, Vol. 27A, No. 8, pp. 1185-1201 (1993).
For more information, contact: George A.
Klouda, National Institute of Standards and Technology,
100 Bureau Dr., Mailstop 8375, Gaithersburg, MD 20899-8375;
301-975-3931; (Fax: 301-417-1321); george.klouda@nist.gov
Return to Top of Page
TOPIC #3 . . .
How does the sample affect the measurement
of different carbon fractions? How do properties of particles
on a filter differ from those in ambient air? How do
different compounds react with heat and among themselves to
create pyrolized carbon? How do different filter loadings affect
optical measures of pyrolysis? Under what conditions might other
carbon-containing components (e.g., carbonates) be detected
as OC or EC? What additional information should be reported
with OC and EC values to evaluate the precision and validity
of an OC/EC split?
Judith Chow, Research Professor, Environmental Analysis Facility,
Desert Research Institute
Click
Here to view Dr. Chow's abstract in Microsoft Word
Follow Up Presentation Abstract
1
Effects of iron oxides on the determination of organic
and elemental carbon using thermal optical techniques
Kochy Fung, AtmAA, Inc. Calabasas, CA, U.S.A.
Ambient particulate samples are routinely analyzed for organic
and elemental carbon (OC/EC) using either thermal volatilization-pyrolysis
correction methods, as in Interagency Monitoring of PROtected
Visual Environments (IMPROVE) with correction by reflectance,
or in the National Institute of Occupational Safety and Health
(NIOSH) Method 5040 using transmittance (TOT). In both methods,
the filter is heated stepwise with specified temperatures,
rates of temperature increase, and residence times at each
temperature, first in He to volatilize the OC, then in O2/He
for EC combustion. The resulting CO2 is converted to methane
for detection by a flame ionization detector. A recent comparison
study (Chow et al, 2001) found that NIOSH EC was 107 + 77.5%
lower than IMPROVE.
The difference was thought to be relating to the oxidation
by metal oxides under the NIOSH 900oC-temperature step for
OC. It is already known that MnO2 can oxidize EC at temperatures
higher than 525oC. To study if other metal oxides bear similar
property, soot samples containing iron oxides were created
from combusting ethylene with iron pentacarbonyl vapor. Acetylene
gas was added for soot formation. By varying the level of
acetylene in the combustion mixture, different amount of soot
to iron content can be obtained. Samples with only soot particles,
iron oxides, and a mixture of both were analyzed according
to the IMPROVE and NIOSH protocols.
The results of the analyses indicated that iron oxides attenuated
the laser signals in both the OC and EC phase of the thermal
optical methods. During the OC phase of the NIOSH protocol
(helium atmosphere) EC or other light absorbing carbon was
oxidized or volatilized in the samples containing iron oxides
when the temperature reached 800oC and above. The orange-colored
residue that remained on the filter disc after the IMPROVE
analysis was not visible on NIOSH-analyzed disc. Thus the
presence of significant amount of iron oxides in a sample
can increase the apparent OC content due to their affects
on increasing pyrolysis correction, as well as oxidation of
EC during the OC phase of the NIOSH protocol.
______________________
Chow, J.C.; Watson, J.G.; Crow, D.; Lowenthal, D.H.; Merrifield,
T. Comparison of IMPROVE and NIOSH carbon measurements; Aerosol
Sci. Technol. 2001, 34(1), 23-34.
Follow-up Presentation Abstract
2
Measurement of Carbonate
Minerals in Aerosol Samples
Johann Engelbrecht, Associate Research Professor, Desert
Research Institute
Carbonate minerals such as calcite, dolomite and magnesite
commonly occur in aerosols from arid regions of the western
USA. Screened as well as PM10 and PM2.5 filter samples of
ten re-suspended soils were analyzed for their mineral contents
by X-ray Diffraction (XRD). Mineral species identified in
these samples include quartz, calcite, mica, feldspar, dolomite,
rutile, gypsum, and clay minerals. These samples were also
chemically analyzed for their elements, ions, OC and EC, as
well as carbonate contents. The mineralogical and chemical
results are compared, with emphasis on the measurement of
carbonate minerals.
Return to Top of Page
TOPIC #4 . . .What
are the important parameters that need to be defined for a carbonaceous
aerosol analysis and how should these be documented for different
analysis protocols?
Hélène Cachier, LSCE/CFR, laboratoire mixte CEA-CNRS,
France
Abstract
Intercomparisons have shown for a long time important discrepancies
among EC/OC ratio results produced with different analytical protocols.
Only very recently have efforts been deployed to understand the
reasons. Although some answers clearly reveal which things to
avoid, answers for what to do are still pending.
Keeping in mind firstly, the chemical complexity and the diversity
of carbonaceous aerosols and, secondly, the fact that there is
no clear separation between high molecular weight OC and EC, the
“thermal” definition of EC is obviously operationally-dependent.
The aim is however to uniformize results and ultimately to reconcile
both thermal and optical definitions of BC and EC.
1) Indeed thermal-optical analyses rely on both properties
of BC/EC, the light absorption and the refractory behaviour.
This is an additional argument to have only one name for this
component: BC, whereas EC can induce some conceptual errors
in the chemical nature of the particles.
2) There are two main classes of analyses:
- analyses under oxidative atmosphere, producing a gradual
combustion (oxidation) of the carbonaceous particles;
- 2-step analyses with the first step under an inert gas
(helium or nitrogen), producing a pyrolysis of the OC component.
Applying such thermal treatments modifies the particles differently.
This is particularly crucial for EC particles. In conclusion,
any mix of both treatments in the analytical protocol must be
considered with caution.
3) The analysis by itself brings artefacts: charring of OC
and untimely departure of BC.
In all cases, it is better to minimize artefacts than to try
to quantify them. This is primarily due to the fact that the different
fractions of EC have different optical properties (absorption
coefficient) and that their time (temperature) evolution may overlap.
When entering the problem of carbon analysis, some questions
arise:
a) The choice of method: under oxygen/or mixed atmosphere?
- Why choose the second method, which apparently brings more
artefacts?
- Does the EGA/O2 bring a satisfactory split between OC and
BC?
- For this methodology, why has the importance of the plateau
duration (almost) never been discussed?
b) Do we actually need the split of OC and BC into several
fractions?
- Hasn’t it a historical origin?
- Have these fractions any environmental significance?
c) How do dust particles affect the signal (laser signal, evolution
of EC)?
d) How do we treat the problem of carbonates?
- In other words, is it possible to avoid removal of carbonates
prior to thermal analysis? If so, may carbonate peak be clearly
identified?
- Does acidic treatment affect the carbon content and the
split EC/BC?
e) What is the limitation of the analysis in terms of EC concentration
per area unit?
- Indeed, this parameter may affect the gas penetration in
the particle layer, and additionally may affect the laser
sensitivity for pyrolysis correction.
f) If we chose the EGA/He/O2 analysis:
- How to minimize charring; although some fractions char more
than others, it may be seen that charring occurs very early
in the temperature program.
g) Finally, a side problem is the absorption of VOC (or CO2?)
during storing.
- Should we treat the sample prior to analysis?
- How does the cleaning of the filter affect its capability
of positive (passive and active) artefacts?
In light of the above-mentioned considerations, there is a need
to delineate the main determinants which have the potential to
create discrepancies in the determination of EC/TC ratios.
Specific parameters to be defined for EGA/O2 analysis:
- decarbonatation: indeed, due to a matrix effect, carbonates
evolve at lower temperature than assessed for pure material,
and the generally adopted temperature for the evolution of EC
(720 to 750°c) actually creates some decomposition of carbonate.
- The temperature increase rate which is especially critical
between 300°c to 400°c where there is some overlap of
the OC and EC fractions.
- optimum concentration range of the filter loading.
- beginning and end temperature.
Specific parameters to be defined for EGA/He/O2 analysis:
- number of plateaux, duration, temperature.
- should we link the plateau duration to background recovery?
- is there any special procedure in the transition zone? (lower
increase of temperature)
- laser monitoring wavelength, type of optical measurement:
T or R?
- description of flash heating if any.
Finally in the discussion 2 other items will appear:
- the role of the nature (origin, age) of the aerosol will
be introduced which probably leads to the discussion that some
methods are “better” for a given type of aerosol.
- How to analyze cascade impactor samples?
Return to Top of Page
TOPIC #5 . . .
What specific compounds are likely to evolve during different
temperature fractions of thermal evolution methods used to analyze
carbonaceous aerosols?
Joellen
Lewtas, US EPA/Office of Research & Dev’t/Nat’l
Exposure Research Lab
Follow Up Presentation
Abstracts
Organic Carbon Concentration and Composition
in Fine Particulate Matter Collected During ARIES Study
Barbara Zielinska, Desert Research Institute, 2215 Raggio
Parkway, Reno, NV 89512
Fine particulate matter (PM2.5) was measured daily from August
1998 to December 1999 over a 24-hr sampling period in an Atlanta
residential area to provide data for an epidemiological study.
Samples were collected using the DRI Sequential Fine Particulate/Semi-Volatile
Organic Compounds (PSVOC) sampler, which allows unattended collection
of up to four samples. To account for the semi-volatile organic
compounds associated with particles, a quartz filter of 10 cm
diameter was backed by polyurethane foam plugs (PUF) in combination
with the polystyrene-divinylbenzene resin XAD-4.
All quartz filters were analyzed for organic and elemental
carbon (OC/EC) by thermal/optical reflectance method (TOR) prior
to extraction. Filters and PUF/XAD cartridges were extracted
together, using three solvents of different polarity (dichloromethane
followed by acetone, followed by water). The mass of the extracts
was determined by weighing the residue of 20 microliter aliquots
of the extracts deposited on a pre-fired quartz filter punch.
The organic carbon content of the extracts was determined by
subsequent analysis of the punch by the TOR method. Select punches
were also analyzed by gas chromatography/mass spectrometry (GC/MS)
using a thermal desorption method. Thermal desporption was conducted
in a sequential fashion with increasing temperatures that try
to mimic the temperature ramps used with the TOR method (1:
120 ºC; 2: 250 ºC; 3: 380 ºC).
This presentation will discuss the major compounds identified
during sequential desorption analyses of the extracts deposited
on quartz punches. The desorption method can provide some insight
into the composition of the organic fractions observed during
the TOR temperature ramps.
Topic
5 Follow Up Abstract 2—Insights
from Thermal Analysis of Individual Organic Compounds, Mixtures,
Black Carbon Surrogates, Airborne Particulate Matter and Extracts.
L.
A. Gundel, Environmental
Energy Technologies Division, Lawrence Berkeley National Lab
L.A. Gundel, T.W. Kirchstetter, R.L. Dod, Y. Pang, Environmental
Energy Technologies Division, Lawrence Berkeley National
Laboratory, 1 Cyclotron Road, Berkeley, CA 94720; C.S. Claiborn,
Department of Civil and Environmental Engineering, 101 Sloan
Hall, Washington State University, Pullman, WA 99164-2910.
This presentation compares thermograms of organic compounds,
ambient, source and 'surrogate' particulate matter. Here 'surrogate
PM' refers to well-characterized mixtures of fine activated
carbon (AC) particles and known compounds. In addition, thermograms
of extracts and extracted ambient PM are compared to 'reconstructed'
PM (extracts + fine AC particles) and the original ambient PM.
The goal is to shed light on the chemical characteristics of
the temperature-defined carbon fractions in relation to what
is currently known from detailed speciation efforts for organic
compounds in source and ambient PM.
Thermograms of individual compounds suggested that volatility
and MW, rather than class or functional group, controlled the
evolution of OC in the fractions through 250 C. For compounds
that evolved between 250 and 500 C, formation of light absorbing
carbon in situ was more likely for polyfunctional molecules
with at least one aromatic ring. Complex molecules with saturated
rings also pyrolyzed easily.
When adsorbed onto clean fine AC particles, individual compounds
showed thermograms with more complexity: in surrogate PM the
compounds could distribute over more than one temperature fraction,
for example (not counting the EC from AC). However, adding extracts
of ambient PM to fine AC did not necessarily yield the original
ambient PM OC-EC pattern. Reconstructed PM did not contain the
high-molecular weight unextractable OC that was usually associated
with ambient PM.
Return to Top of Page
TOPIC # 6
. . .
How does carbonaceous particle composition, shape, and
size affect optical properties in the air and when sampled on
a filter?
Kirk Fuller, Research Scientist, National Space
Science and Technology Center, Univ. of Alabama
Organic carbon aerosols free of any graphitic (elemental) carbon
may be both spherical and hygroscopic. To the extent that this
is so, their in situ optical properties are amenable to analysis
with Lorenz-Mie theory, with the greatest unknown being the complex
refractive index (i.e., the exact composition) of the particles.
Soot is typically comprised of graphitic carbon mixed with incompletely
oxidized fuel. Rather than being spherical, fresh soot particles
are loose (ramiform), fractal-like agglomerations of graphitic
carbon spherules which may become more tightly packed upon aging.
The spherules are composed of nanometer-sized domains of graphite
crystals. The crystalline structure of the graphite itself is
imperfect, with frequent substitution of carbon atoms in the hexagonal
C6 molecules comprising the crystals, or intercalation of molecules
between the stacked C6 rings. Such variations in structure and
composition leave the complex refractive index of the material
variable, as well. The theoretical analysis of soot optics is
more difficult, but such analysis indicates that aggregation of
the spherules tends to enhance their scattering and absorption
cross sections. The greatest enhancement in absorption cross section,
however, arises from encapsulation of elemental carbon by nonabsorbing
host particles, such as solution droplets of ammonium sulfate.
Absorption coefficients of atmospheric aerosol are routinely
inferred from reflection and/or transmission measurements of particle
deposits on quartz, nuclepore, or Teflon filters. In such procedures,
the filters may be regarded as optical elements which are often
used in concert with integrating spheres, integrating plates,
or both. The technique was originally developed by Lin et al.
(1973), and a number of variations have followed (cf. Rosen et
al., 1978; Sadler et al., 1981; Edwards et al., 1983; Clarke et
al. 1987; Campbell et al., 1995; and Bond et al., 1999). A proper
accounting of scattering in the radiative transfer is a requirement
of all such measurements. Of particular relevance to filter samples,
but much less understood, are the alterations of a particle's
absorption (and scattering) spectrum by its close proximity to
other particles and by its attachment to substrate surfaces.
Absorption by particles in a filter deposit differs from in situ
absorption because of (1) multiple scattering in the deposit/substrate
system, (2) alteration of absorption and scattering cross sections
by electromagnetic coupling between particles, (3) electromagnetic
coupling of particles to filter surfaces, (4) optical coherence
between particles with separations comparable to the wavelength
of the interrogating radiation, (5) induced alignment of nonspherical
particles along filter surfaces, (6) shape distortion of liquid
droplets, and (7) reactions among different chemical species,
especially over extended sampling times.
Raman spectroscopy may provide a useful analysis of soot in that
there is a pair of vibrational signatures present in the spectra
of graphitic carbon that are not active in the infrared. The band
strength of one relates to the quality of the crystal structure,
while the other relates to the size of grain boundaries within
the cryptocrystalline spherules that comprise the soot fractal
(Rosen et al., 1978). Since the degree of molecular ordering in
the spherules is a function of initial combustion conditions and
thermal annealing, such features may be useful in distinguishing
between biomass and fossil fuel soot, as well as in studying the
modification of carbonaceous material during thermo-optic measurements.
If soot initially exists as an inclusion in sulfate aerosol or
the amount of OC coating EC is great enough, then evaporation
of those materials may increase the fractal dimension of the aggregates,
thereby altering its mass-specific absorption cross section. Compositional
changes during heating seem especially difficult to predict, though
it seems that monitoring the infrared (in addition to the Raman)
spectra of the deposits may provide some insight.
The overview by the topic leader will provide a bit more detail
and some examples of the effects described above.
REFERENCES (in addition to suggested reading)
Bond, T. C., T. L. Anderson, and D. Campbell. Calibration and
intercomparison of filter-based measurements of visible light
absorption by aerosol particles. Aerosol Sci.
Tech., 30:582-600, 1999.
Campbell, D., S. Copeland, and T. Cahill. Measurement of aerosol
absorption coefficient from teflon filters using integrating plate
and integrating sphere techniques. Aerosol Sci.
Techno., 22:287-292, 1995.
Clarke, A. D., K. J. Noone, J. Heintzenberg, S. G. Warren, and
D. S. Covert. Aerosol light absorption measurement techniques:
Analysis and intercomparisons. Atmos. Environ., 21:1455-1465,
1987.
Edwards, J. D., J. A. Ogren, R. E. Weis, and R. J. Charlson.
Particulate air pollutants:
A comparison of British “smoke” with optical absorption
coefficient and elemental carbon concentration. Atmos. Environ.,
17:2337-2341, 1983.
Lin, C.-I., M. Baker, and R. J. Charlson. Absorption coefficient
of atmospheric aerosol:
A method for measurement. Appl. Opt., 12:1356-1363, 1973.
Rosen, H., A. D. A. Hansen, L. Gundel, and T. Novakov. Identification
of the optically
absorbing component in urban aerosol. Appl. Opt., 17:3859-3861,
1978.
Sadler, M., R. J. Charlson, H. Rosen, and T. Novakov. An intercomparison
of the integrating plate and the laser transmission methods for
determination of aerosol absorption coefficients. Atmos. Environ.,
15:1265-1268, 1981.
TOPIC #
7 . . .
How might current analysis methods
be enhanced or combined to obtain more information about the nature
of OC, EC, and other carbon fractions in filter samples?
What can be done with existing analysis methods and samples? What
might be provided by collocated measurements? What hardware and
software changes would permit more of the commonly applied protocols
to be applied with the same analytical instruments?
Hans Hansson, Air Pollution Laboratory, Institute
of Applied Environmental Research and Department of Meteorology,
Stockholm University, Sweden
Abstract
Interest in carbon compounds in the atmospheric
aerosol is not mainly from the point of view of basic science,
but rather for their potentially large influence on climate and
health. The climate influence is especially due to the absorbing
properties of what is commonly called soot, Elemental Carbon or
Black Carbon. This has been noted in the recently published results
of the large field experiment INDOEX. The air pollution plume
originating over India and flowing south over the India Ocean,
has shown indications of increasing climate forcing by 10 W/m2
in the air above ground, due to the high content of light absorbing
material.
Cloud drop formation depends on the size and chemical composition
of the particles on which the droplets are formed. The amount
of inorganic salts has been considered to totally determine whether
or not a particle will nucleate (activate) as a cloud droplet,
as described through the original Köhler equations. However
recent theoretical investigations, where the Köhler equations
had to be further developed, indicate that organic components
have a crucial influence on the activation properties and possibly
the resulting cloud droplet size distribution, as well. The global
climate forcing due to this phenomenon is not known.
A major source of particulate carbon to the atmosphere is the
combustion engine. The emission factor by mass varies greatly
and depends on the type of engine, its age and the kind of fuel
it uses. New developments have strongly reduced the carbon mass
emitted, but this reduction has been achieved by reducing the
size of the particles, not the number. The new particle traps
for diesel engines decrease particle number and mass by about
99%, but measurements show that, due to nucleation after the particle
trap, the number of particles, in the nanometer range, increase
strongly, and that the total number of particles is comparable
to what is found without the particle trap.
Globally, the total number of particles emitted to the atmosphere
from combustion in general, and specifically from engines, is
very large. A simple box model calculation for Germany gives an
estimate of several thousand particles per cc at the border of
the emission area. Such concentrations can prevail over several
thousand kilometers assuming a lifetime of 4-8 days for OC/EC.
Even if the particles are small, some tenths of nm, they might
have a profound influence on particle formation and resulting
size distribution far away from the source area.
The recent WHO review on the health effects of ambient particles,
states that there is strong evidence that fine particles (<
2.5 mm, PM2.5) are more hazardous than larger ones (coarse particles)
in terms of mortality and cardiovascular and respiratory endpoints
in panel studies. This does not imply that the coarse fraction
of PM10 is innocuous. Amongst the characteristics found to contribute
to toxicity in epidemiological and controlled exposure studies,
are metal content, presence of PAHs and other organic components,
endotoxin content as well as small (< 2.5 mm) and extremely
small size (< 100 nm). Further, it states that a number of
source types are associated with health effects, in particular
motor vehicle emissions and coal combustion. These sources produce
primary as well as secondary particles, both of which have been
associated with adverse health effects.
In summary, several important characteristics of particles, other
than mass, such as number size distribution and physical and chemical
properties, are necessary to estimate the effects of OC/EC. But
still TC/EC/OC has to be measured. However efforts have to be
made to minimize artifacts, like adsorption of VOC’s and
losses of semi-volatile particulate matter, in the sampling procedure.
The starting point for the discussion will be to question if
one standard for measuring EC/OC will be useful for addressing
all different effects. Different approaches will be suggested
depending on the effect to be estimated.
Return to Top of Page
TOPIC # 8 .
. .
What new and innovative sampling, analytical, and interpretive
techniques are needed to determine the properties and sources
of carbonaceous aerosol in the atmosphere?
Topic Leader: Hans Moosmüller, Desert
Research Institute, 2215 Raggio Parkway, Reno, NV 89512
Abstract
The total carbon content (TC) of carbonaceous
aerosol can be determined accurately with thermal analysis methods.
In addition, thermal analysis methods utilize the refractory nature
of elemental carbon (EC) and the volatility of organic carbon
(OC) to quantify these two components of TC. However, while heating
filter samples in an inert atmosphere to volatilize OC, part of
the OC pyrolyzes into EC. Despite procedures correcting for this
charring, the resulting EC/OC split depends sensitively on the
heating process and the charring correction. Therefore, EC and
OC determined by thermal-optical analysis are method-dependent
operational definitions, poorly suited as primary standards. However,
this situation may change if the charring process and compounds
removed and created through charring can be better quantified.
Alternative techniques for determining the EC/OC split are needed
for source apportionment, radiative transfer (including visibility),
and health effect applications. Of particular interest are real
time techniques with high (1 s) time resolution, which can be
used to characterize both ambient aerosol and emissions.
OC is a conglomerate of hundreds or thousands of individual chemical
compounds and it is difficult to define common properties which
can be used for OC quantification. On the other hand, EC consists
of graphitic micro crystals, which can be quantified using several
unique properties including 1) strong light absorption, 2) graphitic
Raman spectrum, 3) insoluble in polar and non-polar solvents,
and 4) thermally refractory.
1) Measurement of light absorption directly yields the primary
parameter needed for radiative transfer applications and also
quantifies EC or rather BC (black carbon). Both filter-based and
photoacoustic techniques can be used for this measurement, yielding
some real time instruments with excellent time resolution.
2) Measurement of the graphitic Raman spectrum is currently undergoing
a small resurgence. It quantifies graphitic carbon (GC) thought
to be responsible for carbonaceous light absorption and potentially
identical with BC. In addition, Raman spectroscopy may yield information
on the size of graphitic micro crystals.
3) Dissolving the OC and quantifying the remaining carbon yields
non-extractable carbon (NEC), which is closely related to EC.
Dissolving part of the OC may also be useful as pretreatment for
thermal-optical analysis to reduce charring. However, extraction
methods are not well suited for continuous or real time analysis.
4) Measurement of EC through its thermally refractory properties
has been discussed above.
Topic 8 Follow
Up Abstract—Source
Apportionment Using Semi-Continuous Measurement of OC, EC, and
Other Markers of Combustion Emissions
Delbert J. Eatough, Brigham Young University,
Provo, Utah
Abstract
There has been a significant increase in the number
of semi-continuous instruments available for the determination
of PM2.5 mass and composition. Commercial instruments are now
available for the semi-continuous measurement of nonvolatile
mass, total mass, EC, OC, sulfate, nitrate and other fine particulate
components. Data are often provided as 1-h average measurements.
The availability of data on a time period comparable to or shorter
than meteorological and emission changes in an urban atmosphere
significantly improves the ability of programs such as UNMIX
to do source apportionment in an urban air shed. EC, in particular,
is a good marker for diesel emissions. However, the separation
of diesel, gasoline and wood combustion emissions is of vital
importance and difficult to de using EC and OC data alone. The
addition of other 1-h average gaseous data for potential markers
of primary emissions (e.g. NOx, CO, SO2) and secondary formation
processes (e.g. NO2, O3, H2O2) significantly improves the use
of semi-continuous EC data in the apportionment of EC and OC
from various combustion emissions and secondary formation processes.
The potential use of such data will be illustrated with data
obtained at the Pittsburgh Supersite and in the Wasatch Front
EMPACT study. The development of other 1-h average carbon fraction
data sets would be expected to further improve the power of
this approach.
Return to Top of Page
|
