UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON D.C. 20460
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
March 6, 2009
EPA-CASAC-09-006
The Honorable Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, D.C. 20460
Subject: Consultation on Monitoring Issues Related to the NAAQS for Particulate Matter
Dear Administrator Jackson:
The Clean Air Scientific Advisory Committee (CASAC) Ambient Air Monitoring & Methods
Subcommittee (AAMMS) held a public teleconference on February 11, 2009, to consult with staff from
EPA's Office of Air and Radiation (OAR) Office of Air Quality Planning and Standards (OAQPS) on
issues related to monitoring and speciation sampling for coarse particles (PMi0-2.5). When EPA issued the
final rule to revise both the primary and secondary NAAQS for particulate matter (PM) in October 2006,
a new Federal Reference Method (FRM) was also promulgated for measuring the mass concentration of
PMio-2.5 in ambient air. As part of the revisions to the Ambient Air Monitoring Regulations, PMi 0-2.5
speciation monitoring will be required at National Core (NCore) multi-pollutant monitoring stations by
January 1, 2011.
The CASAC uses a consultation as mechanism for technical experts to provide comments to the
Agency on the issues under consideration. Several areas of concern have been identified in the written
comments provided by the individual Subcommittee members. They include (but are not limited to): the
omission of organic compounds; the current lack of consensus on sampling and analytical methods for
both PM] 0-2.5 and its speciated components; and the haste with which the Agency is committed to launch
the PMio-2.5 monitoring network without sufficient time to analyze the data from the pilot study. As this is
a consultation, we do not expect a formal response from the Agency. We thank the Agency for the
opportunity to provide advice in the NAAQS review process.
Sincerely,
/Signed/
Dr. Armistead (Ted) Russell, Chair
CASAC AAMMS
Enclosures
-------
Enclosure A
U.S. Environmental Protection Agency
Clean Air Scientific Advisory Committee
Ambient Air Monitoring and Methods Subcommittee
CASAC MEMBERS
Dr. Armistead (Ted) Russell (Chair), Professor, Department of Civil and Environmental
Engineering, Georgia Institute of Technology, Atlanta, GA
Dr. Donna Kenski, Data Analysis Director, Lake Michigan Air Directors Consortium,
Rosemont, IL
SUBCOMMITTEE MEMBERS
Mr. George A. Allen, Senior Scientist, Northeast States for Coordinated Air Use Management
(NESCAUM), Boston, MA
Dr. Judith Chow, Research Professor, Desert Research Institute, Air Resources Laboratory,
University of Nevada, Reno, NV
Mr. Bart Croes, Chief, Research Division, California Air Resources Board, Sacramento, CA
Dr. Kenneth Demerjian, Professor and Director, Atmospheric Sciences Research Center, State
University of New York, Albany, NY
Dr. Delbert Eatough, Professor of Chemistry, Department of Chemistry and Biochemistry ,
Brigham Young University, Provo, UT
Dr. Eric Edgerton, President, Atmospheric Research & Analysis, Inc., Gary, NC
Mr. Henry (Dirk) Felton, Research Scientist, Division of Air Resources, Bureau of Air Quality
Surveillance, New York State Department of Environmental Conservation, Albany, NY
Dr. Philip Hopke, Bayard D. Clarkson Distinguished Professor, Department of Chemical
Engineering, Clarkson University, Potsdam, NY
Dr. Rudolf Husar, Professor, Mechanical Engineering, Engineering and Applied Science,
Washington University, St. Louis, MO
Dr. Kazuhiko Ito, Assistant Professor, Department of Environmental Medicine, School of
Medicine, New York University, Tuxedo, NY
-------
Dr. Thomas Lumley, Associate Professor, Biostatistics, School of Public Health and
Community Medicine, University of Washington, Seattle, WA
Dr. Peter H. McMurry, Professor and Head, Department of Mechanical Engineering,
University of Minnesota, Minneapolis, MN
Mr. Richard L. Poirot, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT
Dr. Kimberly A. Prather,* Professor, Department of Chemistry and Biochemistry, University
of California, San Diego, La Jolla, C A
Dr. Jay Turner, Visiting Professor, Crocker Nuclear Laboratory, University of California,
Davis, CA
Dr. Warren H. White, Research Professor, Crocker Nuclear Laboratory, University of
California - Davis, Davis, CA
Dr. Yousheng Zeng, Air Quality Services Director, Providence Engineering & Environmental
Group LLC, Baton Rouge, LA
Dr. Barbara Zielinska, Research Professor, Division of Atmospheric Sciences, Desert Research
Institute, Reno, NV
SCIENCE ADVISORY BOARD STAFF
Ms. Kyndall Barry, Designated Federal Officer, 1200 Pennsylvania Avenue, N.W. (Mailcode
1400F), Washington, DC, Phone: 202-343-9868, Fax: 202-233-0643, (barry.kyndall@epa.gov)
*Dr. Prather did not participate in this CAS AC AAMM Subcommittee activity.
-------
Enclosure B
Comments received:
Mr. George Allen 5
Dr. Judith C. Chow 10
Mr. Bart Croes 31
Dr. Kenneth Demerjian 35
Dr. Delbert J. Eatough 38
Mr. Dirk Felton 48
Dr. Philip Hopke 53
Dr. Rudolf Husar 60
Dr. Donna Kenski 61
Dr. Thomas Lumley 63
Dr. Peter H. McMurry 65
Mr. Richard L. Poirot 69
Dr. Jay R. Turner 74
Dr. Warren H. White 78
Dr. YoushengZeng 80
Dr. Barbara Zielinska 81
-------
Mr. George Allen
PMm-2.5 Speciation Measurement
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Are there
additional PM 10-2.5 target species or methods that can be used to help identify the source of
unidentified mass in order to obtain better mass closure?
For all references to Table 1 in my revised comments, I am referring to the simpler version
included with the EPA slide presentation, not Table 1 from the white paper. Other species:
None that I am aware of; uncertainties in OC (estimated) mass measurements (there are many)
may be part of the missing mass puzzle. Methods: ICPMS may be needed to get robust
measurements of some of the listed acid/water soluble species at the range of concentrations
expected from a dichotomous coarse channel filter. Ions could be done on quartz filter media
(after punches are taken for carbon and protein analysis); if this is being considered, I would
recommend that for the pilot study ions be done on both the Teflon and quartz filters to evaluate
both and to provide a consistency check.
2. Various sampling devices, including dichotomous samplers, MetOne SASS speciation
monitors, PMw andPM2.s FRMs are potential sampling devices (with the appropriate filter
types) for PM 10-2.5 speciation. Among these sampler types, which should be included or excluded
from the pilot network design? Are there other sampling devices not listed here that should be
considered?
For PM-coarse, I strongly recommend using only a dichotomous (dichot) sampler (virtual
impactor, VI) for coarse PM speciation sampling, since it has multiple advantages over a simple
difference method (accuracy and precision, separation of pm-coarse from most of the pm-fme).
The "classic" dichot VI combined with the existing low-volume PM10 inlet is a well
characterized sampler. It is not an FRM or FEM for PM-coarse since there is no NAAQS for
PM-coarse at this time.
3. What are the PM 10-2.5 speciation sampling artifacts that may be encountered using the
samplers mentioned above and how should they be addressed? Is speciation by the difference
method problematic for PM 10-2.5 speciation and if so what specific issues make it problematic?
There are the usual fuzzy issues with OC artifacts, both positive and negative; we will probably
have to live with these. The issue of how OC blanks will be created, handled, and applied to the
data, and how samples are transported and stored (warm or cold) before analysis is important to
clearly define. Nitrate in the coarse mode is complex there will be a mixture of ammonium
nitrate and reaction products of nitric acid with alkaline crustal components. The latter may be
more dominant.
Speciation by difference is strongly discouraged. Any species that is primarily in the fine mode,
5 5
-------
such as sulfate, will be difficult to measure with useful precision by a difference method that is
run routinely in SLT networks. This would be looking at a small difference between two large
measurements, and measurement errors and biases get magnified dramatically when this is done.
For a more detailed discussion of this topic, see: "Techniques for High-Quality Ambient Coarse
Particle Mass Measurements", George A. Allen, Jung-a Annie Oh, and Petros Koutrakis, Journal
of the Air & Waste Management Association, Volume 49, September 1999, PM-133-141.
While it is possible to generate highly precise PM-coarse mass data by difference under very
carefully controlled "research-grade" conditions, routine networks are unlikely to provide this
level of precision. Also, although the difference method can work well for PM when great care
is taken to insure high precision, it remains unclear if analytical techniques for speciation can be
made as precise as mass measurements without excessive cost in a routine network.
4. The current and most widely usedPM2.s speciation sampler is the MetOne SASS and it has a
flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRMfor
PMw-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for PMw-
2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PMw-2.5 FRM?
Yes, this is likely to be problematic for many reasons. I do not recommend anything other than a
16.7 Lpm dichotomous sampler for coarse pm speciation.
5. Is the amount of particle mass collected on the dichotomous filters (especially the minor
flow) sufficient for speciation chemical analysis?
Yes it is, for the "classic" dichot inlet flow of 16.7 Lpm. Note that although the coarse channel
flow is only 1.67 Lpm, the coarse channel dichot filter includes (in theory) all the coarse particles
present in the 16.7 inlet flow. Thus, the low flow for the coarse channel is not a relevant issue,
and the comment on page 8 of EPA's background document for this consultation about adequacy
of XRF for dichot samples is not relevant. Another important parameter for evaluation of dichot
performance is the filter diameter - it does not have to be 47 mm. The early dichots used 37 mm
filters (33 mm exposed area). If enhanced loading is needed on the coarse channel, a 25 mm
filter or a mask could be used; the coarse channel flow is low and large particles do not plug
pores readily, so pressure drop should be minimal even with a much smaller filter diameter.
There are loading-related issues with XRF and coarse particle analysis; self-absorption for large
particles is a major issue and gets worse with high loading.
One area that is still of potential concern is loss of particles from the dichot coarse mode filter
during shipping. It should be clearly demonstrated that shipping does not cause mass loss; early
work by Spengler and Thurston showed shipping loss with a 15 um inlet in the range of 30 to
50% (Spengler and Thurston, JAPCA December 1983, 33:12; and Dzubay and Barbour, JAPCA
August 1983, 33:7). Higher losses would be expected from filters with a 15 um inlet cutpoint
compared to 10 um. More recent work by Bob Vanderpool (EPA-ORD) did not show this loss
using various shipping methods with one exception, where significant losses were observed.
-------
Dichot shipping loss tests done by Environment Canada in 2002 with a PM10 inlet showed
losses ranging from 0 to 25 ug on a filter, but the actual filter loading was not reported and it is
not clear if the original weighings were done before any shipping of the filters had occurred. My
conclusion is that shipping can be an issue for coarse channel dichot filters; an established
protocol that has been shown to not lose mass must be used. The Vanderpool work noted above
can be used to establish a reasonable shipping protocol.
PMm-2 s Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM 10-2.5 species or components missing
from this list? Are there important analysis methods missing from this list?
Species priority: I agree with the order as listed in Table 1, with elements first, soluble ions
second, then carbon and lastly bioaerosols. One aspect of pm-coarse speciation not on this list is
particle surface composition. For coarse mode particles, the surface composition - not the bulk
mass composition may be much more relevant for health effects. For example: urban coarse
mode particles are black, not earth colored, since they have a coating of soot (and things that
absorb on soot...) on the surface (Figure 1 below, from Boston at a location approximately 100
feet above street level, downtown). But by mass, EC would be very low. Thus, simple bulk
mass measurements may not be that relevant for health effect studies. Methods for measurement
of particle surface composition do exist, but may be expensive compared to traditional bulk
chemical analysis.
Figure 1. Deposit of particles greater than 2.5 um on an impactor surface in Boston.
-------
Finally, it may not be worth the added analytical expense to fully characterize carbonate carbon
with a separate analysis. A TOA analysis time-temperature profile that goes higher than 580C-
helium only can give a reasonably useful indication of carbonate carbon. Carbonate comes off as
a sharp peak between 700-750 C, during the transition from 615 to 870 C, the highest helium
only temperature used in the Sunset Labs laboratory carbon analyzer protocol. With some care,
the modest amount of OC that comes off in this last helium only temperature step can be
differentiated from carbonate, since the OC usually evolves off more slowly than carbonate.
Also, it is important to get carbonate off at this time in the run profile, since otherwise it could be
classified as EC; while this is not much of an issue for fine mode pm, it may be a major issue for
coarse mode pm.
2. In the consideration of potential ion measurements for PMw-2.s species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM 10-2.5?
The ions listed are appropriate, with the exception of ammonium. Ammonium ion
concentrations will be biased low in areas where nitrate is the major source of ammonium ion, as
noted above. I do not consider ammonium ion by itself important enough to justify the extra
effort needed to measure it correctly.
3. The 2004 CD included a list of important PMw-2.s components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
These species are at the bottom of my priority list. The analysis is expensive, and at least for
bioaerosols, there is no way to control them. I wouldn't expect a significant amount of fly ash in
the coarse mode at most sites unless they are unusually impacted by a local source.
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
This may be a large problem for coarse pm speciation by difference, since PM10 loading will be
larger than that on a pm-fine or pm-coarse dichot filter, and different [and imprecise] absorption
correction factors would need to be used, resulting in additional coarse-mode by difference
errors. Absorption correction factors are more practical for dichot filters. Still, ICPMS should
be considered as an alternative analysis technique for acid/water soluble components since it is
both more sensitive and does not have any issue with self-absorption.
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PM 10-2.5 far OC and EC given the large expected soil component? If so, how should this
interference be addressed?
-------
Probably not. Metal oxides in crustal material can be sources of O2 or catalysts that can move
the EC/OC split point in TOA analysis, but this does not necessarily mean the EC/OC split result
is wrong it is just a limited amount of "pre-oxidation". However, another potential issue with
TOA is interference from biogenic coarse mode humic material; it may have some absorption of
the -660 nm wavelength used for pyrolysis correction. If biogenic mass is much greater than EC
mass, the interference could be significant.
Network Design
1. Are sites with high PMw and low PM2.s good candidate sites for PM'10-2.5 speciation? Given
that there will be some urban and rural NCore monitoring sites with PM 10-2.5 speciation, what
other factors should be considered in selecting the pilot monitoring and long-term sites or
locations?
If a site's PM10 is nearly all coarse-mode on a mass basis, it still may be useful to do dichot-
based pm-coarse speciation. Some species of interest do not contribute much to the total PM
mass. Pilot sites should be primarily in areas with large populations, to assist health effect
studies. A subset in very large cities should be daily monitoring, since that is needed for health
studies to properly account for any lag effect.
2. If there is an opportunity to modify the NCore PMw-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For example,
changing the total number of required monitors and/or the required locations?
There are issues of large spatial gradients in urban pm-coarse; siting must be urban scale to be
useful. It may be appropriate to have a second urban site (traffic oriented?) with at least limited
speciation at mid-scale locations near sources of coarse PM. Results from the EPA-funded PM-
coarse STAR grant projects on "Sources, Composition, and Health Effects of Coarse Particulate
Matter"
http://cfpub.epa.gov/ncer abstracts/index.cfm/fuseaction/recipients.displav/rfa id/450/records_p
er_page/ALL or http://tinyurl.com/d65fcy will (with one exception) not be complete until after
January 2011, but perhaps preliminary results from projects under this RFA may be able to
inform network design.
Finally, I am concerned about full deployment of this network before an appropriate pilot study
period, including analysis of the data and method performance, is completed. Given the present
status of the PM-coarse speciation program, I doubt that this can be done in time to have a full
network in operation by January 2011. At the least another year, perhaps two years, is needed to
prevent a premature deployment. Going forward with full network deployment before we are
confident of what we can and should do for PM-coarse methods and species would be a massive
waste of limited resources and is to be avoided if at all possible.
-------
Dr. Judith C. Chow
PMm-2.5 Speciation Measurement
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Are there
additional PM 10-2.5 target species or methods that can be used to help identify the source of
unidentified mass in order to obtain better mass closure?
The Revised Table 1 provides a comprehensive list of elements, ions, carbon, and bioaerosol
indicators. Additional PMio-2.5 species (e.g., filter light transmission [babs] and organic species)
can be measured to assist in health assessment and for source attribution, but these additional
marker species will not help to achieve better mass closure. It is not clear how the ORD report
(U.S. EPA, 2006) found that 10-50% of the PMi0-2.5 mass is unaccounted for. The sum of
species is not a good indicator for PMio-2.5 mass closure calculations, since crustal material
accounts for the major portion of PMio-2.5. A material balance that commonly groups by source
types (e.g., geological material, secondary inorganic aerosols, carbon, salt, and others) may be a
better way to evaluate mass closure.
There are two major sources of uncertainties in PMio-2.5 mass closure calculations: 1)
assumptions about the form of metal oxides present in the atmosphere, and 2) the multiplier used
to convert OC to organic mass (OM). Oxides of major crustal components (Pettijohn, 1949) are
used in the calculation for geological material. There are two commonly used metal oxide
formulas that give similar estimates:
Geological Material = 1.89 x Al + 2.14 x Si + 1.4 x Ca+ 1.43 x Fe (1)
Geological Material = 2.2 x Al + 2.49 x Si + 1.63 x Ca + 2.42 x Fe + 1.94 x Ti (2)
(Eq. 2 is the IMPROVE formula;Watson, 2002)
Using OM = 1.2 x QC (and Eq. 1; Solomon et al., 1989), Chow et al. (2002a) found that
geological material accounted for 72-81% of PMio-2.5, with 7-15% organics. The unidentified
mass was negligible (<1%) at six urban locations in Mexico City. PMio-2.5 mass concentrations
were high and variable, ranging from 19 to 56 |ig/m3. The nature of OM is expected to be
different between PM2.5 and PMio-2.5. A 1.2 multiplier is appropriate for hydrocarbons or freshly
emitted vehicle exhaust (composed mainly of H and C atoms), but humic-like substances
(HULIS; Kerley and Jarvis, 1997; Schulten and Leinweber, 2000; Thomsen et al., 2002) and
bioaerosols (Alexis et al., 2006; Bauer et al., 2008; Heinrich et al., 2003; Lee et al., 2007;
Menetrez et al., 2007a) in PMio-2.5 probably justify a larger multiplier to account for unmeasured
oxygen (O) and nitrogen (N) atoms. Elemental analysis of total H, C, N, O, and sulfur (S) may
be applied in the pilot network to estimate OC/OM. These measurements can be compared with
solvent extraction or other calculation methods (El Zanan et al., 2005; 2009). Mass closure
within ħ10-20% is considered sufficient as a quality check.
10 10
-------
2. Various sampling devices, including dichotomous samplers, MetOne SASS speciation
monitors, PMw and PM2.5 FRMs are potential sampling devices (with the appropriate filter
types) for PM 10-2.5 speciation. Among these sampler types, which should be included or excluded
from the pilot network design? Are there other sampling devices not listed here that should be
considered?
Collocated BGIPQ2000 PM2.5 and PMi0 samples have been designated as PMi0-2.5 FRM
(U.S.EPA, 2008). Nevertheless, all filter samplers with well-characterized PM2.5 and PMio inlets
should be considered. The PMi0 sampling efficiency (Keywood et al., 1999; Watson et al., 1983;
Wedding and Carney, 1983) is most important because the ambient size distribution often peaks
near the 10 jim cut-point (Burton and Lundgren, 1987; Lundgren et al., 1984; Lundgren and
Burton, 1995). Since PMio and PMio-2.5 are highly variable in space and time (Baldauf et al.,
2002; Burton et al., 1996; Chow et al., 1981; 1992; 1999; 2000; 2002a), less expensive and
portable samplers should be considered (e.g., the BGI OMNI and the Airmetrics MiniVol) to
allow more of them to be deployed, especially for short-term studies. These 5 L/min units have
well-characterized inlets. These filter samplers can be collocated with PMio-2.5 FRM.
To detect short-duration events, optical particle counters (OPCs, e.g., Grimm Dust Monitor, TSI
DRX DustTrak II) should also be considered. They can be comparable with filter measurements
under certain circumstances (Grimm and Eatough, 2009; Grover et al., 2006; Peters et al., 2006;
Teikari et al., 2003). Collocated TEOMs or BAMs with appropriate inlets are also useful for this
purpose.
Other low- and medium-volume sampling systems are still in use by different state and local
agencies (e.g., Andersen RAAS, URG MASS, R&P2025 Partisol sequential PM2.5, and R&P
2300 samplers). Sampling efficiencies should be estimated based on sampling effectiveness
curves and anticipated size distributions at potential monitoring locations prior to conducting
collocated sampler comparisons. It will be important to verify the homogeneity of PMio-2.5
levels across the sampling array (Mathai et al., 1990) using similar instruments for a valid
comparison study.
3. What are the PM10-2.s speciation sampling artifacts that may be encountered using the
samplers mentioned above and how should they be addressed? Is speciation by the difference
method problematic for PM'10-2.5 speciation and if so what specific issues make it problematic?
Carbon gases are adsorbed onto quartz-fiber filters, and this artifact is imperfectly assessed and
subtracted using field blanks exposed for the entire sample duration and backup filters (Arp et
al., 2007; Chow et al., 2008; McDow, 1986; Noll and Birch, 2008; Watson et al., 2008; 2009).
The magnitude of the gaseous organic sampling artifact correction is reduced by using higher
sample volumes and lower filter exposed areas so that more carbon is collected per unit of filter
area. A better understanding of the adsorbed compounds will be needed to go beyond current
PM2.5 practices that can also be applied to PMio-2.5.
Nitrate is less likely to evaporate from the coarse fraction as it does for PM2.s (Ashbaugh and
Eldred, 2004; Chow et al., 2002b; 2005). PMio-2.5 nitrate often occurs as non-volatile sodium
nitrate (Mamane and Gottlieb, 1992; Mamane and Mehler, 1987; Wu and Okada, 1994). While
11 11
-------
there is abundant nitric acid in the atmosphere, it may react with alkaline PMio-2.s (Goodman et
al., 2000; Hodzic et al., 2006; Laskin et al., 2005; Ooki and Uematsu, 2005; Umann et al., 2005;
Underwood et al., 2001), thereby resulting in a positive nitrate artifact for non-denuded samples
at certain locations.
Collocated PMi0/PM2.5 samplers and virtual impactor (dichotomous) samplers require difference
calculations for which the uncertainties of the individual samples should be propagated (Evans
and Ryan, 1983; Watson et al., 2001). The propagated uncertainties of differences are about the
same for collocated PMio/PM2.s samplers, and may be higher for the dichotomous samplers
because the 10 to 1 flow rate ratio between the fine and coarse channels often varies. As pointed
out by Allen et al. (1999), variability in PMio-2.5 is expected to be higher either by virtual
impactor or by the difference method than with an inertial impactor stage. Inertial impactors also
have uncertainties due to particle bounce and non-uniform deposits.
Most collocated sampling studies (see Chow, 1995; Chow et al., 2008 for study summaries) have
compared PM2.5 and PMio separately. PMio comparisons are often poorer than those for PM2.5
when: 1) PMi0-2.5 constitutes a large fraction of the PMi0; 2) inlets have different sampling
effectiveness curves (even minor ones); and 3) inlet sampling effectiveness changes with inlet
loading.
4. The current and most widely usedPM2.s speciation sampler is the MetOne SASS and it has a
flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRMfor
PM 10-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for PMio.
2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PM 10-2.5 FRM?
The MetOne SASS with a flow rate of 6.7 Lpm should be adequate to collect enough PMio-2.5 for
analysis in urban networks when equipped with a 6.7 Lpm PMio inlet (Kenny et al., 2000; Kenny
and Gussman, 2000). The flow rates should be the same through all filters so that filter
adsorption and evaporation artifacts would be similar. The URG 3000N sampler used in the
current Chemical Speciation Network (CSN) for carbon analysis uses a flow rate of 22.7 Lpm,
which is three times higher than the MetOne SASS. The tenfold difference in face velocity
(107.2 cm/sec for the URG 3000N versus 9.5 cm/sec for the SASS) and its effect on sampling
artifact needs to be further explored. This is more of a PM2.s than a PMio-2.5 issue, however. The
MetOne SASS filter holder is costly to obtain, process, and ship, and greater operational savings
might be obtained by using a simpler sampler pair (e.g., BGI OMNI or PQ2000) that uses the
standard Delron FRM filter ring.
5. Is the amount of particle mass collected on the dichotomous filters (especially the minor
flow) sufficient for speciation chemical analysis?
Yes, the minor flow rate of 1.67 Lpm in the dichotomous sampler is not an issue, since PMio-.2.s
is taken from the 16.7 Lpm drawn through the PMio inlet. The dichotomous sampler with either
a PMis or PMio initial inlet has been used since the mid to late 1970s. The Harvard Six-City
Study used elemental analysis by XRF on dichotomous sampler filters for health assessments
12 12
-------
(e.g., Ferris, Jr. et al., 1979; Spengler et al., 1983; Spengler and Thurston, 1983; Thurston and
Spengler, 1985).
During the early 1980s, dichotomous samplers (Beckman Instruments) were installed in the U.S.
EPA's Inhalable Particulate Matter network for PM2.5 and PMis at 73 locations (Watson et al.,
1981). A subset of samples was submitted to elemental (Al, S, Cl, K, Ca, Ti, Mn, Fe, Ni, Cu, Zn,
Br, Sn, and Pb) and ionic (N(V, SO,f) speciation. The California Air Resources Board (ARE)
also operated dichotomous (Andersen 244) samplers from 1983-2001 at 20 urban sites with
elemental analysis for 30 species on both PM2.5 and PMio-2.5 samples (Motallebi et al., 2003a;
2003b; 2003c). There are several previous studies using virtual impactors that have involved
chemical speciation and used these data for air quality assessment (Dzubay et al., 1977; Dzubay,
1980; Dzubay and Stevens, 1975; John et al., 1988; Li et al., 2001; Lin, 2002; Lin and Tai, 2001;
Magliano, 1988; Mamane and Dzubay, 1990; Rashid and Griffiths, 1993; Sprovieri and Pirrone,
2008; Watson et al., 1981; Witz et al., 1982). PM2.5 and PMi0-2.5 mass and chemical data should
be assembled and evaluated to better understand the nature and chemical composition of PMio-
2.5-
For very low loading PMio-2.5 samples, the analytical instrument for XRF analysis and extraction
volume for ion chromatography (1C) can be optimized to improve the minimum detectable limits
(MDLs).
PMm-2.s Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM 10-2.5 species or components missing
from this list? Are there important analysis methods missing from this list?
PMio-2.5 is enriched in soil-related elements (e.g., Al, Si, K, Ca, Ti, Mn, and Fe) at inland sites
and marine-related elements (e.g., Na and Cl) at coastal sites (e.g., Chow et al., 1994a). Crustal
elements, salt (Na and Cl), ions (e.g., N(V) and carbon (e.g., OC and carbonate) are also
important PMio-2.5 components to be considered.
Without acidification for carbonate analysis, the presence of carbonate could bias both OC and
EC measurements (Chow and Watson, 2002). Cao et al. (2005) found carbonate accounted for
8% of PM2.5 mass during Asian dust storms, and -5% between storms in Xian, China. Although
PM2.5 carbonate may be negligible at non-urban IMPROVE sites (Chow and Watson, 2002), its
abundances may be important in PMio-2.5 at some locations and it should be measured, especially
for Asian dust intrusions (VanCuren, 2003; VanCuren and Cahill, 2002).
Decomposition temperatures for various forms of carbonate vary according to the chemical and
crystalline forms of the carbonate, catalytic influences of other materials, particle size, rate of
heating, and the atmosphere surrounding the material being heated (Webb and Kriiger, 1970).
Figure A, adapted from Webb and Kriiger (1970), shows differential thermal analysis (DTA)
curves for various forms of carbonate. Different heating rates, shown in Figure B, also result in
variations of DTA curves for magnesian limestone. Murray et al. (1951) found that the presence
of salt altered the decomposition temperature of the first dolomite peak by -235 to 40 °C, as
13 13
-------
shown in Table A.
Figure A. Differential Thermal Analyses (DTA) curves for: A) typical calcite (Faust, 1950); B)
aragonite (Faust, 1950); C) natural calcite; D) sample giving curve C, calcined, hydrated, and
partially recarbonated (from Webb and Kriiger, 1970).
3CC
Figure B. Differential Thermal Analyses (DTA) curves for magnesian limestone at a heating
rate of: A) 15 °/min; and B) 5 °/min (from Webb and Kriiger, 1970).
14
14
-------
Table A. Effects of salts on the "characteristic" temperature of the first dolomite peak (Murray
et al., 1951, from Webb and Kriiger, 1970).
; ' 775 +40
,'.-.<, 750 + 15
750 +15
' , '., 745 -HO
. Ğ , 735
* . . 725 -10
1 710 -25
., " , 655 -80
630 -105
565 -170
. .. 560 -175
' 560 -175
.. ' 535 -200
' .,- 520 -215
515 -220
515 -220
- ' , 510 -225
500 -235
Organic carbon in urban and rural PMio road dusts and agricultural soils ranges from 3-20%
(Chow et al., 2003). There is growing evidence that coarse particle organic markers differ from
those for PM2.5 and can be used to identify and quantify source contributions (Boon et al., 1998;
Labban et al., 2006; Omar et al., 2002; Rogge et al., 2006; Rogge et al., 2007; Simoneit et al.,
2004; Song et al., 1999). Analyses for organic species (e.g., polycyclic aromatic hydrocarbons
[PAHs]) should be considered on a subset of samples, possibly by thermal desorption-gas
chromatography/mass spectrometry (TD-GC/MS) where only a small portion of quartz-fiber
filters are needed to speciated non-polar organics (e.g., n-alkanes, alkenes, hopanes, steranes,
PAHs; Chow et al., 2007; Hays and Lavrich, 2007; Ho et al., 2008; Ho and Yu, 2004; Yang et
al., 2005)
Coarse particles, especially minerals, have different light absorption characteristics at different
wavelengths that can be used to distinguish among sources (Fialho et al., 2005; Kim et al., 2004).
Filter light transmission (or babs) has been shown to be highly correlated with PM2.5 or PMio EC
and can serve as a surrogate for PM black carbon (BC). PMi0-2.5 babs measurements at various
wavelengths might be considered, especially for locations where only Teflon-membrane filter
samples will be collected.
2. In the consideration of potential ion measurements for PMw-2.s species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM 10-2.5?
15 15
-------
The anions (i.e., Cl", NCV, and SC>4 ) and cations (i.e., Na+, Ca++, K+, NH4+) listed in the revised
Table 1 should all be measured.
Coarse particle NCV (e.g., sodium nitrate [NaNOs]) has been found in the desert southwest and
environments with abundant coarse PM (Mamane and Mehler, 1987; Watson et al., 1994).
Chemical reactions convert sodium chloride (NaCl) to hydrochloric acid (HC1) and thermally
stable NaNOs (e.g., Ellis et al., 1983; Pilinis and Seinfeld, 1987; Russell and Cass, 1984). These
reactions may dominate coarse particle NCV and occur when NaCl from sea salt, road sanding,
or dry lake beds is an abundant atmospheric constituent. Since most of the PMio-2.5 contains
alkaline crustal material, it is not necessary to use preceding acid gas denuders or nylon filters.
Water-soluble Ca++, Mg+, and PC>43" are also found in coarse particles and may be indicative of
sources. Water-soluble Ca+2 is formed by the reaction of acid gases with calcium carbonate
(Krueger et al., 2004).
Ion analysis is important for PMi0-2.5, but it is unnecessary to allocate a separate nylon channel
for sampling and analysis. If a URG3000N sampler is used for carbon analysis, a 47 mm quartz-
fiber filter should be used (with an exposed area of 11.8 cm2) to allow for both carbon (0.5 cm2
punch) and ionic (1/2 of the 47 mm quartz-fiber filter) speciation. The current 25 mm filter
configuration in the URG 3000N with an exposed area of 3.5 cm2 is not adequate for both carbon
and ion speciation, especially if carbonate or other analyses are considered.
3. The 2004 CD included a list of important PMw-2.s components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
Fly ash may be important if the sampling site is located near a coal-fired power plant or if there
are fly ash storage piles nearby. Most modern industrial stacks have particle control devices that
remove most of the coarse particles, and fugitive emissions are the most probable sources.
Fugitive fly ash can be collected as grab samples, sieved, and resuspended in a laboratory
resuspension chamber (Chow et al., 1994b) for both PM2.5 and PMio size fractions, and
submitted for the same chemical speciation as those for ambient PM. Resuspended fly ash
samples may be useful for in vitro analysis in health assessment. Source profiles for fly ash can
also be used in receptor modeling to estimate coal-fired power plant source contributions.
Both optical and electron microscopy (SEM) can be applied to filter samples (Casuccio et al.,
1983; Edgerton et al., 2009; Gwaze et al., 2007; Kim, 2007; Ott et al., 2008; Watson et al., 2007;
Zhao et al., 2006), but hundreds of particles must be classified to obtain quantitative results.
Microscopy has the advantage of identifying unknown sources, but automated image recognition
methods perform best with low particle densities (i.e., low flow rates of 1-2 Lpm) and a
consistent background, such as that provided by flat polycarbonate filters; this implies a separate
filter channel for microscopic analysis.
16 16
-------
Bioaerosols, including plant pollen and microorganisms (mold and bacteria) have been
recognized as potential health hazards (e.g., Heederik, 2003; Schulze et al., 2006; Steerenberg et
al., 2003; Targonski et al., 1995), but have not been widely studied (e.g., Heinrich et al., 2003;
Menetrez et al., 2000; 2001; 2007b). If standard operating procedures (SOPs) can be established
and target species determined, a subset of PM2 5 and PMi0-2.5 samples (e.g., near agricultural
fields, livestock, and vacant lands) should be analyzed for fungi, endotoxins, and proteins.
5. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
Yes, corrections have been successfully devised and applied (Adams and Billiet, 1976; Adams
and van Grieken, 1975; Berry et al., 1969; Criss, 1976; Dzubay and Nelson, 1975; Hunter and
Rhodes, 1972; Jackson and Hampel, 1992; Rhodes and Hunter, 1972; Sewell et al., 1985;
Wagman et al., 1978). Some further work is needed to compare the different approaches and
evaluate deviations from basic assumptions (e.g., size distributions). The corrections are specific
to the XRF configuration, so it would be good to develop a software system that could be used
by different analysis laboratories.
Self-absorption is most prominent for Al and Si, which have the least energetic absorption edges
and emitted X-rays. Dzubay and Nelson (1975) showed that attenuation changes with 10 and 15
|im particles as shown in Figure C. The derived attenuation factor is shown in Table B (Dzubay
and Nelson, 1975) for their XRF set-up.
1.0
9e
o
<:
r3
z
LU
(
Ğe
0.8
O.S
0,4
0.2
Ex=Ğ.SkeV
1 J 1 L J J
^
o BOTANICAL 10 pn
O QUARTZ 10 pm
A LIMESTONE 15 ^.m
ALMADINE 15 pm
Ex =18 keV
J IL_LJ_L_L_L
14 16 18 20 20 22 24 26
ATOMIC NUMBER OF EXCITED ELEMENT
28 30
Figure C. Attenuation factors for Ka X-rays of various elements calculated for four different
coarse particle compositions (from Dzubay and Nelson, 1975).
17
17
-------
Table B. Midband attenuation and uncertainty for Ka X-ray fluorescence of various elements in
coarse particles deduced from Figure C (from Dzubay and Nelson, 1975).
Element
Al
Si
P
S
Cl
K
Attenuation Factor
(4.5 KeV excitation)
0.41
0.48
0,12
0.15
0.58 + 0.24
0.64 + 0.22
0.70 + 0.20
0.78 + 0.15
0.81 + 0.13
Element
K
Ca
TI
V
Cr
Mn
Fe
Ni
Cu
Zn
Attenuation Factor
(18 KeV excitation)
83
86
0.87 +
0.90 ħ
0.92 +
0.93 +
0.94 +
0.96 +
0.94 +
0.95 +
0.10
0.10
0.08
0.07
0.06
0.05
0.03
0.06
0.05
ICP-MS presents its own difficulties in that Al, Si, and other elements are closely bound as
oxides and need extreme acid digestion to free them for analysis (Anzano and Ruiz-Gil, 2005;
Herner et al., 2006; Lu et al., 2003; Margui et al., 2005; Melaku et al., 2005). ICP-MS is not as
precise as XRF for some key elements, including Si, P, S, and Cl, although it has lower detection
limits for some of the transition metals. ICP-MS should be considered as complementary to, not
as a replacement of, XRF.
6. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PM10-2.5 for OC and EC given the large expected soil component? If so, how shouldthis
interference be addressed?
Metal oxides and catalysts such as NaCl, if present in significant amounts, could be a source of
interference in thermal/optical analysis (Chow et al., 2001; 2006; Fung et al., 2004; Han et al.,
2007a; 2007b; Lin and Friedlander, 1988a; 1988b; 1988c). Several effects occur: 1) mineral
oxides provide oxygen during the inert-atmosphere analysis phased (Fung and Wright, 1990); 2)
the color of the minerals interferes with transmittance and reflectance corrections (Fung et al.,
2004); and 3) catalysts increase the EC oxidation rate (Lin and Friedlander, 1988a; 1988b).
More systematic study is needed to quantify the uncertainties introduced by these effects and to
determine thresholds below which they are tolerable.
Network Design
1. Are sites with high PM10 and low PM2.s good candidate sites for PM'10-2.5 speciation? Given
that there will be some urban and rural NCore monitoring sites with PMw-2.5 speciation, what
other factors should be considered in selecting the pilot monitoring and long-term sites or
locations?
Sites with high PMio and low PM2.5 (e.g., Las Vegas, NV, during spring; Chow et al., 1999) or
18 18
-------
low PMio and high PM2 5 (e.g., wintertime in central California; Chow et al., 2005) are good
candidates to test the high and low PMio-2.s mass and composition, respectively. The PM2.5 Ca
from CSN sites can be used to select sites with the potential for high carbonate concentrations.
A combination of sites that are influenced by paved and unpaved roads, agricultural activities,
storage piles, and bioaerosols are ideal for pilot and long-term monitoring to assess the impacts
from specific sources.
2. If there is an opportunity to modify the NCore PM'10-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For example,
changing the total number of required monitors and/or the required locations?
While there is a desire to collocate PMio and PM2.5 at sites for source attribution and for
assessment of health impacts, this may be an opportunity to modify the NCore sites to
accommodate special needs of more source-oriented PMi0-2.5 sites.
References
Adams, F.C.; van Grieken, R.E. (1975). Absorption correction for x-ray fluorescence analysis of
aerosol loaded filters. Anal. Chem., 47: 1767-1773.
Adams, F.C.; Billiet, J. (1976). Experimental verification of the x-ray absorption correction in
aerosol loaded filters. X-Ray Spectrometry, 5(4): 188.
Alexis, N.E.; Lay, J.C.; Zeman, K.; Bennett, W.E.; Peden, D.B.; Soukup, J.M.; Devlin, R.B.;
Becker, S. (2006). Biological material on inhaled coarse fraction particulate matter
activates airway phagocytes in vivo in healthy volunteers. Journal of Allergy and
Clinical Immunology, 117(6): 1396-1403.
Allen, G.A.; Oh, J.A.A.; Koutrakis, P.; Sioutas, C. (1999). Techniques for high-quality ambient
coarse particle mass measurements. J. Air Waste Manage. Assoc., 49(PM): PM133-
PM141.
Anzano, J.M.; Ruiz-Gil, M. (2005). Comparison of microwave acid digestion with the wet
digestion and ashing methods for the determination of Fe, Mn, and Zn in food samples
by flame AAS. Atomic Spectroscopy, 26(1): 28-33.
Arp, H.P.H.; Schwarzenbach, R.P.; Goss, K.U. (2007). Equilibrium sorption of gaseous organic
chemicals to fiber filters used for aerosol studies. Atmos. Environ., 41: 8241-8252.
Ashbaugh, L.L.; Eldred, R.A. (2004). Loss of particle nitrate from teflon sampling filters: Effects
on measured gravimetric mass in California and in the IMPROVE Network. J. Air Waste
Manage. Assoc., 54(1): 93-104.
Baldauf, R.W.; Wiener, R.W.; Heist, D.K. (2002). Methodology for siting ambient air monitors
at the neighborhood scale. J. Air Waste Manage. Assoc., 52(12): 1433-1442.
19 19
-------
Bauer, H.; Claeys, M.; Vermeylen, R.; Schueller, E.; Weinke, G.; Berger, A.; Puxbaum, H.
(2008). Arabitol and mannitol as tracers for the quantification of airborne fungal spores.
Atmos. Environ., 42(3): 588-593.
Berry, P.P.; Furuta, T.; Rhodes, J.R. (1969). Particle size effects in x-ray spectrometry. In
Advances inX-Ray Analysis, Barrett, C. S., Ed.; Plenum Press: New York, NY, 612-632.
Boon, K.F.; Kiefert, L.; McTainsh, G.H. (1998). Organic matter content of rural dusts in
Australia. Atmos. Environ., 32(16): 2817-2823.
Burton, R.M.; Lundgren, D.A. (1987). Wide-Range Aerosol Classifier - A size selective sampler
for large particles. Aerosol Sci. Technol, 6(3): 289-301.
Burton, R.M.; Suh, H.H.; Koutrakis, P. (1996). Spatial variation in particulate concentrations
within metropolitan Philadelphia. Enivron. Sci. Technol., 30(2): 400-407.
Cao, J.J.; Lee, S.C.; Zhang, X.Y.; Chow, J.C.; An, Z.S.; Ho, K.F.; Watson, J.G.; Fung, K.K.;
Wang, Y.Q.; Shen, Z.X. (2005). Characterization of airborne carbonate over a site near
Asian dust source regions during spring 2002 and its climatic and environmental
significance../. Geophys. Res. -Atmospheres, 110(D03203): 1-8.
doi:10.1029/2004JD005244.
Casuccio, G.S.; Janocko, P.B.; Lee, R.J.; Kelly, J.F.; Dattner, S.L.; Mgebroff, J.S. (1983). The
use of computer controlled scanning electron microscopy in environmental studies. J. Air
Poll. ControlAssoc., 33(10): 937-943.
Chow, J.C.; Shortell, V.; Collins, J.F.; Watson, J.G.; Pace, T.G.; Burton, R.M. (1981). A
neighborhood scale study of inhalable and fine suspended particulate matter source
contributions to an industrial area in Philadelphia. Philadelphia, PA.
Chow, J.C.; Liu, C.S.; Cassmassi, J.C.; Watson, J.G; Lu, Z.; Pritchett, L.C. (1992). A
neighborhood-scale study of PMio source contributions in Rubidoux, California. Atmos.
Environ., 26A(4): 693-706.
Chow, J.C.; Fujita, E.M.; Watson, J.G.; Lu, Z.; Lawson, D.R.; Ashbaugh, L.L. (1994a).
Evaluation of filter-based aerosol measurements during the 1987 Southern California Air
Quality Study. Environ. Mon. Assess., 30(1): 49-80.
Chow, J.C.; Watson, J.G.; Houck, I.E.; Pritchett, L.C.; Rogers, C.F.; Frazier, C.A.; Egami, R.T.;
Ball, B.M. (1994b). A laboratory resuspension chamber to measure fugitive dust size
distributions and chemical compositions. Atmos. Environ., 28(21): 3463-3481.
Chow, J.C. (1995). Critical review: Measurement methods to determine compliance with
ambient air quality standards for suspended particles. J. Air Waste Manage. Assoc.,
45(5): 320-382.
Chow, J.C.; Watson, J.G.; Green, M.C.; Lowenthal, D.H.; DuBois, D.W.; Kohl, S.D.; Egami,
R.T.; Gillies, J.A.; Rogers, C.F.; Frazier, C.A.; Cates, W. (1999). Middle- and
neighborhood-scale variations of PMio source contributions in Las Vegas, Nevada. J. Air
Waste Manage. Assoc., 49(6): 641-654.
Chow, J.C.; Watson, J.G.; Green, M.C.; Lowenthal, D.H.; Bates, B.A.; Oslund, W.; Torres, G.
20 20
-------
(2000). Cross-border transport and spatial variability of suspended particles in Mexicali
and California's Imperial Valley. Atmos. Environ., 34(11): 1833-1843.
Chow, J.C.; Watson, J.G.; Crow, D.; Lowenthal, D.H.; Merrifield, T.M. (2001). Comparison of
IMPROVE and NIOSH carbon measurements. Aerosol Sci. TechnoL, 34(1): 23-34.
Chow, J.C.; Watson, J.G. (2002). PM2 5 carbonate concentrations at regionally representative
Interagency Monitoring of Protected Visual Environment sites. J. Geophys. Res. -
Atmospheres, 107(D21): ICC 6-1-ICC 6-9. doi: 10.1029/2001JD000574.
Chow, J.C.; Watson, J.G.; Edgerton, S.A.; Vega, E.; Ortiz, E. (2002a). Spatial differences in
outdoor PMio mass and aerosol composition in Mexico City. J. Air Waste Manage.
Assoc., 52(4): 423-434.
Chow, J.C.; Watson, J.G.; Edgerton, S.A.; Vega, E. (2002b). Chemical composition of PMi0 and
PM2.5 in Mexico City during winter 1997. Sci. Total Environ., 287(3): 177-201.
Chow, J.C.; Watson, J.G.; Ashbaugh, L.L.; Magliano, K.L. (2003). Similarities and differences
in PMio chemical source profiles for geological dust from the San Joaquin Valley,
California. Atmos. Environ., 37(9-10): 1317-1340. doi: 10.1016/81352-2310(02)01021-
X.
Chow, J.C.; Watson, J.G.; Lowenthal, D.H.; Magliano, K.L. (2005). Loss of PM2.5 nitrate from
filter samples in central California. J. Air Waste Manage. Assoc., 55(8): 1158-1168.
Chow, J.C.; Watson, J.G.; Doraiswamy, P.; Chen, L.-W.A.; Sodeman, D.A.; Ho, S.S.H.; Kohl,
S.D.; Trimble, D.L.; Voepel, H.; Fung, K.K. (2006). Climate change - Characterization
of black carbon and organic carbon air pollution emissions and evaluation of
measurement methods. Phase I: Method intercomparison. Volume I: Interim report.
Report Number DRI Contract No. 04-307; prepared by Desert Research Institute, Reno,
NV, for California Air Resources Board and the California Environmental Protection
Agency, Sacramento, CA;
Chow, J.C.; Yu, J.Z.; Watson, J.G; Ho, S.S.H.; Bohannan, T.L.; Hays, M.D.; Fung, K.K. (2007).
The application of thermal methods for determining chemical composition of
carbonaceous aerosols: A Review. Journal of Environmental Science and Health-Part A,
42(11): 1521-1541.
Chow, J.C.; Doraiswamy, P.; Watson, J.G; Chen, L.-W.A.; Ho, S.S.H.; Sodeman, D.A. (2008).
Advances in integrated and continuous measurements for particle mass and chemical
composition. J. Air Waste Manage. Assoc., 58(2): 141-163.
Criss, J.W. (1976). Particle size and composition effects in x-ray fluorescence analysis of
pollution samples. Anal. Chem., 48(1): 179-186.
Dzubay, T.G.; Stevens, R.K. (1975). Ambient air analysis with dichotomous sampler and x-ray
fluorescence spectrometer. Enivron. Sci. Technol., 9(7): 663-668.
Dzubay, T.G.; Nelson, R.O. (1975). Self absorption corrections for x-ray fluorescence analysis
of aerosols. In Advances in X-Ray Analysis, Vol. 18, Pickles, W. L., Barrett, C. S.,
Newkirk, J. B., Rund, C. O., Eds.; Plenum Publishing Corporation: New York, NY, 619-
21 21
-------
631.
Dzubay, T.G.; Stevens, R.K.; Peterson, C.M. (1977). Application of the dichotomous sampler to
the characterization of ambient aerosols. In X-ray Fluorescence Analysis of
Environmental Samples, Dzubay, T. G., Ed.; Ann Arbor Science Publishers Inc.: Ann
Arbor, MI, 95-105.
Dzubay, T.G. (1980). Chemical element balance method applied to dichotomous sampler data.
Annals of the New York Academy of Sciences., 338: 126-144.
Edgerton, E.S.; Casuccio, G.S.; Saylor, R.D.; Lersch, T.L.; Hartsell, B.E.; Jansen, J.J.; Hansen,
D.A. (2009). Measurements of OC and EC in Coarse Particulate Matter in the
Southeastern United States. J. Air Waste Manage. Assn., 59: 78-90.
Ellis, E.G.; Farber, R.J.; Eatough, DJ. (1983). Formation of Coarse Particle Secondary Nitrate in
the Los Angeles Basin. AerosolSci. Technol., 2: 176.
El Zanan, H.S.; Lowenthal, D.H.; Zielinska, B.; Chow, J.C.; Kumar, N.K. (2005). Determination
of the organic aerosol mass to organic carbon ratio in IMPROVE samples. Chemosphere,
60(4): 485-496.
El Zanan, H.S.; Zielinska, B.; Mazzoleni, L.R.; D.A. Hansen (2009). Analytical Determination
of the Aerosol Organic Mass-to-Organic Carbon Ratio. J. Air Waste Manage. Assn., 59:
58-69.
Evans, J.S.; Ryan, P.B. (1983). Statistical uncertainties in aerosol mass concentrations measured
by virtual impactors. Aerosol Sci. Technol, 2: 531-536.
Faust, G.T. (1950). Thermal analysis studies on carbonates I. Aragonite and calcite. Am. Miner.,
38: 207-224.
Ferris, E.G., Jr.; Speizer, F.E.; Spengler, J.D.; Dockery, D.W.; Bishop, Y.M.M.; Wolfson, M.;
Humble, C. (1979). Effects of sulfur oxides and respirable particles on human health.
Am. Rev. Respir. Dis., 120: 767-779.
Fialho, P.; Hansen, A.D.A.; Honrath, R.E. (2005). Absorption coefficients by aerosols in remote
areas: a new approach to decouple dust and black carbon absorption coefficients using
seven-wavelength Aethalometer data. J. Aerosol Sci., 36(2): 267-282.
Fung, K.K.; Wright, B. (1990). Measurement of formaldehyde and acetaldehyde using 2,4-
dinitrophenylhydrazine- impregnated cartridges during the Carbonaceous Species
Methods Comparison Study. Aerosol Sci. Technol., 12: 44-48.
Fung, K.K.; Chow, J.C.; Chen, L.-W.A.; Chang, M.-C.O.; Watson, J.G. (2004). Evaluation of
pyrolysis correction by reflectance and transmittance in thermal/optical OC/EC analysis
techniques. In Proceedings, Symposium on Air Quality Methods and Technology, Air and
Waste Management Association: Pittsburgh, PA, 10.4-1-10.4-10.
Goodman, A.L.; Underwood, G.M.; Grassian, V.H. (2000). A laboratory study of the
heterogeneous reaction of nitric acid on calcium carbonate particles. J. Geophys. Res.,
22 22
-------
105(D23): 29053-29064.
Grimm, H.; Eatough, DJ. (2009). Aerosol Measurement: The Use of Optical Light Scattering
for the Determination of Particulate Size Distribution, and Particulate Mass, Including
the Semi-Volatile Fraction. J. Air Waste Manage. Assn., 59: 101-107.
Grover, B.D.; Eatough, N.L.; Eatough, D.J.; Chow, J.C.; Watson, J.G.; Ambs, J.L.; Meyer, M.B.;
Hopke, P.K.; Al-Horr, R.; Later, D.W.; Wilson, W.E. (2006). Measurements of both non-
volatile and semi-volatile fractions of fine parti culate matter in Fresno, CA. Aerosol Sci.
Technol.,40: 811-826.
Gwaze, P.; Annegarn, H.J.; Huth, J.; Helas, G. (2007). Comparison of particle sizes determined
with impactor, AFM and SEM. Atmospheric Research, 86(2): 93-104.
Han, Y.M.; Cao, J.J.; Chow, J.C.; Watson, J.G; An, Z.; Jin, Z.D.; Fung, K.K.; Liu, S. (2007a).
Evaluation of the thermal/optical reflectance method for discrimination between char-
and soot-EC. Chemosphere, 69(4): 569-574.
Han, Y.M.; Cao, J.J.; An, Z.; Chow, J.C.; Watson, J.G; Jin, Z.D.; Fung, K.K.; Liu, S. (2007b).
Evaluation of the thermal/optical reflectance method for quantification of elemental
carbon in sediments. Chemosphere, 69(4): 526-533.
Hays, M.D.; Lavrich, R.J. (2007). Developments in direct thermal extraction gas
chromatography-mass spectrometry of fine aerosols. Trac-Trends in Analytical
Chemistry, 26(2): 88-102.
Heederik, D. (2003). Biological Agents Monitoring and Evaluation of Bioaerosol Exposure. In
International Modern Industrial Hygiene, Vol. 2; Biological Aspects, Perkins, J. L., Ed.;
American Conference of Governmental Industrial Hygiensist: Cincinnati, OH, 293-327.
Heinrich, J.; Pitz, M.; Bischof, W.; Krug, N.; Borm, P.J.A. (2003). Endotoxin in fine (PM2.5) and
coarse (PM2.5_io) particle mass of ambient aerosols. A temporo-spatial analysis. Atmos.
Environ., 37: 3659-3667.
Herner, J.D.; Green, P.G.; Kleeman, M.J. (2006). Measuring the trace elemental composition of
size-resolved airborne particles. Environ. Sci. Technol, 40(6): 1925-1933.
Ho, S.S.H.; Yu, J.Z. (2004). In-injection port thermal desorption and subsequent gas
chromatography-mass spectrometric analysis of poly cyclic aromatic hydrocarbons and n-
alkanes in atmospheric aerosol samples. J. Chromatogr. A, 1059(1-2): 121-129.
Ho, S.S.H.; Yu, J.Z.; Chow, J.C.; Zielinska, B.; Watson, J.G; Sit, E.H.L.; Schauer, J.J. (2008).
Evaluation of an in-injection port thermal desorption-gas chromatography/mass
spectrometry method for analysis of non-polar organic compounds in ambient aerosol
samples. Journal of Chromatogr aphy A, 1200(2): 217-227.
doi:10.1016/j.chroma.2008.05.056.
Hodzic, A.; Bessagnet, B.; Vautard, R. (2006). A model evaluation of coarse-mode nitrate
heterogeneous formation on dust particles. Atmos. Environ., 40(22): 4158-4171.
Hunter, C.B.; Rhodes, J.R. (1972). Particle size effects in x-ray emission analysis: Formulae for
23 23
-------
continuous size distribution. X-Ray Spectrometry, 1: 107-111.
Jackson, T.L.; Hampel, J. (1992). Size effects in the energy-dispersive x-ray fluorescence
(EDXRF) analysis of archaeological obsidian artifacts. 23 March 92 A.D.; Los Angeles,
CA.
John, W.; Wall, S.M.; Ondo, J.L. (1988). A new method for nitric acid and nitrate aerosol
measurement using the dichotomous sampler. Atmos. Environ., 22(8): 1627-1635.
Kenny, L.C.; Gussman, R.A. (2000). A direct approach to the design of cyclones for aerosol-
monitoring applications. J. Aerosol Sci., 31(12): 1407-1420.
Kenny, L.C.; Gussman, R.A.; Meyer, M.B. (2000). Development of a sharp-cut cyclone for
ambient aerosol monitoring applications. Aerosol Sci. Technol., 32(4): 338-358.
Kerley, S.J.; Jarvis, S.C. (1997). Variation in 15N natural abundance of soil, humic fractions and
plant materials in a disturbed and an undisturbed grassland. Biology and Fertility of Soils,
24(2): 147-152.
Keywood, M.D.; Ayers, G.P.; Gras, J.L.; Gillett, R.W.; Cohen, D. (1999). An evaluation of PMio
and PM2 5 size selective inlet performance using ambient aerosol. Aerosol Sci. Technol.,
30(4): 401-407.
Kim, K.W.; He, Z.S.; Kim, Y.J. (2004). Physicochemical characteristics and radiative properties
of Asian dust particles observed at Kwangju, Korea, during the 2001 ACE-Asia intensive
observation period. J. Geophys. Res. -Atmospheres, 109(D19).
Kim, K.W. (2007). Physico-chemical characteristics of visibility impairment by airborne pollen
in an urban area. Atmos. Environ., 41(17): 3565-3576.
Krueger, B.J.; Grassian, V.H.; Cowin, J.P.; Laskin, A. (2004). Heterogeneous chemistry of
individual mineral dust particles from different dust source regions: the importance of
particle mineralogy. Atmos. Environ., 38(36): 6253-6261.
Labban, R.; Veranth, J.M.; Watson, J.G.; Chow, J.C. (2006). Feasibility of soil dust source
apportionment by the pyrolysis-gas chromatography/mass spectrometry method. J. Air
Waste Manage. Assoc., 56(9): 1230-1242.
Laskin, A.; Wietsma, T.W.; Krueger, B.J.; Grassian, V.H. (2005). Heterogeneous chemistry of
individual mineral dust particles with nitric acid: A combined CCSEM/EDX, ESEM, and
ICP-MS study. J. Geophys. Res. -Atmospheres, 110(D10): D10208.
doi:10.1029/2004JD005206.
Lee, A.K.Y.; Lau, A.P.S.; Cheng, J.Y.W.; Fang, M.; Chan, C.K. (2007). Source identification
analysis for the airborne bacteria and fungi using a biomarker approach. Atmos. Environ.,
41(13): 2831-2843.
Li, W.-W.; Orquiz, R.; Garcia, J.H.; Espino, T.T.; Pingitore, N.E.; Gardea-Torresdey, J.; Chow,
J.C.; Watson, J.G. (2001). Analysis of temporal and spatial dichotomous PM air samples
in the El Paso-Cd. Juarez air quality basin. J. Air Waste Manage. Assoc., 51(11): 1551-
1560.
24 24
-------
Lin, C.; Friedlander, S.K. (1988a). Soot oxidation in fibrous filters. 1. Deposit structure and
reaction mechanisms. Langmuir, 4(4): 891-898.
Lin, C.; Friedlander, S.K. (1988b). Soot oxidation in fibrous filters. 2. Effects of temperature,
oxygen partial pressure, and sodium additives. Langmuir, 4(4): 898-903.
Lin, C.I.; Friedlander, S.K. (1988c). A note on the use of glass fiber filters in the thermal analysis
of carbon containing aerosols. Atmos. Environ., 22(3): 605-607.
Lin, J.J.; Tai, H.S. (2001). Concentrations and distributions of carbonaceous species in ambient
particles in Kaohsiung City, Taiwan. Atmos. Environ., 35(15): 2627-2636.
Lin, JJ. (2002). Characterization of water-soluble ion species in urban ambient particles.
Environ. Int., 28(1-2): 55-61.
Lu, A.; Zhang, S.; Shan, X.; Wang, S.; Wang, Z. (2003). Application of microwave extraction
for the evaluation of bioavailability of rare earth elements in soils. Chemosphere, 53(9):
1067-1075.
Lundgren, D.A.; Hausknecht, B.J.; Burton, R.M. (1984). Large particle size distribution in 5
United States cities and the effect on a new ambient paniculate matter standard (PMi0).
Aerosol Sci. Technol, 3(4): 467-473.
Lundgren, D.A.; Burton, R.M. (1995). Effect of particle size distribution on the cut point
between fine and coarse ambient mass fractions. Inhal. ToxicoL, 7(1): 131-148.
Magliano, K.L. (1988). Level 1 PMi0 assessment in a California air basin. In Transactions,
PMio- Implementation of Standards, Mathai, C. V., Stonefield, D. H., Eds.; Air Pollution
Control Association: Pittsburgh, PA, 508-517.
Mamane, Y.; Mehler, M. (1987). On the nature of nitrate particles in a coastal urban area. Atmos.
Environ., 21(9): 1989-1994.
Mamane, Y.; Dzubay, T.G. (1990). Dichotomous samplers modified for use with electron
microscopy. Aerosol Sci. Technol., 13: 241-248.
Mamane, Y.; Gottlieb, J. (1992). Nitrate formation on sea-salt and mineral particles - A single
particle approach. Atmos. Environ., 26A(9): 1763-1769.
Margui, E.; Queralt, I; Carvalho, M.L.; Hidalgo, M. (2005). Comparison of EDXRF and ICP-
OES after microwave digestion for element determination in plant specimens from an
abandoned mining area. Anal. Chim. Acta., 549(1-2): 197-204.
Mathai, C.V.; Watson, J.G.; Rogers, C.F.; Chow, J.C.; Tombach, I.H.; Zwicker, J.O.; Cahill,
T.A.; Feeney, P.J.; Eldred, R.A.; Pitchford, M.L.; Mueller, P.K. (1990). Intercomparison
of ambient aerosol samplers used in western visibility and air quality studies. Enivron.
Sci. Technol., 24(7): 1090-1099.
McDow, S.R. (1986). The effects of sampling procedures on organic aerosol measurement.
Oregon Graduate Center.
Melaku, S.; Dams, R.; Moens, L. (2005). Determination of trace elements in agricultural soil
samples by inductively coupled plasma-mass spectrometry: Microwave acid digestion
25 25
-------
versus aqua regia extraction. Anal. Chim. Acta., 543(1-2): 117-123.
Menetrez, M.Y.; Foarde, K.K.; Ensor, D.S. (2000). Fine Biological PM: Understanding Size
Fration Transport and Exposure Potential (Extended Abstract). In The Air & Waste
Management Association Specialty Conference, PM2000: Paniculate Matter and Health
- The Scientific Basis for Regulatory Decision-making, Proceedings of the The Air &
Waste Management Association Specialty Conference, PM2000: Particulate Matter and
Health - The Scientific Basis for Regulatory Decision-making.
Menetrez, M.Y.; Foarde, K.K.; Ensor, D.S. (2001). An analytical method for the measurement of
nonviable bioaerosols. J. Air Waste Manage. Assoc., 51(10): 1436-1442.
Menetrez, M.Y.; Foarde, K.K.; Dean, T.R.; Betancourt, D.A.; Moore, S.A. (2007a). An
evaluation of the protein mass of parti culate matter. Atmos. Environ., 41: 8264-8274.
Menetrez, M.Y.; Foarde, K.K.; Esch, R.K.; Dean, T.R.; Betancourt, D.A.; Moore, S.A.;
Svendsen, E.R.; Yeatts, K. (2007b). Aerosol Sci. Technol. 41(9): 884-893.
Motallebi, N.; Taylor, C.A.; Croes, B.E. (2003a). Particulate matter in California: Part 2 -
Spatial, temporal, and compositional patterns of PM2 5, PMio-2.5, and PMi0. J. Air Waste
Manage. Assoc., 53(12): 1517-1530.
Motallebi, N.; Taylor, C.A.; Turkiewicz, K.; Croes, B.E. (2003b). Particulate matter in
California: Part 1 - Intercomparison of several PM2.5, PMio-2.5, and PMio monitoring
networks. J. Air Waste Manage. Assoc., 53(12): 1509-1516.
Motallebi, N.; Tran, H.; Larsen, L.C.; Croes, B.E. (2003c). Day-of-week patterns of particulate
matter and its chemical components at selected sites in California. J. Air Waste Manage.
Assoc., 53(7): 876-888.
Murray, J.A.; Fischer, H.C.; Shade, R.W. (1951). Report on fundamental research fellowship at
the Massachusetts Institute of Technology 1950-51. Proceedings, National Lime
Association, 4995-116.
Noll, J.; Birch, M.E. (2008). Effects of sampling artifacts on occupational samples of diesel
particulate matter. Environ. Sci. Technol, 42(14): 5223-5228.
Omar, N.Y.M.J.; Abas, M.R.B.; Ketuly, K.A.; Tahir, N.M. (2002). Concentrations of PAHs in
atmospheric particles (PMio) and roadside soil particles collected in Kuala Lumpur,
Malaysia. Atmos. Environ., 36(2): 247-254.
Ooki, A.; Uematsu, M. (2005). Chemical interactions between mineral dust particles and acid
gases during Asian dust events. J. Geophys. Res. -Atmospheres, 110(D3):
ISL000226901400002.
Ott, D.K.; Cyrs, W.; Peters, T.A. (2008). Passive measurement of coarse particulate matter,
PMio-2.5. J. Aerosol Sci., 39(2): 156-167.
Peters, T.M.; Ott, D.; O'Shaughnessy, P.T. (2006). Comparison of the Grimm 1.108 and 1.109
portable aerosol spectrometer to the TSI 3321 aerodynamic particle sizer for dry
particles. Ann. Occup. Hyg., 50(8): 843-850.
26 26
-------
Pettijohn, FJ. (1949). Sedimentary Rocks, Second Edition. Harper & Brothers: New York, NY.
Pilinis, C.; Seinfeld, J.H. (1987). Continued development of a general equilibrium model for
inorganic multicomponent atmospheric aerosols. Atmos. Environ., 21(11): 2453-2466.
Rashid, M.; Griffiths, R.F. (1993). Ambient K, S, and Si in fine and coarse aerosols of Kuala
Lumpur, Malaysia. J. Aerosol Sci., 24(1): S5-S6.
Rhodes, J.R.; Hunter, C.B. (1972). Particle size effects in X-ray emission analysis: simplified
formulae for certain practical cases. X-Ray Spectrometry, 1(3): 113-117.
Rogge, W.F.; Medeiros, P.M.; Simoneit, B.R.T. (2006). Organic marker compounds for surface
soil and fugitive dust from open lot dairies and cattle feedlots. Atmos. Environ., 40(1):
27-49.
Rogge, W.F.; Medeiros, P.M.; Simoneit, B.R.T. (2007). Organic marker compounds in surface
soils of crop fields from the San Joaquin Valley fugitive dust characterization study.
Atmos. Environ., 41: 8183-8204.
Russell, A.G.; Cass, G.R. (1984). Acquisition of regional air quality model validation data for
nitrate, sulfate, ammonium ion and their precursors. Atmos. Environ., 18(9): 1815-1827.
Schulten, H.R.; Leinweber, P. (2000). New insights into organic-mineral particles: composition,
properties and models of molecular structure. Biology and Fertility of Soils, 30(5-6): 399-
432.
Schulze, A.; van Strien, R.; Ehrenstein, V.; Schierl, R.; Kuchenhoff, H.; Radon, K. (2006).
Ambient Endotoxin Level in an Area with Intensive Livestock Production. Ann Agric
Environ Med, 13: 87-91.
Sewell, D.A.; Love, G.; Scott, V.D. (1985). Universal correction procedure for electron-probe
microanalysis: II. The absorption correction. J. Phys. D. App. Phys., 18: 1245-1267.
Simoneit, B.R.T.; Elias, V.O.; Kobayashi, M.; Kawamura, K.; Rushdi, A.I.; Medeiros, P.M.;
Rogge, W.F.; Didyk, B.M. (2004). Sugars - Dominant water-soluble organic compounds
in soils and characterization as tracers in atmospheric particulate matter. Environ. Sci.
Technol, 38(22): 5939-5949. ISL000225272100018.
Solomon, P.A.; Fall, T.; Salmon, L.G.; Cass, G.R.; Gray, H.A.; Davidson, A. (1989). Chemical
characteristics of PMio aerosols collected in the Los Angeles area. J. Air Poll. Control
Assoc.,39(2): 154-163.
Song, X.H.; Hopke, P.K.; Bruns, M.A.; Graham, K.; Scow, K. (1999). Pattern recognition of soil
samples based on the microbial fatty acid contents. Environ. Sci. Technol., 33(20): 3524-
3530.
Spengler, J.D.; Thurston, G.D. (1983). Mass and elemental composition of fine and coarse
particles in six U.S. cities. JAPCA, 33(12): 1162-1171.
Spengler, J.D.; Duffy, C.P.; Letz, R.; Tibbitts, T.W.; Ferris, E.G., Jr. (1983). Nitrogen dioxide
inside and outside 137 homes and implications for ambient air quality standards and
health effects research. Enivron. Sci. Technol., 17(3): 164-168.
27 27
-------
Sprovieri, F.; Pirrone, N. (2008). Particle size distributions and elemental composition of
atmospheric particulate matter in southern Italy. J. Air Waste Manage. Assoc., 58(6):
797-805.
Steerenberg, P.A.; Bischoff, E.W.M.A.; de Klerk, A.; Verlaan, A.P.J.; Jongbloets, L.M.N.; van
Loveren, H.; Opperhuizen, A.; Zomer, G.; Heisterkamp, S.H.; Hady, M.; Spieksma,
F.T.M.; Fischer, P.H.; Dormans, J.A.M.A.; van Amsterdam, J.G.C. (2003). Acute effect
of air pollution on respiratory complaints, exhaled NO and biomarkers in nasal lavages of
allergic children during the pollen season. Int. Arch. Allergy Immunol., 131(2): 127-137.
Targonski, P.; Persky, V.; Rameskirshan, V. (1995). Effect of Environmental Molds on Risk of
Death from Asthma During the Pollen Season. J. Allerg. Clin. Immunol., 95: 955-961.
Teikari, M.; Linnainmaa, M.; Laitinen, J.; Kalliokoski, P.; Vincent, J.; Tiita, P.; Raunemaa, T.
(2003). Laboratory and field testing of particle size-selective sampling methods for
mineral dusts. AIHA Journal, 64(3): 312-318.
Thomsen, M.; Lassen, P.; Dobel, S.; Hansen, P.E.; Carlsen, L.; Mogensen, B.B. (2002).
Characterization of humic materials of different origin: A multivariate approach for
quantifying the latent properties of dissolved organic matter. Chemosphere, 49(10):
1327-1337. doi:10.1016/S0045-6535(02)00335-l.
Thurston, G.D.; Spengler, J.D. (1985). A multivariate assessment of meteorological influences
on inhalable particle source impacts. Journal of Climatology and Applied Meter ology,
24: 1245-1256.
U.S. EPA (2006). Multi-Site Evaluations of Candidate Methodologies for Determining Coarse
Particulate Matter (PMio-2.s): August 2005 Updated Report Regarding Second-
Generation and New PMio-2.s Samplers. Report Number EPA-600/R-06/093; prepared by
U.S. EPA, Research Triangle Park, NC,
U.S.EPA (2008). Method 5 - Particulate Matter (PM). prepared by U.S. Environmental
Protection Agency, Research Triangle Park, NC,
http://www.epa.gov/ttn/emc/methods/method5.html
Umann, B.; Arnold, F.; Schaal, C.; Hanke, M.; Uecker, J.; Aufmhoff, H.; Balkanski, Y.; van
Dingenen, R. (2005). Interaction of mineral dust with gas phase nitric acid and sulfur
dioxide during the MINATROC II field campaign: First estimate of the uptake
coefficient gamma(HNO3) from atmospheric data. J. Geophys. Res. -Atmospheres,
110(D22): ISL000233936200002.
Underwood, G.M.; Song, C.H.; Phadnis, M.; Carmichael, G.R.; Grassian, V.H. (2001).
Heterogeneous reactions of NC>2 and HNOs on oxides and mineral dust: A combined
laboratory and modeling study. J. Geophys. Res., 106(D16): 18055-18066.
VanCuren, R.A.; Cahill, T.A. (2002). Asian aerosols in North America: Frequency and
concentration of fine dust. J. Geophys. Res., 197(D24): AAC 19-1-AAC 19-16.
doi: 10.1029/2002JD002204.
VanCuren, R. A. (2003). Asian aerosols in North America: Extracting the chemical composition
and mass concentration of the Asian continental aerosol plume from long-term aerosol
28 28
-------
records in the western United States. J. Geophys. Res., 108(D20): ACH 1-1-ACH 1-16.
doi: 10.1029/2003 JD003459.
Wagman, J.; Miller, J.L.; Griffin, RJ. (1978). Verification of a particle size correction method
for x-ray fluorescence spectrometric analysis of environmental samples. Anal. Chem.,
50(1): 37-39.
Watson, J.G.; Chow, J.C.; Shah, JJ. (1981). Analysis of inhalable and fine particulate matter
measurements. Report Number EPA-450/4-81-035; prepared by U.S. Environmental
Protection Agency, Research Triangle Park, NC,
Watson, J.G.; Chow, J.C.; Shah, J.J.; Pace, T.G. (1983). The effect of sampling inlets on the
PMio and PMi5 to TSP concentration ratios. J. Air Poll. Control Assoc., 33(2): 114-119.
Watson, J.G.; Chow, J.C.; Lurmann, F.W.; Musarra, S. (1994). Ammonium nitrate, nitric acid,
and ammonia equilibrium in wintertime Phoenix, Arizona. J. Air Waste Manage. Assoc.,
44(4): 405-412.
Watson, J.G.; Turpin, B.J.; Chow, J.C. (2001). The measurement process: Precision, accuracy,
and validity. In Air Sampling Instruments for Evaluation of Atmospheric Contaminants,
Ninth Edition, 9th; Cohen, B. S., McCammon, C. S. J., Eds.; American Conference of
Governmental Industrial Hygienists: Cincinnati, OH, 201-216.
Watson, J.G. (2002). Visibility: Science and regulation. J. Air Waste Manage. Assoc., 52(6):
628-713.
Watson, J.G.; Chow, J.C.; Chen, L.-W.A.; Kohl, S.D. (2007). Non-destructive XRF and SEM
analyses on beta attenuation filters for elemental concentrations at the Craig Road
monitor, prepared by Desert Research Institute, Reno, NV, for Clark County Department
of Air Quality and Environmental Management, Las Vegas, NV;
Watson, J.G; Chow, J.C.; Chen, L.-W.A.; Kohl, S.D.; Tropp, R.J.; Trimble, D.L.; Chancellor,
S.; Sodeman, D.A.; Ho, S.S.H. (2008). Assessment of carbon sampling artifacts in the
IMPROVE, STN/CSN, and SEARCH networks, prepared by Desert Research Institute,
Reno, NV.
http://www.epa.gov/air/airtrends/special studies/20080822_sampling_artifact_rev.pdf
Watson, J.G.; Chow, J.C.; Chen, L.W.A. (2009). Methods to assess carbonaceous aerosol
sampling artifacts for IMPROVE and other long-term networks. J. Air Waste Manage.
Assoc., accepted.
Webb, T.L.; Kriiger, I.E. (1970). Carbonates. In Differential Thermal Analysis Vol. 1 -
Fundamental Aspects, MacKenzie, R. C., Ed.; Academic Press: London, 303-341.
Wedding, J.B.; Carney, T.C. (1983). A quantitative technique for determining the impact of non-
ideal ambient sampler inlets on the collected mass. Atmos. Environ., 17: 873-882.
Witz, S.; Smith, T.B.; Shu, M.; Moore, A.B. (1982). A comparison of mass, lead, sulfate and
nitrate concentrations in a field study using dichotomous, size-selective and standard Hi-
Vol samplers. J. Air Poll. Control Assoc., 32(3): 276-278.
Wu, P.M.; Okada, K. (1994). Nature of coarse nitrate particles in the atmosphere - A single
29 29
-------
particle approach. Atmos. Environ., 28(12): 2053-2060.
Yang, H.; Yu, J.Z.; Ho, S.S.H.; Xu, J.H.; Wu, W.S.; Wan, C.H.; Wang, X.D.; Wang, X.R.;
Wang, L.S. (2005). The chemical composition of inorganic and carbonaceous materials
in PM2.5 in Nanjing, China. Atmos. Environ., 39(20): 3735-3749.
Zhao, Y.C.; Zhang, J.Y.; Sun, J.M.; Bai, X.F.; Zheng, C.G. (2006). Mineralogy, chemical
composition, and microstructure of ferrospheres in fly ashes from coal combustion.
Energy & Fuels, 20(4): 1490-1497.
30 30
-------
Mr. Bart Croes
These comments also reflect input from California Air Resources Board (ARB) staff responsible
for implementing U.S. EPA monitoring requirements and using the data in source apportionment
and health studies.
PMm-2.s Speciation Measurement
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Are there
additional PMw-2.s target species or methods that can be used to help identify the source of
unidentified mass in order to obtain better mass closure?
Not that I am aware.
2. Various sampling devices, including dichotomous samplers, MetOne SASS speciation
monitors, PMw andPM2.$FRMs are potential sampling devices (with the appropriate filter
types) for PM 10-2.5 speciation. Among these sampler types, which should be included or excluded
from the pilot network design? Are there other sampling devices not listed here that should be
considered?
A dichotomous sampler would be a superior choice. While California has dismantled its 20-site
network that operated from 1983 to 2001 (due to resource constraints once the PM2.5 FRM
network was deployed), the data are in reasonable agreement for PM2 5 and PMi0 (Motallebi, et
al., 2003a) which has allowed us to determine long-term PM trends (Motallebi, et al., 2003b) and
perform particle pollution assessments in areas of California that exceed state and federal
ambient air quality standards (Motallebi, 1999; Motallebi and Dolislager, 1999; Motallebi, et al.,
2003bc).
I understand the concerns expressed by Delbert Eatough and Peter McMurry, but PM-coarse
speciation by difference should also be considered. The approach is to continue use of the
MetOne SASS sampler with a 2.5-um cut point for ions and metals, add a 10-um cut point for
ions and metals, and calculate PM-coarse species by difference. Even though the flow is lower
(6.7 Lpm) we have found in the California network that we always have sufficient material for
accurate analyses (few species < LOD), although that may not be the case for other parts of the
country. This approach is economical because the standard SASS has three flow controlled
channels and two orifice channels. To implement this proposal would be to upgrade one orifice
channel to active flow control.
Total, organic and elemental carbon is sampled by the URG 3000N for PM2 5. To implement the
carbon speciation by difference would require a second URG with a 10-um cut point.
3. What are the PM 10-2.5 speciation sampling artifacts that may be encountered using the
samplers mentioned above and how should they be addressed? Is speciation by the difference
method problematic for PM 10-2.5 speciation and if so what specific issues make it problematic?
31 31
-------
Any potential artifacts would be consistent with the PM2.5 speciation program per the difference
approach described above.
4. The current and most widely usedPM2.5 speciation sampler is the MetOne SASS and it has
a flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRMfor
PM10-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for
PMw-2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PMw-2.s FRM?
In California, it has not been problematic with the PM2.5 speciation program except with
elemental carbon, which has been switched to the URG 3000N sampler. We've found measured
SASS mass compares well with measured FRM mass.
5. Is the amount of particle mass collected on the dichotomous filters (especially the minor
flow) sufficient for speciation chemical analysis?
Yes, California has an extensive dataset that confirms there is sufficient mass for trace metal
speciation (Motallebi, et al., 2003a).
PMm-2.s Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM 10-2.5 species or components
missing from this list? Are there important analysis methods missing from this list?
Even though it's not explicitly mentioned, I assume sulfur will be included in the elemental
analyses (at minimum it provides a cross check of sulfate ion by 1C).
2. In the consideration of potential ion measurements for PM 10-2.5 species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM 10-2.5?
Nitrate and ammonium are always important. At least in Los Angeles, coarse particle nitrate
appears to be significant, but it is usually in the form of sodium nitrate.
3. The 2004 CD included a list of important PM]0-2.5 components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
I defer to others.
32 32
-------
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
I defer to others, but we understand these correction factors to be theoretical and crude and that
there is no consensus on their use. In California, we do not use any correction factors for PM2.5
elemental analyses. We understand that at most these correction factors would be no greater than
10%, and possibly inconsequential.
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PMio-2.sfor OC and EC given the large expected soil component? If so, how should this
interference be addressed?
I defer to others.
Network Design
1. Are sites with high PMw and low PM2.5 good candidate sites for PM'10-2.5 speciation? Given
that there will be some urban and rural NCore monitoring sites with PM 10-2.5 speciation, what
other factors should be considered in selecting the pilot monitoring and long-term sites or
locations?
Sites with high PMi0 and low PM2.s suggest areas dominated by wind blown dust. If the source
is obvious for a particular site, I question the value (and diversion of resources) for investigating
an air quality issue that may already be well understood. Because of higher deposition rates for
coarse particles, a PMio-2.5 monitoring site is representative of sources relatively nearby, so it
should be obvious if a single source (e.g., windblown dust, coal pile, tire and brake wear) is
predominant.
The pilot monitoring study should include a southern California site that is high in both PMio and
PM2.5 with potential for significant coarse particle contributions from sources other than wind-
blown dust, such as coarse particle nitrate. Other sites for pilot studies and long-term monitoring
can be selected based on an analysis of existing PMio and PM2.5 network data to identify areas
(using the difference method) that are high in both PMio-2.5 and PM2.5.
2. If there is an opportunity to modify the NCore PM 10-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For example,
changing the total number of required monitors and/or the required locations?
The Health Effects Institute or other representations of the air pollution epidemiology
community should be consulted as to where year-round PM-coarse measurements would be
useful for epidemiologic cohort studies. Otherwise, I agree with USEPA staff that the candidate
NCore site may not be the best choice for PMio-2.5 speciation since the highest concentrations of
PM2.5 and PMio-2.5 are usually not collocated. A study by ARB staff (VanCuren, 1999) concludes
33 33
-------
that, in winter, most coarse (PM10-2.5) particles collected are less than 2 hours old, and even on
days that are not truly "stagnant," samplers are very strongly influenced by their immediate
surroundings (distances less than 10 km), and only weakly influenced by regional emissions. The
implications for interpretation of sample analyses are as follows: (1) Typical PM sampling
networks are unlikely to represent regional conditions; (2) Similarity of samples in time and
space between widely separated samplers probably arises from sampling analogous local
environments rather than a uniformly mixed regional air mass; (3) Even weak sources near a
sampler will prevent regionally representative samples, so that "background" specification in
models can be strongly skewed by misapplication of sampler data; (4) Source-receptor
relationships within a single modeling grid cell can cause measured and modeled source impacts
at a sampler to diverge by orders of magnitude, even for grid cells as small as 1 km; and (5)
Differential deposition of coarse and fine particles will skew source apportionment by chemical
tracers unless the tracers and the source emissions have the same size distribution.
VanCuren, T. (1999) Spatial Factors Influencing Winter Primary Particle Sampling and
Interpretation. J. Air Waste Manage. Assoc., 49: PM-3-15.
Motallebi, N. (1999) Wintertime PM2.5 and PMio Source Apportionment at Sacramento,
California. J. Air Waste Manage. Assoc., 49: PM-25-34.
Motallebi, N. and L. J. Dolislager (1999) Characterization of Particulate Matter in California. J.
Air Waste Manage. Assoc., 49: PM-45-56.
Motallebi, N., C. A. Taylor, Jr., K. Turkiewicz, and B. E. Croes (2003a) Particulate matter in
California: Part 1 - Intercomparison of several PM2 5, PMi0-25, and PMio monitoring
networks. J. Air Waste Manage. Assoc., 53: 1509-1516.
Motallebi, N., C. A. Taylor, Jr., and B. E. Croes (2003b) Particulate matter in California: Part 2
- Spatial, temporal, and compositional patterns of PM2.5, PMio-25, and PMio. J- Air
Waste Manage. Assoc., 53: 1517-1530.
Motallebi, N., H. Tran, B. E. Croes, and L. C. Larsen (2003 c) Day-of-week patterns of PM and
its chemical components at selected sites in California. J. Air Waste Manage. Assoc., 53:
867-888.
34 34
-------
Dr. Kenneth Demerjian
PMm-2.5 Speciation Measurement
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Are there
additional PM10-2.s target species or methods that can be used to help identify the source of
unidentified mass in order to obtain better mass closure?
See items 2 and 3 below.
2. Various sampling devices, including dichotomous samplers, MetOne SASS speciation
monitors, PM10 and PM2.5 FRMs are potential sampling devices (with the appropriate filter
types) for PM 10-2.5 speciation. Among these sampler types, which should be included or excluded
from the pilot network design? Are there other sampling devices not listed here that should be
considered?
Run two (2) EC - OC semi-continuous (hourly average) measurements outfitted with PMio and
PM2.5 inlets respectively. Although this measurement is not totally free of artifacts, volatilization
losses of OC should be significantly less than those resulting from 24-hr average sampling.
3. What are the PM 10-2.5 speciation sampling artifacts that may be encountered using the
samplers mentioned above and how should they be addressed? Is speciation by the difference
method problematic for PM 10-2.5 speciation and if so what specific issues make it problematic?
Evaluate OC artifacts by comparing the 24-hr OC filter based measurements with the integrated
semi-continuous measurements recommended in (2.) above.
4. The current and most widely usedPM2.s speciation sampler is the MetOne SASS and it has a
flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRMfor
PMio-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for PM10.
2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PM 10-2.5 FRM?
The lower flow will likely impact the precision and accuracy of the MetOne SASS as compared
to PMio-2.5, but why speculate, perform the field test experiments!
5. Is the amount of particle mass collected on the dichotomous filters (especially the minor
flow) sufficient for speciation chemical analysis?
Survey the concentration ranges from existing CSN measurement networks and calculate the
expected reductions resulting from flow differences in the dichotomous sampler. If the range in
chemical speciation concentrations estimated for these adjusted flows remain within the
detection limits of the analytical methods, case closed. If not, you will need to consider other
options.
35 35
-------
PMm-2.s Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM 10-2.5 species or components
missing from this list? Are there important analysis methods missing from this list?
The compiled list is reasonable for the majority of contributing sources to PMi0-2.5. There is
always the possibility of a unique PM coarse source whose chemical composition is not on this
list and would require the application of species specific analytical technique.
2. In the consideration of potential ion measurements for PM 10-2.5 species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM 10-2.5?
The measurement of nitrate and ammonium ions is important and the measurement technique
requires the application of appropriate denuders.
3. The 2004 CD included a list of important PMw-2.5 components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
SEM analyses are very labor intensive and not very quantitative. I would suggest a plot study
should be carried out to demonstrate its potential utility in this application.
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
Absorption correction factors are likely to be variable and only reproducible in controlled
environments. Consider the application of ICP/MS analytical technique.
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PM 10-2.5 far OC and EC given the large expected soil component? If so, how should this
interference be addressed?
This is an interesting research question that requires an investment of resources to study the
phenomena in the laboratory and the field.
Network Design
1. Are sites with high PMw and low PM2.5 good candidate sites for PM 10-2.5 speciation?
36 36
-------
Given that there will be some urban and rural NCore monitoring sites with PM'10-2.5 speciation,
what other factors should be considered in selecting the pilot monitoring and long-term sites or
locations?
The majority (if not all) of urban NCore monitoring sites are not deployed in locations to
adequately sample PMio-2.5 exposures. The likely source regions impacting PMio-2.5 exposures in
urban areas are traffic related and associated with populations situate within 500m of major
highways.
2. If there is an opportunity to modify the NCore PM 10-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For example,
changing the total number of required monitors and/or the required locations?
The design criteria for locating urban NCore level 2 monitoring sites must consider
measurements in neighborhoods in proximity (e.g. within 500m) of significant traffic sources
(major highways). Estimates of population with 500m of major roadways within metropolitan
areas is one metric for determining the distribution of traffic impacted vs. central urban monitors
to be deployed. One in three urban monitors deployed to characterize exposures within traffic
impacted neighborhoods would be a reasonable starting point.
37 37
-------
Dr. Delbert J. Eatough
I. General comments:
The speciation of coarse particles being undertaken by EPA as part of the revisions to the
Ambient Air Monitoring Regulations in connection with the EPA final rule to revise the NAAQS
for PM is a new undertaking for the agency. EPA is to be commended for the commitment to
speciation of the PMi0-2.5 (coarse) particle fraction of PMio, even though a NAAQS for coarse
particles has not been set by the agency. I encourage the agency to not be limited to the thinking
of the past 30+ years with respect to monitoring and speciation of fine particles. The speciation
program for fine particles has been necessarily tied to the PM2.5 sampling FRM because of the
requirement to understand the speciation of PM2.5 as traditionally sampled by conventional filter
pack techniques, the basis of the standard. This has lead to a disparity between our
understanding of sampling of fine particles, the optimization of the understanding of components
of fine particles which are not well sampled by the FRM and any subsequent improvement in our
understanding of the etiology of health effects related to fine particles to the extent that the mass
and composition have been inaccurately measured by these techniques. A coarse particle
standard does not yet exist. The health effects which may be associated with exposure to coarse
particles are not currently well understood. Therefore, I recommend that EPA include important
aspects of improved PM monitoring methods in the coarse particle speciation program to allow
for a better tie of all components to any observed health effects exacerbated by exposure to
coarse particles.
While many components of coarse particles will be unaffected by the sampling protocols
used, this is not true for all components. The components listed on slide 9 in the presentation by
Joann Rice represents the current focus of EPA in this sampling effort. Most of the components
listed in that slide will not be significantly affect by the sampling methods used. However, this
is not true for Ammonium and nitrate ions, nor for the semi-volatile fraction of organic C. Both
are components which need to be well understood. Inattention to sampling artifacts in the
program will bias the results towards those components which are not affected by the sampling
method. There are at least two areas where the sampling methods used will significantly affect
the obtained outcome.
In the review of these two areas in the material to follow, I have focused mainly on our
past work. I have done this because of the short time available to provide comments and because
this work is pertinent to the charge questions asked by EPA. However, it should be recognized
that the effects noted below have been observed by several investigators.
38 38
-------
II. Alteration of the Mass and Chemical Composition of Coarse Particles During
Sampling by Mixing Both Fine and Coarse Particles Together.
The origins of fine and coarse particles in the atmosphere are very different, and as a
result their compositions are very different (U.S. EPA 2004). Fine particles are dominated by
primary and secondary material associated with anthropogenic emissions. Ammonium sulfate,
ammonium nitrate, organic material and elemental carbon are major components of these fine
particles. If the aerosol is acidic, the concentrations of ammonium nitrate are low. In contrast,
coarse particles are dominated by crustal material, some biogenic material and materials formed
from abrasive action, e.g. tire debris, etc. Sulfate concentrations are generally very low. Nitrate
can be present in these fine particles, but is generally there are a result of the conversion of sea
salt to sodium nitrate by the reaction with nitric acid (e.g., Ellis et al., 1983).
NaCl(s) + HNO3(g) -> NaNO3(s) + HCl(g)
When fine and coarse particles are collected together reactions can occur in the collected mixture
because of the introduction of dissimilar compositions. The best documented of these changes is
the loss of coarse particulate nitrate by reaction with acidic salts in the fine particles when the
two size fractions are mixed on the filter (e.g., Eatough and Ellis, 1983),
NaNO3(s) + HN4HSO4(s) -> NaHN4SO4(s) + HNO3(g)
and similar reactions. In addition, changes in coarse particulate organic material can be expected
to occur from reaction with fine particulate reactive species. The presence of these artifacts,
some of which are certainly presently unknown, suggests that coarse particulate mass and
composition is best determined if the fine and coarse particles are collected and analyzed
separately. This argues for the use of a dicot sampler or similar device. The EPA heard several
strong pleas fore use of dichot samplers and the avoidance of FM integrated PMio filter collected
sample (for use in a differential speciation calculation) in the coarse particle speciation network.
I add my strong vote to those views for the additional reasons listed here and in Section III.
III. Alteration of the Mass and Chemical Composition of Coarse Particles During
Sampling by the Loss of Semi-volatile Material During Sampling.
Exposure to fine particulate matter (PM2.5, particles with an aerodynamic diameter less
than 2.5 |j,m) has been implicated as a contributor to adverse human health effects including
increases in cardiovascular and pulmonary disease which leads to elevated human mortality and
morbidity (Pope and Dockery, 2007). The role coarse particulate matter (PMio-2.5) may play in
exacerbating health problems is not as well understood. PM2.5 in the atmosphere is not
composed of a single pollutant but consists of both stable and semi-volatile species. Stable
species in the atmosphere include trace metals (including toxic, crustal, and transition metals),
black carbon (BC), and sulfate. Semi-volatile material (SVM) exists in dynamic equilibrium
between the gas and particle phase and includes ammonium nitrate (Finlayson-Pitts and Pitts,
2004) and low molecular weight organic species (e.g., SVM that exists in the atmosphere in the
39 39
-------
particle phase can be lost from particles during sampling making it difficult to measure, Eatough
et al., 2003). The composition of coarse particles may also include these semi-volatile
components, however this semi-volatile material is less well understood in coarse particles.
Several integrated samplers have been developed which accurately determine both
nonvolatile material (NVM) and SVM concentrations (Eatough et al., 2003). These techniques
could be applied to the chemical characterization of coarse particles. Although these samplers
can accurately speciate PM2.5, including SVM, there are several drawbacks to their use.
Integrated samplers are very labor and cost intensive. Collection of filter media and in-lab
analysis are time consuming and cost intensive, resulting in data interpretation weeks and months
from the time of collection. The potential for sample contamination is increased with collection,
transport, and laboratory analysis. Furthermore, 1-h time resolved data is usually not possible
with integrated samplers which inhibit the ability to temporally resolve short term changes in
pollution levels that often occur in urban environments.
To overcome these problems, the development of real-time or near real-time instruments
has been attempted. One of the widely used semi-continuous PM2 5 measurement techniques is
the Tapered-Element Oscillating Microbalance (TEOM, Patashnick et al., 1991) developed by
Rupprecht & Patashnick Co., Inc. The TEOM does not accurately determine total PM2 5 mass
because the particle collection filter is heated to 30-50°C to remove particle bound water, which
also results in loss of SVM. The Rupprecht & Patashnick Filter Dynamic Measuring System
(FDMS, Meyer et al., 2002), and GRIMM monitors (Grimm and Eatough, 2009) have been
developed to measure total PM2 5 mass, including SVM. The GRIMM monitor will measure
both PM2.5 and PMio-2.5- A dicot based FDMS has now been released by Thermo, and is in use at
some state monitoring sites, which will also measure both PM2 5 and PMio-2.5. Real-time
instruments have several advantages including, reliability, cost effectiveness, ease of sampling
and reduction in labor requirements. One prominent advantage of real-time instruments is the
ability to temporally resolve short term episodes of PM2.5 that occur in urban environments that
may be relevant to human health effects. One-hour averaged semi-continuous data has also been
shown to increase the performance (i.e., reduce uncertainty) of source apportionment techniques
to determine sources, both primary and secondary, of urban PM (Grover and Eatough, 2008;
Eatough et al., 2008a).
Both the FDMS TEOM and GRIMM Technologies Inc. (Model 1.180) optical particle
monitor have Nafion dryers at the inlet to remove particulate water. PM2.5 measurements with
these two instruments were compared during the winter of 2007 in Lindon, Utah. The excellent
agreement between these two instruments (Figure 1, from Hansen et al., 2009) suggest that both
measure particulate SVM in a similar manner.
40 40
-------
E
"3)
u.
o
120
100 -
y = (0.995ħ0.007)x + 3.0ħ5.0 |jg/m3
N = 567, R2 = 0.97 o
Sigma = 4.0 pg/m3 (ħ9.6%)
20
40 60 80
FDMS TEOM, |jg/m3
100
120
Figure 1. Comparison of
PM2.5 hourly average mass
measurements by an FDMS-
TEOM and a GRIMM
monitor for the 19 January
through 16 February 2007
study at Lindon, Utah. The
linear regression fit to the
results are given in the Figure
Mass balance studies between 1-h measurements of PM25 mass with an R&P FDMS
TEOM and 1-h measurement of fine particulate components, including the semi-volatile nitrate
and organic material in Riverside, California have also shown excellent agreement (Figure 2,
from Grover et al., 2008b), indicating the semi-volatile material is well measured by the FDMS
TEOM. In This study, the SVM averaged 36% of the total PM2.5 mass.
22 23 24 25 26 27 28 29 30 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Date (July - August 2005)
80
a eo
in
in
ra
| 40
I
| 20
o
O
20 40 60
FDMS (ug/m3)
80
Figure 2. Semi-continuous
constructed mass (y axis in
both plots) vs. FDMS
measured mass. Constructed
mass is the sum of all the
major components of the
aerosol including; NVOM,
SVOM, EC, NH4NO3 and
ammonium sulfate.
41
-------
The above referenced studies indicate that PM2.5 mass and components, including the semi-
volatile material can be reliably determined with technology currently available commercially.
The ability of the GRIMM monitor to measure the coarse particle fraction and the recent
introduction of the dicot FDMS TEOM should allow similar comparisons to be made for PMi0.
2.5.
A measure of the amount of semi-volatile material in both fine and coarse particles can
also be using the conventional TEOM (30 or 50 EC), coupled with FDMS TEOM and GRIMM
monitor data. This is illustrated by results obtained during the study in Lindon (Hansen et al.,
2009). Concentrations of PM2.5 mass were made with a conventional TEOM, and FDMS TEOM
and the GRIMM monitor. PMio was measured with a conventional TEOM and the GRIMM
monitor. The conventional TEOM does not measure the semi-volatile material. The differences
between the conventional TEOM and either the FDMS TEOM or GRIMM monitor give the total
semi-volatile material in the particles (Grover et al., 2008b). The calculated PM2.5 and PMio
SVM concentrations obtained from the FDMS TEOM, GRIMM and conventional TEOM
monitor data are compared in Figure 3 (from Hansen et al., 2009). Included in Figure 3 are error
bars indicated the estimated uncertainty in the comparison. As indicated, the data tend to scatter
around the slope equals one line, however, some of the points show a bias towards higher SVM
concentrations for the PMio. This indicates the presence of semi-volatile material in the coarse
fraction, small, but statistically significant. The SVM material will be composed of semi-volatile
ammonium nitrate and organic material. These secondary components would be expected to be
concentrated in the fine particles. Thus the results shown in Figure 3 are consistent with these
expectations based on atmospheric formation processes. However, a small, but significant
amount of semi-volatile material is present in the coarse particles. Chemical analysis of 24-h
integrated filter samples confirmed that little nitrate was present in the coarse particles,
suggesting the coarse particulate semi-volatile material is organic.
60
50
40
30
20
10
O O o
10
20
30
40
50
SVM, PM25,
FigureS. Comparison of the
calculated concentrations of
PM2.5 and PMio semi-volatile
material at Lindon UT, see
text.
60
-------
Comparisons such as those shown here should be obtained in the EPA coarse speciation
monitoring program to determine if semi-volatile material is important in coarse particles in a
variety of urban areas. This will in turn, indicate if conventional filter pack sampling can be
expected to give an accurate picture of total coarse particle mass and composition. If significant
semi-volatile material is identified in an Urban environment then semi-continuous monitors
which will measure the expected semi-volatile components (e.g. ion chromatographic monitors
such as the URG AIM and carbon analyzers such as the Sunset Dual Oven monitor which
determine both nonvolatile and semi-volatile carbonaceous material) can help elucidate the
sources of the semi-volatile material. This will help to close a hole in the chemical speciation of
coarse particles from the start which has never been adequately addressed in fine particle
research in past EPA monitoring programs.
PMm-2 s Speciation Measurement
1. Table Iprovides a list of proposed PM]0-2.5 species and analysis methods. Are there
additional PMw-2.s target species or methods that can be used to help identify the source of
unidentified mass in order to obtain better mass closure?
Because of the probable alteration of PMi0-2.5 composition and mass when fine and coarse
particles are mixed during sampling (see above discussion), all analyses should be done on a
PMio-2.5 discrete sample and not inferred by a PMio - PM2.5 difference measurement, i.e. a dicot
(or equivalent) sampler should be used.
A significant problem with the protocols given in Table 1 is that they provide for the
determination of species retained on a filter in an integrated sample only. As I have outlined
above, EPA should take advantage of the requirement to characterize coarse particles for better
health effect assessment by insuring that semi-volatile particulate matter which is not retained in
the particles is quantified. This could first involve the semi-continuous determination of total
semi-volatile material by comparison of PMio-2.5 mass using a conventional TEOM to measure
nonvolatile and either a FDMS TEOM or a GRIMM monitor to measure total mass. The
difference between the two is a measure of semi-volatile mass. These measurements should all
be made on a PMio-2.5 basis and not inferred from difference measurements. The 24-h average of
the 1-h conventional TEOM data would be compared to the 24-h integrated filter measured mass.
If these comparison indicates significant semi-volatile material is present in an air shed, then the
composition of the semi-volatile material could be determined using monitors such as the URG
AIM (for nitrate) of the Sunset Dual Oven monitor (for semi-volatile organic material). For the
reasons outlined above, it is important that all these measurements be done on a PMio-2.5
collected sample and not determined by difference.
Fly Ash. We have done extensive work on the analysis of fly ash in collect PM (e.g. Eatough et
al., 2000). SEM analysis is only practicable on Polycarbonate filters. If the data are intended to
be use in source apportionment, a PMi0 sample should be analyzed. We found it best to pre-treat
the sample with an acid wash and firing before analysis to optimize detection of insoluble fly ash
particles.
-------
2. Various sampling devices, including dichotomous samplers, MetOne SASS speciation
monitors, PMw andPM2.s FRMs are potential sampling devices (with the appropriate filter
types) for PM 10-2.5 speciation. Among these sampler types, which should be included or excluded
from the pilot network design? Are there other sampling devices not listed here that should be
considered?
For reason outlined in #1 and discussed in more detail above, chemical speciation should only be
done on a dichotomous sampler (or equivalent device to provide a discrete PMi0-2.5 sample).
This would rule out the use of PMio FRM samplers.
3. What are the PM 10-2.5 speciation sampling artifacts that may be encountered using the
samplers mentioned above and how should they be addressed? Is speciation by the difference
method problematic for PM 10-2.5 speciation and if so what specific issues make it problematic?
Speciation by the difference method is problematical. Details on artifacts and the approach
which should be taken in speciation are discussed in detail above.
4. The current and most widely usedPM2.s speciation sampler is the MetOne SASS and it has a
flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRM for
PM 10-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for PMw-
2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PMw-2.s FRM?
It has been my experience, based on extensive work with diffusion denuder samplers, that in a
24-h sample the results obtained are largely independent of flow rate. This is because the
kinetics for changes which occur during sampling are generally of the order of 10s of minutes or
shorter. However, a difference method suffers the problems discussed above at any flow rate.
5. The current and most widely usedPM2.s speciation sampler is the MetOne SASS and it has a
flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRM for
PM 10-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for PMw-
2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PMw-2.s FRM?
Since the total mass collected in the minor flow is a function of the inlet flow and not the minor
flow, I would not think this would be a problem.
PMm-2.s Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM 10-2.5 species or components missing
from this list? Are there important analysis methods missing from this list?
If your goal is characterization for source apportionment and health effect etiology applications,
none of the components should be eliminated. The importance of quantifying semi-volatile
material in coarse particles has been discussed in the previous section and in the introductory
material.
44
-------
2. In the consideration of potential ion measurements for PMw-2.s species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM'10-2.5?
Nitrate is expected to be present mainly as the sodium salt. But knowing both nitrate and
ammonium ion would be valuable. If, as we suspect, ammonium nitrate is negligible in most
Urban areas in coarse particles, it would not seem worth the effort to use a denuder in integrated
sampling, _if you also include monitoring for semi-volatile material with conventional TEOM and
either FDMS TEOM or GRIMM monitors as outlined above. This would highlight the possible
presence of ammonium nitrate and followup with an AIMS would verify the ammonium nitrate
presence. If however, you do not monitor for semi-volatile species, then using a denuder system
for the coarse particle collection is advised. Do you have any data on the transfer of coarse
particles through an annular denuder?
3. The 2004 CD included a list of important PM 10-2.5 components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
I have commented on fly ash above. I am not familiar with monitoring for biological materials.
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
I will leave this to those who know.
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PMw-2.sfor OC and EC given the large expected soil component? If so, how should this
interference be addressed?
This is certainly a potential interference but I do not know to what extent. Those knowledge in
this area did not seem to think this would be a large problem.
Network Design
1. Are sites with high PMw and low PM2.s good candidate sites for PM 10-2.5 speciation? Given
that there will be some urban and rural NCore monitoring sites with PM 10-2.5 speciation, what
other factors should be considered in selecting the pilot monitoring and long-term sites or
locations?
This question seems to imply the use of difference methods. I have outlined why this is not a
good idea above.
45
-------
2. If there is an opportunity to modify the NCore PM'10-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For example,
changing the total number of required monitors and/or the required locations?
I think the routine inclusion of monitoring for semi-volatile species using the sem-continuous
mass monitors I have mentioned above would be a valuable addition. This would proved input
on both PMio-2.5 mass and components as well as give diurnal data which would be very helpful
in understanding the effects of changing meteorology, emissions and atmospheric chemistry on
the observed concentrations and composition.
References
Eatough, D.J.; Grover, B.D.; Woolwine, W.R.; Eatough, N.L.;Long, R.; Farber, R. (2008)
Source Apportionment of 1-hr Semi-Continuous Data During the 2005 Study of Organic
Aerosols in Riverside (SOAR) Using Positive Matrix Factorization, Atmos. Environ. 42,
2706-2719.
Eatough, D.J.; Long, R.W.; Modey, W.K. and Eatough, N.L. Semi-Volatile Secondary Organic
Aerosol in Urban Atmospheres: Meeting a Measurement Challenge (2003) Atmos.
Environ. 37, 1277-1292.
Eatough D.J.; Farber RJ. and Watson J.G. (2000) Second Generation Chemical Mass Balance
Source Apportionment of Sulfur Oxides and sulfate at the Grand Canyon during the
Project MOHAVE Summer Intensive, J. Air & Waste Management Association 50, 759-
774.
Eatough DJ. and Ellis E.G. (1983) A Comparison of High and Low Volume Particulate Matter
Nitrate Concentrations in the Los Angeles Basin, Aerosol Science and Technology, 2,
177.
Ellis E.G.; Farber RJ. and Eatough DJ. (1983) Formation of Coarse Particle Secondary Nitrate
in the Los Angeles Basin, Aerosol Science and Technology., 2, 176.
Finalyson-Pitts B J.; Pitts, J.N. Chemistry Of The Upper And Lower Atmosphere: Theory,
Experiments and Applications; Elsevier, 2004.
Grimm, H. and Eatough, D J. (2009) Aerosol Measurement: The Use of Optical Light Scattering
for the Determination of Particulate Size Distribution, and Particulate Mass, Including
the Semi-Volatile Fraction; J. Air Waste Manage. Assoc. 59, 101-107.
Grover, B.D. and Eatough, D J. (2008a) Source Apportionment of One-Hour Semi-Continuous
Data using Positive Matrix Factorization with Total Mass (Nonvolatile plus Semi-
Volatile) Measured by the R&P FDMS Monitor; Aerosol Sci. Technol. 42, 28-39.
Grover, B.D.; Eatough N.L.; Woolwine, W.R.; Cannon, J.P.; Eatough, D J.; Long, R.W. (2008b)
Semi-continuous Mass Closure of the Major Components of Fine Particulate Matter in
46
-------
Riverside, CA; Atmos. Environ. 42, 250-260.
Hansen, 1; Woolwine, W.R. Ill; Bates, B.L, Clark, J.M.; Kuprov, R.Y.; Mukherjee, P.; Murray,
J.A.; Simmons, M.A.; Waite, M.F.; Eatough, N.L.; Eatough, D.J.; Long R. and Grover
B.D. (2009) Semi-Continuous PM2.5 and PMio Mass and Composition Measurements in
Lindon, Utah During Winter, 2007. J. Air Waste Manage. Assoc., in press.
Meyer, M.B.; Patashnick, H.; Ambs, J.L. (2002) Ongoing Development of a Continuous
Reference Standard Particulate Matter Mass Monitor for Ambient Air; paper presented at
the Symposium on Air Quality Measurement Methods and Technology, San Francisco,
CA, November.
Patashnick, H.; Rupprecht, E.G. Continuous PMio Measurements Using the Tapered Element
Oscillating Microbalance; J. of Air & Waste Manage. Assoc. 41, 1079-1083.
Pope, C.A. Ill; Dockery, D.W. Health Effects of Fine Particulate Air Pollution: Lines that
Connect; J. Air & Waste Manage. Assoc. 2007, 57, 709-742.
U.S. EPA (2004) Air Quality Criteria for Particulate Matter. Report No. EPA/600/P-99/002aF,
Environmental Protection Agency, Research Triangle Park, NC.
47
-------
Mr. Dirk Felton
PMm-2.5 Speciation Measurement
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Are there
additional PM10-2.s target species or methods that can be used to help identify the source of
unidentified mass in order to obtain better mass closure?
The conversion to OM from OC is contributing to the unidentified mass issue. The EPA should
endeavor to reduce this uncertainty by performing detailed OC component and molecular marker
analysis on some of the filter samples.
2. Various sampling devices, including dichotomous samplers, MetOne SASS speciation
monitors, PM10 andPM2.s FRMs are potential sampling devices (with the appropriate filter
types) for PM 10-2.5 speciation. Among these sampler types, which should be included or excluded
from the pilot network design? Are there other sampling devices not listed here that should be
considered?
The SASS PM-10 inlet is not yet viable and the SASS may not have enough sample volume for
some PMc components in some areas of the country. There are many existing Cascade Impactor
designs and it is possible that one of those could be modified to make it appropriate for
collection of some of the components of interest for the PMc pilot. Other technologies also not
on this list include continuous instruments and particle mass spectrometry. Again these latter
two new technologies would have to be modified with suitable inlets and they are not user
friendly but they have the potential of producing the best results.
3. What are the PM 10-2.5 speciation sampling artifacts that may be encountered using the
samplers mentioned above and how should they be addressed? Is speciation by the difference
method problematic for PM 10-2.5 speciation and if so what specific issues make it problematic?
The difference method is likely to suffer the most in areas with high nitrate and high OC. The
problem is that component conversion and volatilization are likely to occur preferentially for one
component size fraction (PM-10, PM-2.5) or the other depending on the source. This will make
the bias between the size fractions component specific.
The Dicot method also has the disadvantage of requiring an adjustment for fine particle
intrusion. This adjustment to Dicot mass concentration is less attractive for component analysis
because it is not known if the correction or the presence of fine particles on the coarse mode
filter will preferentially affect one component over another.
4. The current and most widely usedPM2.s speciation sampler is the MetOne SASS and it has a
flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRMfor
PM 10-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for
2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PMw-2.s FRM?
48
-------
In Queens, NY, typical 16.7 L/min FRM filter loadings are 260 ug for PM-2.5 and 480 ug for
PM-10. It is expected that a 6.7 L/min sampler would have loadings of 104 ug and 192 ug
respectively. These lower loadings certainly increase the significance of the precision of the
weighing procedure and the magnitude of the filter blanks which are permitted to be as high as
30 ug.
During the recent FEM equivalency evaluations performed by ThermoFisher Scientific, Franklin,
MA, the vendor was required to run filter based low volume FRM PM-2.5 and PM-10 samplers
in triplicate. The average standard deviation of the three calculated PMc values over 44
sampling days was 0.8 ug/m3. If we assume that the 6.7 L/min sampler performs as well as a
carefully operated FRM during an equivalency test, which should be considered a best case, 0.8
ug would indicate a standard deviation of about 20 ug of filter loading on the 6.7 L/min sampler.
That makes the precision of the method an issue since the expected PMc by difference is only 88
ug for the 6.7 L/min sampler.
5. Is the amount of particle mass collected on the dichotomous filters (especially the minor
flow) sufficient for speciation chemical analysis?
During the Thermo Fischer Scientific FEM evaluation in Queens NY last summer, the
correlation between the virtual impactor Coarse mode data and PMc by filter difference was not
as good as the correlation with the fine mode. This is not too surprising considering the low
PMc concentrations and the issue of large particle retention on filters prior to gravimetric
analysis. The amount of mass should be sufficient for most analysis but the precision of the PMc
mass measurements by Dicot will not be as good as that for PMc by difference.
PMm-2.5 Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM10-2.s species or components missing
from this list? Are there important analysis methods missing from this list?
The first question will have a different answer depending on the end use of the data. The PMc
component mix will vary greatly by location, season and proximity to sources. In the northeast,
the concentrations of elements in PMc tend to be around 30% of the PMc mass. Again from a
northeast perspective, the only element missing from Table 1 is Vanadium which can have
significant PMc concentrations in some areas. The New York State Department of
Environmental Conservation (NYSDEC) undertook a pilot study for PMc speciation by
difference. Low volume PM-2.5 and PM-10 FRM samplers were installed in New York City
and in Niagara Falls. The samplers were operated on a 1/6 day schedule for a year and the filters
were weighed and analyzed by XRF. The results demonstrated that some species such as Pb
were found nearly equally in the fine and coarse mode. Arsenic was unique in that it appears to
have a distinct source in both fine and coarse modes.
49
-------
The concentrations of the elemental data for all but about 10 elements were very low in the
coarse mode.
PMcoarse to PM-10 Ratio: Canal St 2005 (Mass and Elements)
1 -,
0.8
0.6
0.4
0.2
Se S Ni Zn Br As PM Pb K IVh Ma Cu V Sr Zr Fe Al Ti Ca Si Cl Cr Mg
-Median 25th %tile 75th %tile
NYC PMcoarse and PM-2.5 Ave Elemental Concentrations
Calendar Year 2005 (Sulfur omitted)
Fe Si Ca Na Cl Al K Mg Zn V Ti Cu P Mn Pb Ni Br La Sm Zr Sr Ce
Ğ PMc Annual Ave Cone
PMc 75%tile
PM-2.5 Annual Ave Cone
- PM-2.5 25%tile
PM: 25%tile
- PM-2.5 75%tile
Carbon is probably the most important overall group of PM components. It includes many
compounds that are not uniquely identified and are likely to be correlated with health effects.
The poor understanding of OC's conversion to organic mass is also not well understood which
makes it difficult to accurately characterize OC PMc. Continuous carbon monitoring is the most
cost effective means of obtaining daily datasets. The Sunset Labs instrument is available as a
prototype in a configuration for PM-10. The pilot should establish a few sites in different parts
of the country with hourly OC/ECcoarse by difference. The EPA should also attempt to include
the volatile portion of all of the components collected in the pilot.
50
-------
2. In the consideration of potential ion measurements for PMw-2.s species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM'10-2.5?
Soluble ions are less important for eastern sites. The ion concentrations from Table 1, except for
Nitrate, can be estimated from the elemental data. This will not be the case for the west where
ammonium nitrate is very important and cannot be estimated from other measurements. A
denuder should be employed and the sampler should be modified to keep the sample cool after
and during the sampling event if possible.
3. The 2004 CD included a list of important PMw-2.5 components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
As with all PMc data, the relative importance of one component over another is dependent on
location and sources. In and around industrial areas and large, older urban areas such as cities in
the midwest and northeast, fly ash and oil soot are found in significant concentrations in the air.
Often the majority of these particles are larger than 10 microns and are likely to be excluded in
PMc measurements. Since these particles may be correlated with health effects, the pilot
monitoring program should consider evaluating concentration differences in results between
samples collected with PMc samplers and samplers with low volume TSP inlets such as the
URG-2000-30DG.
Teflon filters are suitable for SEM work but are not ideal. Their depth structure prevents reliable
quantitation because it tends to hide the smaller particles.
Biological materials such as mold spores, insect parts and moth scales will partially be collected
with PMc samplers. It certainly is plausible that mold spores could have a correlation with
health effects but it is unlikely that generalized assay tests would sufficiently identify specific
components of PM that probably affect only a limited portion of the population.
Pollen for the most part is too large to be included in PMc measurements. Specialized samplers
such as the Burkhard spore trap and Rotorod sampler are used for Pollen sampling and counting.
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
The XRF analysis for the more heavily loaded PM-10 filters is modified by extending the count
time for the detector. Extending the count time may or may not eliminate the uncertainty
associated with an increase in X-Ray absorption. The EPA should sponsor experimentation by
loading filters with different component mixes of PMc and then compare the XRF results with
ICPMS digestive analysis.
51
-------
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PMw-2.sfor OC and EC given the large expected soil component? If so, how should this
interference be addressed?
The IMPROVE program should have some insight into this.
Network Design
1. Are sites with high PMw and low PM2.5 good candidate sites for PM'10-2.5 speciation? Given
that there will be some urban and rural NC'ore monitoring sites with PM 10-2.5 speciation, what
other factors should be considered in selecting the pilot monitoring and long-term sites or
locations?
The best site is the one that has significant concentrations of the PMc component that has the
greatest correlation with health effects. To that end, the pilot should be designed to provide
survey information covering the entire range of component mix and concentration ranges. This
is the most effective way to determine the importance of specific components of PMc for the
design of a future more relevant PMc network. NCore sites are situated to represent large spatial
scales and generally will tend to have relatively low contributions from known sources of larger
particles.
There are no standards for components of PMc or for mass of PMc so the pilot as well as the
initial longer term deployment should be designed to help ascertain the highest concentrations of
PMc components likely to be associated with health effects. This strategy will accelerate the
linking of health effects and individual PMc component concentrations
2. If there is an opportunity to modify the NCore PM 10-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For example,
changing the total number of required monitors and/or the required locations?
It is appropriate to tailor the PMc speciation program to the species that are prevalent and likely
to be related to health effects in specific regions of the country. In northeastern cities for
instance, PMc OC/EC are likely to be much more important than PMc elements. In western
cities, soil and nitrate are likely to more relevant. Once health effect correlations are found in
areas with high concentrations of specific components of PMc, the network can be modified to
include that component in sampling efforts in the rest of the network.
Health researchers are continually stressing a preference for daily data due to lag structure in
health effects studies and the PMc pilot should be designed to the extent possible to provide daily
or continuous data.
52
-------
Dr. Philip Hopke
PMm-2 s Speciation Measurement
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Are there
additional PM 10-2.5 target species or methods that can be used to help identify the source of
unidentified mass in order to obtain better mass closure?
Part of the issue here is what sampling method is to be used. Is there really sufficient precision to
be able to use the difference method which unfortunately is still being put forth as the standard
approach? I suggest that the key is less on the analytical method as on the sampling approach.
2. Various sampling devices, including dichotomous samplers, MetOne SASS speciation
monitors, PMw and PM2.5 FRMs are potential sampling devices (with the appropriate filter
types) for PM 10-2.5 speciation. Among these sampler types, which should be included or excluded
from the pilot network design? Are there other sampling devices not listed here that should be
considered?
I strongly recommend against using the difference method. There is the potential for substantial
interactions between the fine and coarse particles on the PM10 filter. For example, fine NH4NO3
can dissociate to produce HNO3 that would react with alkaline coarse particles. This could
produce an overestimate of coarse NO3.
Dichotomous samplers complicate the analyses by putting about 10% of the fine particle mass
onto the coarse particle filter.
It probably makes more sense to develop an impactor system to collect specific coarse particle
samples where there would be little supercoarse or fine particle contamination. Such an
impactor would have two stages. The first would have as sharp as possible PM10 cut and the
second would have a sharp PM2.5 cut. The PM2.5 stage would be the one to be analyzed.
However, since this approach was not adopted when the idea of coarse particles was first
considered (such that it would now be available), it makes sense to go with the dichot.
3. What are the PM 10-2.5 speciation sampling artifacts that may be encountered using the
samplers mentioned above and how should they be addressed? Is speciation by the difference
method problematic for PM 10-2.5 speciation and if so what specific issues make it problematic?
See above.
One question is whether low volume systems provide sufficiently large samples for all of the
analyses envisioned. Clarkson is involved in an EPA CRADA to test a high volume (400 LPM)
dichotomous sampler that would provide greater mass for chemical characterization. However,
it is still a virtual impaction system that would still intermix some fine particles into the coarse
particle sample.
53
-------
4. The current and most widely usedPM2.s speciation sampler is the MetOne SASS and it has a
flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRMfor
PM 10-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for PMw-
2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PMw-2.s FRM?
This is why they should develop a reasonable flow, simple impactor system to collect PM10-2.5.
It should be relatively easy to design and build such a system. It should have been done rather
than spending all of the money that has been spent on trying to justify the inherently flawed
difference method and the problematic dichotomous sampler.
5. Is the amount of particle mass collected on the dichotomous filters (especially the minor
flow) sufficient for speciation chemical analysis?
Probably not. However, it is possible to build high flow dichots that would provide sufficiently
large samples to ensure good analytical precision. I have provided a presentation outlining such
a system that is ready for testing.
PMm-2 s Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM 10-2.5 species or components missing
from this list? Are there important analysis methods missing from this list?
We know a lot less about the coarse fraction and what are critical species particularly for primary
biological species that are going to be more prevalent in coarse fraction samples compared to
fine particle samples. Protein alone is highly unlikely to provide a reliable estimate of total
biological particulate mass.
Are there sufficient resources available to perform CCSEM on enough filters for enough
particles to provide statistically valid numbers of particles to adequate quantify the particle types.
Are they envisioning automated analysis of the data? Will they collect secondary electron
images as well as backscatter images and x-ray spectra? One can envision automated image
analysis software that could sort the particles into various major categories, but there is no
discussion of how the analyses are to be conducted. Manual SEM and review of images will not
be sufficient to provide support to a national network of samplers.
2. In the consideration of potential ion measurements for PM]0-2.5 species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM 10-2.5?
Nitrate is important for the coarse sample as well as fine although the coarse particle nitrate is
generally present because of the reaction of gaseous HNO3 with the alkaline coarse particles.
This problem also supports the need for using devices that separate coarse and fine fractions and
not using the difference method.
54
-------
3. The 2004 CD included a list of important PMjo-2.5 components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
See answer to 1 above. Normally Teflon is not an ideal surface for CCSEM. It can be adequate
for manual SEM analysis, but it is not possible to have enough people on enough microscopes to
do enough particles on enough filters to be meaningful. Much of the biological species would be
more carbohydrates than protein so it is not clear that a total protein assay would suffice to
provide an index of total biological material. It may be more useful to look at specific species
such as endotoxin or specific aeroallergens. There are now easy to use kits at reasonable prices
to perform ELIZA analyses. It is likely that additional species tests could be developed for other
identified allergens.
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
Again, this depends critically on what is collected. If you are still using the difference method,
then it makes the XRF analysis much more difficult than if there is a narrower range of particle
sizes in just a coarse particle sample. Also the differences in composition from location to
location also need to be considered as it is the combination of particle composition and size that
affect the penetrability of the fluoresced x-ray.
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PMw-2.sfor OC and EC given the large expected soil component? If so, how should this
interference be addressed?
I am not sure that we really know enough to be able to develop quantitative relationships
between transition metal oxides and changes in relative OC/EC values. There is evidence for
influence but not clear relationships and it certainly will also depend on the nature of the
carbonaceous material as well as the nature of the oxides.
Network Design
1. Are sites with high PMw and low PM2.s good candidate sites for PM'10-2.5 speciation? Given
that there will be some urban and rural NCore monitoring sites with PM10-2.s speciation, what
other factors should be considered in selecting the pilot monitoring and long-term sites or
locations?
Are we picking sites based on a criterion like this rather than on a good exposure science basis?
Do we want PM coarse samples for the sake of having a clean sample or are we sampling
because there are exposure, source identification, or transport issue to be solved? This question
is very strange. Without a clear set of objectives for what we are planning to use the data for, it
is impossible to judge the value of the sites available.
55
-------
2. If there is an opportunity to modify the NCore PM'10-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For example,
changing the total number of required monitors and/or the required locations?
Of course. At this point, when we have very limited knowledge of PM coarse, we are clearly
going to make judgment calls that subsequent data will almost certainly tell us that some of our
choices were wrong. We must always maintain flexibility in the network design. That was part
of the NCore philosophy.
56
-------
Multi-Filter Coarse Particle Chemical
Speciation Sampler Using Virtual
Objective:
Develop a Coarse Particle Chemical
Speciation Sampler to Obtain Filter
Samplers Similar to the Current Fine Particle
Chemical Speciation Sampler Used In
EPA's Fine Particle Monitoring Network with
the Goal of Obtaining a Nearly Complete
Mass Balance on the Collected Coarse PM
-------
Design
Premise
- Need to Separate Fine and Coarse Particles for All the Usual Reason
Better Understand Health Effects of PMc vs PMf
Source Apportionment
Not Mixing Fine and Coarse Chemistry (fine acidic, coarse basic)
- Obtain Separation with a Virtual Impactor
- Design Around the Current PMc and PMf FRM Flow Rates (16.7 LPM),
Option for 6.7 LPM to Match the Current MetOne PMf Speciation
Sampler
- Include Design Flexibility for Additional Measurements
Flexibility
- Allows Multiple Filters, Denuders, Back Up Filters as Needed to
Minimize Interferences
- Bypass Allows for Large Samples to Be Collected for Toxicological
Testing, Organic Speciation, Trace and Ultra-Trace Metals Speciation,
Including Valance States
- Virtual Impactor Can Have 5% Coarse, 95% Fine to Minimize Gas
Phase Artifacts on Coarse Filter*
-------
26
20
Existing Commercial
High Volume Fine and
Coarse Sampler Based
on Virtual Impaction
Modified Sampler for PMc Chemical Speciation
General Schematic
Design and Evaluation by Clarkson
University and U.S. EPS Underway
Currently.
Commercialization by Greentech
Environmental. Available
Prototype for OAQPS Testing, May
15, 2009
1000 LPM
PM 10 INLET
Virtual Impactor
90% Fine, 10% Coarse"
Fine
Particles
Coarse
Particles
Flow Splitters
rj 16.7 LPM, 47 mm Filters
V/////A 49.1 LPM, 90 mm Filter bypass flow
833 LPM, 203 mnm X 250 mm Filter
bypass flow
-------
Dr. Rudolf Husar
PMm-2 s Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM 10-2.5 species or components missing
from this list? Are there important analysis methods missing from this list?
No
2. In the consideration of potential ion measurements for PM 10-2.5 species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM 10-2.5?
Nitrate is important
3. The 2004 CD included a list of important PMw-2.s components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
Not familiar with bio aerosol sampling issues
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
Saturation is real. Keep deposition density low?
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PMio-2.sfor OC and EC given the large expected soil component? If so, how should this
interference be addressed?
I am not familiar with oxides-OC/EC interference.
60
-------
Dr. Donna Kenski
The questions on PMio-2.5 speciation measurement and on PMio-2.5 species or components are
outside my areas of expertise and I respectfully defer to the distinguished panel members'
opinions on these subjects. With regard to network design, I've answered the charge questions
below, but I feel compelled to note that, given the current lack of consensus on how best to
sample for both PMio-2.5 mass and its speciated components, it seems much too premature to
consider rolling out a network of 75 monitors by 2011, just 2 years from now. This is potentially
a huge commitment of dollars and state resources that, once launched, will be very difficult to
steer in a different direction if future developments warrant. The purpose of this pilot network is
for research into coarse particle concentration distributions, sources, and health effects, which is
commendable. But we need to allow enough time for the pilot results to be analyzed and filtered
through the scientific community before launching a long-term network. A reasonable pilot
network would be more like 5 or 10 sites where intensive sampling can take place (maybe even
every day instead of every third day at some sites), characteristics of multiple monitors can be
intercompared, and options for speciation of those multiple monitors can be explored and vetted
until the research community has more collective confidence in the best way to move forward.
If the purpose of speciation is to perform source apportionment, I agree with several panel
members who have noted that the bulk components (OC, EC, ions, metals) will not be sufficient
to separate some important sources of PMio-2.5. It is possible that organic carbon speciation
would help resolve some of these sources. While this is unlikely to be feasible as part of the
routine network, it should have a place in the pilot network since it would shed some light on
whether source apportionment with the bulk species is able to resolve carbon sources
satisfactorily. This should include source sampling for the PMio-2.5 fraction as well, although
that's not necessarily a network function. Continuous sampling for PMio-2.5 mass and
components (at least EC/OC) would certainly be helpful as well.
Network design
1. Are sites with high PM10 and low PM2.5 good candidate sites for PM10_2.5 speciation? Given that there
will be some urban and rural NCore monitoring sites with PM10-2.s speciation, what other factors should
be considered in selecting the pilot monitoring and long-term sites or locations?
One should aim for a selection of sites that represent varying ratios of PMi0:PM2.5, as well as
varying geographic distribution because the composition of both particle size fractions will vary
geographically as well. It will be important to include both coastal and inland sites, urban and
rural (not just pristine wilderness areas but agricultural areas also), and various meteorological
regimes. It seems like most of the NCore sites as currently proposed are generally population-
oriented PM2.5 sites. Sites that reflect maximum and representative PMio-2.5 exposures will not
necessarily coincide with those PM2.5 sites, so the pilot program should have the flexibility to
monitor at locations outside the NCore network. Based on data collected in the next few years,
informed decisions could then be made about long-term sites.
61
-------
2. If there is an opportunity to modify the NCore PM10_2.5 speciation monitoring requirements during a
future rulemaking, should changes to the network design be considered? For example, changing the total
number of required monitors and/or the required locations?
Of course changes to the network should be continually under consideration; hopefully that's an
important objective of a pilot network. Learn what works and what doesn't and adjust
accordingly. As noted above, it seems much too soon to contemplate 75 speciation sites. I
strongly encourage EPA to move ahead slowly and evaluate the pilot study data thoroughly
before expanding the network beyond a handful of research sites. The resources spent on
monitor and method intercomparisons will be a much better investment of our scarce monitoring
dollars than a hasty move to monitor at all 75 NCore sites. Let the research community evaluate
these results, come to agreement on speciation sampling and analysis protocols, then move
forward with a broader network that's truly science-based.
62
-------
Dr. Thomas Lumley
PMm-2 s Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM 10-2.5 species or components
missing from this list? Are there important analysis methods missing from this list?
Distinguishing burning biomass (wood or grass) from other carbon sources with levoglucosan
may be useful and relatively inexpensive.
2. In the consideration of potential ion measurements for PMw-2.s species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM 10-2.5?
Not my field.
3. The 2004 CD included a list of important PMw-2.5 components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy
(SEM) on Teflon filters sufficient to quantify and identify these species? Is the proposed total
protein assay technique important to obtain a quantitative indicator of the total biological
material present?
In areas where wood is used for heating, wood smoke is important to distinguish as there are
straightforward ways to reduce emission. Specific allergenic pollen sources may be important
when studying health effects of the coarse fraction.
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
Not my field
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PMw-2.sfar OC and EC given the large expected soil component? If so, how should this
interference be addressed?
Not my field
Network Design
1. Are sites with high PMw and low PM2.5 good candidate sites for PM 10-2.5 speciation? Given
that there will be some urban and rural NC'ore monitoring sites with PM 10-2.5 speciation,
what other factors should be considered in selecting the pilot monitoring and long-term sites
or locations?
63
-------
Speciation will be relatively straightforward, especially by the difference method, in sites with
high PM10 and low PM2.5. The sources and species contributing to the coarse fraction may well
be different in sites with high PM10 and low PM2.5. If so, it would be important to also carry out
speciation at other sites. It is entirely conceivable that most of the health impact of coarse PM
occurs at sites with moderate coarse:fme ratios rather than at sites with high peak concentrations
of wind-blown soil. The sites should be chosen to capture the range of types of coarse PM
exposure in the population.
2. If there is an opportunity to modify the NCore PM 10-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For
example, changing the total number of required monitors and/or the required locations?
It would be useful to add sites whose coarse-fraction composition is likely to be different from
those already in the network, especially if large numbers of people may be exposed. While it is
true that local sources have a larger contribution to coarse PM than fine PM, some of the highest
levels of soil-based coarse PM will affect fairly large areas (not necessarily at the same time).
Expanding the network is most useful when it captures qualitatively under-represented types of
coarse PM. More monitoring in populated areas California and the Southwest is likely to be
useful for understanding the health effects of coarse PM. Monitoring in areas with low
population density may be informative about sources but is unlikely to be useful for studying
health effects.
64
-------
Dr. Peter H. McMurry
PMm-2.s Speciation Measurement
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Are there
additional PMw-2.s target species or methods that can be used to help identify the source of
unidentified mass in order to obtain better mass closure?
Biological material is a poorly understood category of particulate matter that is likely to be
significant. Matthias-Maser and coworkers {, 2000 #11407} reported that during a complete year
of sampling in the Lake Baikal region, "20% of the number concentration and almost 30% of the
volume concentration were biological, with no seasonal variation." While the carbon content of
biological material ought to be detected by OC/EC analyses, it is important to identify the extent
to which bioaerosols contribute to paniculate mass if this is indeed such a major source. Fuzzi et
al. {, 1997 #6836} reported that "the concentration of airborne bacteria and yeasts is enriched in
foggy conditions up to two orders of magnitude compared to clear air conditions, while
concentration of moulds is not influenced by the presence of fog," suggesting that fog droplets
act as culture media for biological particles. Mohler et al. {, 2007 #21265} report that biological
particles may play an important role in cloud formation. Given the potential significance of
biological particles in coarse particles, an effort should be made to quantify their contribution to
mass. I am not sufficiently knowledgeable to provide useful input regarding suitable analytical
methods.
I agree with the concern expressed by several Committee members that OM/OC may be quite
different for particles of biological origin, secondary OC, and primary. An effort must be mede
to determine those ratios for biological particles.
2. Various sampling devices, including dichotomous samplers, MetOne SASS speciation
monitors, PM10 and PM2.5 FRMs are potential sampling devices (with the appropriate filter
types) for PM 10-2.5 speciation. Among these sampler types, which should be included or excluded
from the pilot network design? Are there other sampling devices not listed here that should be
considered?
I would not favor the use of indirect methods such as the GRIMM optical detector. The
correlations between GRIMM mass and gravimetric mass shown by Delbert Eatough in his
remarks are extremely impressive. However, assumptions about shape, density, and refractive
index must be made to convert optical signals to mass. In my experience, those aerosol
properties vary with sampling time and location and will inevitably introduce uncertainties.
I agree with Delbert Eatough's comments that chemical interactions between fine and coarse
particles could affect speciation measurements. This argues against using the PM10-2.5 FRM as
the speciation sampler. However, it is also true that if a perfect sampler were used to obtain
samples for speciation analysis, then it might not be possible to reconcile measured (FRM) mass
and reconstructed mass. On the other hand, speciation measurements done on FRM samples
could, in principle, produce mass closure. Is "truth" or "agreement" better? I am inclined to
agree with Prof. Eatough: If a dichotomous sampler produces results that are more nearly "true,"
65
-------
then it would be a preferable sampler.
During our conference call Warren White pointed out another reason to inertially separate coarse
and fine particles before collection. In previous meetings of this Committee, the point was made
that coarse urban particles may contribute more to adverse health outcomes than coarse nonurban
particles. This could be because urban coarse particles serve as carriers for toxic compounds
picked up by exposure to the urban environment. If those toxic compounds were also present in
fine particles at comparable or higher concentrations, then measurements by difference would
lead to substantial uncertainties.
3. What are the PM'10-2.5 speciation sampling artifacts that may be encountered using the
samplers mentioned above and how should they be addressed? Is speciation by the difference
method problematic for PM 10-2.5 speciation and if so what specific issues make it problematic?
Volatilization and chemical reactions due to interactions of basic and acidic particles; inaccurate
determination of the fraction of collected mass that is water.
4. The current and most widely usedPM2.s speciation sampler is the MetOne SASS and it has a
flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRMfor
PM 10-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for PMw-
2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PMw-2.s FRM?
I defer.
5. Is the amount of particle mass collected on the dichotomous filters (especially the minor
flow) sufficient for speciation chemical analysis?
If the sampling rate into the dichot is 16.7 1pm, then the sampling rate for coarse particles would
be the same as for the FRM. The dichot would offer the advantage that coarse and fine particles
are not combined to the same extent as occurs on the FRM.
PMm-2.s Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM'10-2.5 species or components missing
from this list? Are there important analysis methods missing from this list?
I feel that highlighting coarse particles of biological origin is appropriate. I am not an expert on
biological particles or on ways to detect them. Eric Edgerton has worked on this topic and made
some interesting recommendations for chemical assays that might provide further information on
biological particles.
The document also suggested that SEM might be used for biological particles. Indeed, visual
observations can provide useful insights into possible origins of particles. However, I am not
familiar with automated SEM techniques that have proven reliable for providing the type of
66
-------
quantitative information that might be expected for a monitoring network. I feel that limited use
of SEM in research by knowledgeable persons should be encouraged. Sampling times and
samplers should be optimized for this purpose. It seems unlikely that heavily loaded PM10
samples would be optimal for SEM analysis.
2. In the consideration of potential ion measurements for PMw-2.s species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM'10-2.5?
George Allen's admonition that coarse particles are different from fine particles is important. As
Bart Croes points out, nitrates (especially sodium nitrate) can be a significant coarse particle
component, especially near the coast. Judy Chow's point that magnesium can be a useful tracer
for agricultural activities is also persuasive.
3. The 2004 CD included a list of important PM 10-2.5 components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
I do not see a clear reason for separating out fly ash as a source category. My other thoughts
pertinent to this question are given above.
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
I defer.
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PMio-2.sfor OC and EC given the large expected soil component? If so, how should this
interference be addressed?
I defer.
Network Design
1. Are sites with high PMW and low PM2.s good candidate sites for PM 10-2.5 speciation? Given
that there will be some urban and rural NCore monitoring sites with PM 10-2.5 speciation, what
other factors should be considered in selecting the pilot monitoring and long-term sites or
locations?
Other committee members provided insightful responses to this question.
67
-------
2. If there is an opportunity to modify the NCore PM'10-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For example,
changing the total number of required monitors and/or the required locations?
I defer.
68
-------
Mr. Richard L. Poirot
PM10-2.5 Speciation Measurement
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Are there
additional PM 10-2.5 target species or methods that can be used to help identify the source of
unidentified mass in order to obtain better mass closure?
Generally the list of proposed species seems OK, although if the difference method is being
considered for sampling, this list may require more samplers than is feasible. It may not make
sense to more or less duplicate the PM2.5 speciation analysis. Possibly it will be more efficient
to include some relatively basic "screening" sampling at most coarse speciation sites , with more
detailed analyses at selected sites and/or based on the screening analysis. I wonder how much
useful information will really be provided by repeating the same TOR & TOT analyses (with the
exact same thermal cut-points) for OC & EC? Some molecular level carbon analysis could be
revealing at a few sites. A measurement of light absorption from Teflon dichot filters might be a
useful, low cost addition to consider. Possibly the ion analysis may not require nylon filters or
denuders, as we really don't expect much coarse ammonium nitrate, nor are there really any very
satisfactory ways to quantify ambient ammonium concentrations. I don't know much about
sampling for biological materials, but think there must be other useful options beyond SEM and
Protein Assay. I'm not sure that SEM on Polycarbonate filters for fly ash can really be justified,
except at sites where a substantial fly ash contribution is otherwise indicated. Possibly also some
sort of toxicity testing might be conducted on (seasonally composited, coarse-only) samples to
get a sense of which sites/seasons have the most harmful coarse particle mixtures. Then
chemical &/or biological analyses could be focused to the samples of interest.
Regarding mass balance issues, I think (assuming crustal oxides are appropriately estimated) that
a large and likely variable fraction unmeasured of non-carbon organic matter could be a
significant contributor to mass. For example the C:mass ratios for carbohydrates & cellulose are
2.4 & 2.5, and carbonate - likely to be present in many coarse samples - has a C:mass ratio of 4.
Rather than adding additional species in ambient sampling, it might be more efficient to analyze
known quantities of different source material by comparable methods to get a better sense of
how the easily measured species relate to the masses.
2. Various sampling devices, including dichotomous samplers, MetOne SASS speciation
monitors, PM10 andPM2.5 FRMs are potential sampling devices (with the appropriate filter
types) for PM10-2.5 speciation. Among these sampler types, which should be included or
excluded from the pilot network design? Are there other sampling devices not listed here that
should be considered?
I think a sampler such as the one Phil Hopke suggests, that allows for both size-separated
particles on multiple filter types, and also has the option to collect relatively large sample sizes to
support other more detailed analyses could be useful to test in the pilot study. It could also be
useful to evaluate different analyses on different filter media. For example, how does nitrate on
coarse dichot quartz and Teflon compare with nitrate on (denuded) dichot Nylon? It would also
69
-------
be useful to evaluate any continuous speciation instruments, such as the paired PM10 and PM2.5
Sunset Labs carbon analyzers - vs filter samples. While there are so many benefits to the use of
dichotomous samplers, one possible disadvantage is that the exposed coarse Teflon filters -
without the sticky fines - may not "travel well". Aside from just comparing samplers & filters, it
might be useful to look specifically for potential sample loss from coarse dichot filters with
rough handling, and to develiop best practices to minimize such losses (if any).
3. What are the PM10-2.5 speciation sampling artifacts that may be encountered using the
samplers mentioned above and how should they be addressed? Is speciation by the difference
method problematic for PM10-2.5 speciation and if so what specific issues make it problematic?
As indicated above, we really don't expect much coarse ammonium nitrate or ammonium sulfate
and so Nylon filters and denuders may not be needed (nor would ammonium measurements like
reflect ambient concentrations). Coarse nitrate might be measured directly from Quartz or
Teflon dichot filters. Loss of some chloride is likely from samples with sea salt or road salt.
OC artifacts are likely to be proportionately less than for fine particle sampling, but a careful
effort should be made to quantify OC blank concentrations.
Speciation by difference should work reasonably well for any species with relatively high
concentrations in the coarse mode, and poorly for any species with relatively high concentrations
in the fine mode.
4. The current and most widely usedPM2.5 speciation sampler is the MetOne SASS and it has
a flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRMfor
PM10-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for
PM10-2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PM10-2.5 FRM?
This may be a concern, but I defer to others for an answer. One potential "advantage" for coarse
speciation is that we are not starting out with a strong advance signal that the mass of coarse
particles is a good health indicator. Achieving a mass balance is less important than
understanding where the various components of coarse particles come from and which are most
injurious to health (or welfare).
5. Is the amount of particle mass collected on the dichotomous filters (especially the minor
flow) sufficient for speciation chemical analysis?
I don't think this should be an issue, except for analyses such as molecular carbon compounds or
specific biological components which may require very large sample sizes.
PM10-2.5 Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM 10-2.5 species or components
missing from this list? Are there important analysis methods missing from this list?
70
-------
See answers to Q 1 under Speciation Measurement Issues. My preference for a minimum Coarse
speciation measurement (for the most sites) would be a dichot with XRF and optical absorption
on a Teflon coarse filter. My next addition would be a dichot with quartz filters for carbon and
ion analysis. Paired (difference method) or virtual impaction (coarse only) mass and OC/EC by
continuous methods may also be a valuable complement to provide better time series
information. Toxicity assays and/or molecular carbon analyses might be considered on
seasonally-composited samples at some sites.
Specific sampling and/or analyses for pollen - or pollen components (often fragmented into or
aerodynamically acting like particles smaller than 10 microns) may be useful at some sites or
seasons. Possibly, sampling for biological analyses may require other methods (see #3 below) for
which absolute particle size cut is less critical than ability to culture or bioassay, and reasonable
assumptions can be made about the domination of coarse mode formation or suspension
mechanisms.
2. In the consideration of potential ion measurements for PM10-2.5 species, what ions should
be on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM10-2.5?
See previous answers. I don't see why there should be much coarse ammonium nitrate in the
ambient air, and I expect coarse nitrate compounds (NaNO3, CaNO3, etc.) to be less volatile -
and therefore likely not needing denuders or Nylon filters. But do the necessary experiments to
confirm or disprove this in the pilot network.
3. The 2004 CD included a list of important PM10-2.5 components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
As previously indicated, I'm not sure fly ash SEM is generally warranted unless otherwise
indicated. For more input on appropriate routine and/or at least occasionally feasible biological
(sampling and) analyses, you may want to consult beyond this subcommittee. See for example
Dale Griffin's (USGS) paper on: Atmospheric Movement of Microorganisms in Clouds of
Desert Dust and Implications for Human Health, Clinical Microbiology Reviews, July 2007, p.
459-477, Vol. 20, No. 3, http://cmr.asm.org/cgi/content/full/20/3/459?view=long&pmid=17630335
In the arid Southwest some specific analyses for Coccidioides fungi could be revealing. See for
example Mark Bultman's (USGS) poster at:
http://www.fws. gov/southwest/Climatechange/poster%20pdfs/VallevFeverPoster_Bultman.pdf
As indicated above alternative sample collection may be warranted for certain kinds of biological
analysis.
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
71
-------
Not my area of expertise. I would imagine that reasonable correction factors can be developed.
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PM 10-2.5 for OC and EC given the large expected soil component? If so, how should this
interference be addressed?
Not my area of expertise. I would imagine that reasonable correction factors can be developed.
Network Design
1. Are sites with high PM10 and low PM2.5 good candidate sites for PM10-2.5 speciation ?
Given that there will be some urban and rural NCore monitoring sites with PM10-2.5 speciation,
what other factors should be considered in selecting the pilot monitoring and long-term sites or
locations?
Sites with high coarse to fine ratios are likely to be dominated by crustal material, and will not be
ideal places to evaluate strengths and weaknesses of different sampling and size-separation
techniques. However, I think a few such sites should probably be included in both the pilot and
longer-term networks and specifically include sites heavily influenced by: windblown dust,
agricultural activities, mining activities and industrial process emissions. In the pilot study, it
would also be useful to challenge the size separation methods by including sites with relatively
high fine to coarse ratios. It would also be desirable to include sites intended to collect and
characterize urban roadway emissions, in both arid and humid regions. Some effort might also
be made in the pilot study toward understanding spatial variability and the differences in coarse
concentration and composition that might be expected at NCore locations compared to locations
closer to sources. A few rural coastal sites on the West Coast in the far Southeast - such as the
Redwood, Point Reyes, Everglades or Virgin Island National Park IMPROVE sites - might be
useful to best characterize the size distributions, chemical and biological content of Asian and
African dust.
2. If there is an opportunity to modify the NCore PM10-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For example,
changing the total number of required monitors and/or the required locations?
Given the many questions & issues raised in the EPA staff memo and by the members of this
review committee regarding sampling and analytical methods, analytes and site characteristics, it
may not be possible or desirable to roll out a network of (identical) coarse speciation samplers at
all 75 NCore locations by 1/1/11. I think the NCore concept was intended to be flexible, and that
pending analysis of existing dichot data (for example from long-term California and Canadian
studies), additional results from the current EPA (Bob Vanderpool) coarse particle monitoring
sites, the planned pilot coarse speciation network, and other developments, it may be logical to
consider changes to the proposed NCore coarse particle speciation network.
Possibly, many of the NCore sites which are otherwise very well suited for assessment of
multiple pollutant exposures may not be ideally sited for measuring coarse particle
concentrations, compositions and representative exposures. I think it would be useful to
72
-------
intentionally locate some coarse speciation sites in source-oriented locations, as indicated above.
Along similar lines, it might make sense to plan for occasional "event analysis" of samples from
many sites on days when certain source influences are otherwise evident - for example wild fires
or prescribed agricultural burns, "local dust storms" from different regions, Asian or African dust
storms, agricultural tilling, pollen or insect events, etc. Seasonal compositing of samples may
also be a useful technique to allow more sophisticated measurements such as for molecular
carbon compounds, which require large sample sizes or are too costly for routine application.
I think an important complement to a coarse speciation network, and something that would
enhance the value of existing fine speciation monitoring sites as well, would be to conduct an
emissions characterization program in which source emissions are characterized by exactly the
same sampling and analytical methods that are applied for ambient sampling (including both fine
and coarse fractions which together provide better detail on sources). This might include a mix
of direct emission sampling (with dilution / settling chambers) and "source oriented sampling -
such as near roadway, etc.
If a relatively large number of ambient monitoring sites is considered important, it might be
useful to consider a "nested" network, with many sites initially taking a "minimal" set of
measurements (for example 1 in 3 day dichots with Teflon filters and XRF and optical
absorption analysis - or possibly a second dichot with quartz filters for carbon and ion analysis).
Additional samplers, higher sample frequencies, continuous analyzers, and/or additional analyses
might be added at smaller numbers of sites or to provide more detail at times and places where
the initial measurements show interesting results. As suggested above, it might be informative to
conduct some screening toxicity tests on seasonally composited samples to determine locations
where more detailed chemical or biological sampling and analyses are most warranted.
Although there was never sufficient funding provided to support NCore "Level 1 sites", the
concept is a good one and it might be especially applicable for coarse particle speciation
monitoring, as it seems likely there will be a lingering need to develop and evaluate new
sampling and analytical methods (at a few sites) even while less intensive "routine" coarse
species monitoring is conducted at larger numbers of Level 2 sites.
73
-------
Dr. Jay R. Turner
Implementation of a monitoring network for coarse particle (PMio-2.5, called PMc in my
comments) speciation will have many benefits including, but not limited to, information about
emission sources and their relative contributions to burdens observed at the monitoring sites.
The Agency is urged to make steady progress on the development and deployment of this
network, but should also conduct the necessary pilot studies and ancillary evaluations (e.g.
optimization of analytical methods) to ensure the network, when fully deployed, is robust in
terms of sampling methodology and monitor siting. It might be necessary to make revisions to
the pilot study design based on preliminary findings from the pilot network towards optimizing
the measurement strategy. In my opinion, optimization of sampling hardware and monitor siting
guidance (based on early results from the pilot network) should take precedence over meeting the
January 1, 2011, date for full deployment of the PMc network at NCore sites. If the deployment
date is delayed, the Agency should provide an explanation of the reason(s) for the delay, the
remaining work to be done, and the revised timeline for network deployment.
PMm-2.s Speciation Measurement
1. Table Iprovides a list of proposed PM 10-2.5 species and analysis methods. Are there
additional PMw-2.s target species or methods that can be used to help identify the source of
unidentified mass in order to obtain better mass closure?
The species and analysis methods in Table 1 (amended from the white paper, and referring to the
Agency presentation at the February 9 teleconference) meet two objectives - identifying the
major chemical components needed to reconstruct PMc mass (as stated in the charge question)
and providing additional information about emission sources and their relative contributions.
Priority should be given to the analyses needed for mass balance closure and this subset of the
list should contain the conventional analyses for gravimetric mass, water soluble ions, XRF
elements, and carbon by thermal-optical analysis. Carbonate carbon will influence the OM/OC
ratio but is expected to be important for only a limited number of environments. Thus, the
carbonate carbon analysis is recommended for the pilot network but if it is expensive then it
might be possible to include it at only a subset of sites in the full network. The protein analysis is
also recommended for the pilot network with an evaluation of whether these data bring sufficient
added value to be performed for the full network.
2. Various sampling devices, including dichotomous samplers, MetOne SASS speciation
monitors, PMw andPM2.s FRMs are potential sampling devices (with the appropriate filter
types) for PM 10-2.5 speciation. Among these sampler types, which should be included or excluded
from the pilot network design? Are there other sampling devices not listed here that should be
considered?
All of these samplers have advantages and disadvantages as has been discussed in previous
deliberations of this subcommittee. Paired PMi0 and PM2.s FRMs might be most challenged at
sites with relatively high PM2.5 /PMio ratios because the measurements uncertainties propagate
for these independent measurements. This does not necessarily preclude the use of paired
74
-------
and PM2.5 FRMs, but their suitability should be evaluated within the context of the DQOs and
MQOs. For example, if source apportionment is an objective then the error structure in the data
would be an important consideration. (We have a small data set of XRF elements for filter
samples collected in St. Louis using paired PM2.5 and PMi0 Harvard Impactors that can be shared
with the Agency; it can be used to demonstrate limitations of the differencing method; there are
additional data set, including sets collected by the Agency, which can be used to probe this
issue). At this time, I give priority to dichotomous samplers. Historically, one concern has been
the potential for losses from the coarse particle filters during handling and shipping since the fine
particles present in a conventional PMio sample act as "glue". I recall the Agency examined this
issue in their recent PMc characterization field campaign. If it is deemed an issue there might be
simple ways to stabilize the deposit on the filters prior to shipping. Samplers with multiple
sampling channels, such as used in the CSN network, are preferred because they allow the
collection of particles on different filter media using a single sampler. However, the use of
parallel samplers (a one sampler-one substrate approach), such as used in the IMPROVE
network, has proven to be robust. In either case, the design should be sufficiently flexible to
permit the use of upstream sample conditioning such as denuders. For mass balance closure, the
low flow rate of the MetOne SASS speciation monitor should not be a major issue since the
focus will be on the major chemical constituents. However, there could be issues with detection
limits for the trace elements that are often important for emission source identification, especially
if paired with XRF instead of ICP-MS.
3. What are the PM'10-2.5 speciation sampling artifacts that may be encountered using the
samplers mentioned above and how should they be addressed? Is speciation by the difference
method problematic for PM 10-2.5 speciation and if so what specific issues make it problematic?
It will be helpful to understand how the virtual impactor cutpoint curve compares to the PM2 5
FRM cutpoint curve. The PM2 5 FRM impactor has been characterized in detail, and we know
that the cutpoint curve is relatively broad (by design). The dichotomous sampler (or any other
sampler being deployed) should also have a well-characterized cutpoint with this information
ideally available when interpreting data from the pilot network, The white paper addresses
potential artifact issues concerning carbon and nitrate; perhaps the nitrate is less of an issue if the
counterion is not ammonium but rather sodium or potassium but it is not clear to me we can
generalize our understanding of the chemical form of coarse PM nitrate based on the data
collected to date. It would be ideal to evaluate the importance of denuders as part of the pilot
network measurements.
4. The current and most widely usedPM2.s speciation sampler is the MetOne SASS and it has a
flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRM for
PM 10-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for PMw-
2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PMw-2.s FRM?
The low flow rate for the MetOne SASS might indeed be an issue for trace elements if XRF is
the de facto analytical method, further compounded if using the difference method. The low
flow rate might be less of a concern if ICP-MS or a similar high-sensitivity method is used, but
in this case there will be other tradeoffs regarding element-specific recoveries, cost, etc.
75
-------
Certainly another layer of complexity is added when comparing reconstructed mass from
samples collected at one flow rate to the mass collected by the PMi0-2.5 FRM at a different flow
rate, with this being more of a sampling artifact issue than a detection limit issue.
5. Is the amount of particle mass collected on the dichotomous filters (especially the minor
flow) sufficient for speciation chemical analysis?
The minor flow still has all of the coarse PM pulled through the inlet (i.e. upstream of the virtual
impactor).
PMm-2 s Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM 10-2.5 species or components missing
from this list? Are there important analysis methods missing from this list?
The core species measurements should be elements, ions, and ECOC, consistent with the PM2.5
speciation network. A key issue is whether XRF can provide adequate data quality for some of
the trace elements for conventional mass loadings on the filters. This is less of an issue for mass
closure but might be important for source identification.
2. In the consideration of potential ion measurements for PMw-2.s species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM 10-2.5?
The target list for ions is reasonable. While conventional wisdom says a denuder is needed for
high quality nitrate measurements, work performed for the IMPROVE program has shown that
certain inlets can be quite effective denuding nitric acid (although I don't know about the
extrapolation of these results to urban environments). Sampling with and without denuders
should be performed in the pilot network to determine whether they are warranted. Also, Nylon
filters should certainly be used in the pilot network and then an evaluation can be made
concerning the importance of ammonium nitrate.
3. The 2004 CD included a list of important PMi0-2.s components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
I do feel strongly that we need a better characterization of the bioaerosol component. However, I
have no specific comments on the method. Analysis for fly ash by SEM could be insightful, but
perhaps used only in cases where trace element compositions suggest there are potentially
significant fly ash contributions to PMc (I repsct there might be some issues in teasing out a fly
ash signature from a soil signature from the elemental data).
76
-------
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
I leave this question to those colleagues with more expertise than me on this matter.
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PMio-2.sfor OC and EC given the large expected soil component? If so, how should this
interference be addressed?
It is generally accepted that the presence of such compounds can dramatically shift the carbon
evolution profile during thermal analysis - whether this leads to errors in the EC/OC split is not
clear. While not focusing on PMc, one option might be to examine data collected from
continuous ECOC instruments in environments with high metal oxide loadings. Does the
relationship between thermal EC and optical EC change in time with increasing time after a
clean filter has been installed (presumably some of the metal oxides will remain on the filter and
accumulate in time).
Network Design
1. Are sites with high PMw and low PM2.s good candidate sites for PM'10-2.5 speciation? Given
that there will be some urban and rural NCore monitoring sites with PM 10-2.5 speciation, what
other factors should be considered in selecting the pilot monitoring and long-term sites or
locations?
The focus should not be exclusively on sites with high PMio and low PM2.5. The pilot sites
should be chosen to represent a broad range of aerosol compositions and environmental
conditions, with the results being used to refine the recommendations for which Ncore sites
should have PMc speciation. There are other considerations, such as providing data sets of use to
health scientists, which would support the selection of sites to collectively include a range of
PMc levels.
2. If there is an opportunity to modify the NCore PM 10-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For example,
changing the total number of required monitors and/or the required locations?
See my introductory comment which addresses the timeline for network deployment. At this
point, I have no additional comments about the network design.
77
-------
Dr. Warren H. White
PMm-2.5 Speciation Measurement
I will start by recognizing, and commending, the Agency's early consultation with this
committee and its plan to start with a pilot study. An eventual network of 75 sites sampling at
least one day in every three will represent a huge commitment of resources that might otherwise
be devoted to more time-resolved measurements, or to biochemical assays and other new
approaches to sample characterization, or to nuclei counters and other instruments more
responsive to ultra-fine particles. Any one of these alternatives to the straightforward extension
to PMio-2.5 of current PM2.5 measurements might yield greater insight into which particles should
be of greatest concern to regulators. I urge the Agency to delay the general introduction of PMio-
2.5 monitoring at NCore sites until the pilot study has had adequate time to develop, explore, and
assess a variety of sampling and analytical methods.
"Since EPA is requiring PM10-2.5 speciation monitoring primarily for scientific purposes..."
It's not at all clear to me what those scientific purposes would be - source attribution?
Assessment of hazardous components as called for by the NRC (2004)? The methods under
discussion, largely derived from the existing PM2.5 speciation network, don't seem very well
adapted to either of these.
Source attribution: We have considerable experience with source attribution using speciated
PM2.5 data. That experience has taught us that it is hard to distinguish between different
carbonaceous sources using EC and OC from thermal-optical analysis, and that it is hard to
distinguish between different sources of dust using elemental analysis. But organic and crustal
materials make up the bulk of PMio-2.5 nearly everywhere. So what can the proposed level of
speciation do? It can distinguish smoke from dust, but how often do local agencies encounter
high PMio-2.5 that they can't easily associate with obvious local and regional sources of smoke or
dust?
Hazard assessment: Its 2005 assessment of the epidemiological evidence lead EPA to
"consideration of thoracic coarse urban particulate matter (UPMj 0-2.5) as an indicator for a
thoracic coarse particle standard ... qualified so as to include any ambient mix of PM 10-2.5
that is dominated by re suspended dust from high-density traffic on paved roads andPM
generated by industrial sources and construction sources, and to exclude any ambient mix of
PMio-2.5 that is dominated by rural windblown dust and soils and PM generated by
agricultural and mining sources. "
In other words it judged that dusts of crustal origin, benign in rural settings, became
contaminated in cities with surface coatings and other urban grime. It will be hard to detect such
trace contaminants in PMio-2.5 samples with the bulk chemical analyses currently used with
PM2.5; it will be still harder against a background of chemically-similar PM2.5 mixed in with a
PMio sample. The pilot study should give particular attention to analytical methods that can
target specific fractions expected to be enriched in PMio-2.5, such as biological markers.
78
-------
Reference
NRC (2004) Research Priorities for Airborne Paniculate Matter IV: Continuing Research
Progress. National Academies Press, Washington.
79
-------
Dr. Yousheng Zeng
PMm-2 s Speciation Measurement
Among the four filter-based sampler types, dichotomous samplers should be used for speciation
measurement. Dichotomous samplers produce physically separate coarse samples for speciation
chemical analysis, which in turn reduces interferences and complications. They have been used
for long time and are better understood.
Although the difference method is FRM for coarse particle mass measurement, it may not be
most suitable for speciation measurement. When the levels of a component (or species) in the
PM2 5 sample and PMi0 sample are close (i.e., the component primarily associated with PM2 5),
the difference between the two measurements may be too small compared to analytical
uncertainty, making the speciation measurement unreliable. This problem is more significant
when the chemical analysis has a sufficient method detection limit, but poor precision
(repeatability). In that case, if dichotomous samplers are utilized, the results will be more
reliable.
Even though dichotomous samplers are not FRM for PM mass measurement, their deviation
from FRM, if any, is small. Considering uncertainties associated with many chemical analyses
at low concentration levels, this possibly small deviation should be acceptable.
-2s Species
Speciation for organic portion seams lacking. If feasible, it would be useful to analyze for semi-
volatile organic species and parameters typically obtained in ultimate analysis. The semi-volatile
organic species should be useful in source apportionment. Also, ultimate analysis may help
close the mass difference. I am not sure this will be feasible or practical. I just want to bring this
up for EPA to consider for the pilot sites.
Network Design
For selection of pilot sites, high PMio and low PM2.5 sites should give more PM coarse fraction
to work with. However, I think it is also important to select some pilot sites that represent low
PMio and high PM2.5. At these sites the PM coarse fraction would be low. It would be a good
test to see how the speciation methods perform when the coarse fraction is low (e.g., will method
detection limits be low enough to cover some important species). Also, when EPA considers
high or low PMio or PM2 5, EPA should consider not only the ratio between PMio and PM2 5, but
also the mass levels of PMio and PM2.5. Both high mass level and low mass level sites should be
included in the pilot program to fully assess candidate speciation methods.
80
-------
Dr. Barbara Zielinska
PMm-2 s Speciation Measurement
1 . Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Are there
additional PMw-2. s target species or methods that can be used to help identify the source of
unidentified mass in order to obtain better mass closure?
In my opinion, Table 1 is missing the speciation of PMio-2.5 organic compounds. Taking into
account that organic carbon (OC) constitutes significant portion of PMi0-2.5 mass, it is essential to
have a general understanding what species comprise this organic mass. Are these species mostly
of biogenic origin, biological material, or perhaps also anthropogenic? This knowledge may
have important health implications and so far is missing. Since the primary objective for PMio-2.5
speciation data is to support further research in understanding their chemical composition and
source, the importance of the organic mass speciation should not be overlooked. It may help in
source apportionment and in determining OM/OC factor. The organic mass speciation should be
attempted during pilot study phase in some selected NCore sites that are representative of urban
and rural settings. The EPA should not specify target organic compounds to be quantified (as it
has been done so far), but should require the identification of as many organics as possible to get
some knowledge of the organic mass composition of coarse particles in different locations. PMi0.
2.5 is very different from PM2.5 and the source tracers that worked for fines may not work for
coarse particles.
2. Various sampling devices, including dichotomous samplers, MetOne SASS speciation
monitors, PMw andPM2.s FRMs are potential sampling devices (with the appropriate filter
types) for PM 10-2.5 speciation. Among these sampler types, which should be included or excluded
from the pilot network design? Are there other sampling devices not listed here that should be
considered?
and PM2.5 FRM are not appropriate for speciation of PMio-2.5, since when the fine and
coarse particles are mixed together, their composition can be potentially altered. Also, the
determination of chemical composition by a difference method is inherently less precise and it is
virtually impossible for organic compound's speciation. Although volatility issues are likely less
important for PMio-2.5, they are important for PM2.5, which may bias the results of the difference
method. Dichotomous sampler (or equivalent), which provides a separate PMio-2.5 fraction are
more appropriate for chemical speciation.
In addition, continuous methods for mass measurements and OC/EC speciation should be
encouraged.
3 . What are the PM 10-2.5 speciation sampling artifacts that may be encountered using the
samplers mentioned above and how should they be addressed? Is speciation by the difference
method problematic for PM 10-2.5 speciation and if so what specific issues make it problematic?
See my answer to Question #2.
81
-------
4. The current and most widely usedPM2.s speciation sampler is the MetOne SASS and it has a
flow rate of 6.7 Liters per minute (Lpm) which is significantly lower than either the FRMfor
PM 10-2.5 mass or the dichotomous sampler (16.7 Lpm). If this sampler was configured for PMw-
2.5 by difference, would the 6.7 Lpm flow rate be problematic, especially with the need to
compare reconstructed mass to the mass collected by the PMw-2.s FRM?
As stated above, I do not recommend speciation by the difference method. Dichotomous
sampler with 16.7 Lpm flow rate is a better choice.
5. Is the amount of particle mass collected on the dichotomous filters (especially the minor
flow) sufficient for speciation chemical analysis?
Since the total mass collected in the minor flow channel is a function of the inlet flow, this
should not be a problem.
PMm-2.5 Species or Components
1. Table 1 provides a list of proposed PM 10-2.5 species and analysis methods. Among these
species, which are most important? Are there important PM 10-2.5 species or components missing
from this list? Are there important analysis methods missing from this list?
Speciation of organic compounds is missing (see my answer to question #1 above). Such
speciation is important from health effects point of view, as well as source attribution and mass
closure.
2. In the consideration of potential ion measurements for PMw-2.s species, what ions should be
on the target list? Are nitrate or ammonium ions important? If so, is an acid gas denuder and
nylon filter required for the proper collection of these species in PM'10-2.5?
I don't think an acid gas denuder or nylon filters are required. The ions and their analysis
methods listed in the corrected Table 1 (slide #9 in Joann Rice presentations) seem to be
adequate
3. The 2004 CD included a list of important PMi0-2.s components which included biological
materials and fly ashes. If these species are important to characterize, what specific types of
biological materials and fly ashes should be included? Is scanning electron microscopy (SEM)
on Teflon filters sufficient to quantify and identify these species? Is the proposed total protein
assay technique important to obtain a quantitative indicator of the total biological material
present?
Fly ashes may only be important at sites close to some local sources (fly ashes were removed
from Table 1 presented by Joann Rice). SEM on Teflon filters is not ideal for identification of
biological materials - polycarbonate filters are better. Total protein assay may be important as an
indicator of allergens, fragments of bacteria, etc. However, this assay is not adequate as an
indicator of the total biological material present. Plant or soil derived material may be an
important contributor. Analysis of sugars, sugar alcohols, phytosterols, triterpenoids, etc., may
82
-------
help in quantification and identification of types of biological materials present in coarse
particles.
4. Can the complication of particle size and absorption effects in XRF be resolved using
absorption correction factors? If not, what other method(s) should be considered?
It would be desirable to compare ICPMS with XRF during a pilot study to answer this question.
5. Are metal oxides a significant source of interference in thermal-optical analysis (TOA) of
PMw-2.sfor OC and EC given the large expected soil component? If so, how should this
interference be addressed?
Metal oxides, if present in significant amounts may affect OC/EC split, however total carbon
should not be affected. Humic material or carbonates are another potential interfering species.
This question should be addressed during a pilot study.
Network Design
1. Are sites with high PMw and low PM2.5 good candidate sites for PM'10-2.5 speciation? Given
that there will be some urban and rural NC'ore monitoring sites with PM 10-2.5 speciation, what
other factors should be considered in selecting the pilot monitoring and long-term sites or
locations?
It would be important to include both types of sites, with low and high PMi0 to PM2.s ratios.
Especially in urban locations and/or close to highways both types may be important. Variations
in geographical distribution, costal and inland locations, urban and rural sites, proximity of
mining sources, etc., should also be considered.
2. If there is an opportunity to modify the NCore PM 10-2.5 speciation monitoring requirements
during a future rulemaking, should changes to the network design be considered? For example,
changing the total number of required monitors and/or the required locations?
Flexibility was one of the major components of NCore design, so yes. However, prior to
implementing 75 PMio-2.5 speciation sites, the throughout evaluation of pilot study data should be
completed. It seems that a goal of launching long-term PMio-2.5 network by 2011 is not very
realistic. We should learn first what works and what doesn't and allow enough time for scientific
community to evaluate pilot study data and come to a consensus regarding the best methods of
coarse particles monitoring and speciation.
83
------- |