MEMORANDUM

SUBJECT:	Key Issues Related to PM10-2.5 Speciation Monitoring

FROM:	Joann Rice, OAQPS/AQAD/AAMG

TO:	PM NAAQS Review Docket (EPA-HQ-OAR-2007-0492)

As part of the recent revision to the Ambient Air Monitoring Regulations
(U.S. EPA 2006a), PM10-2.5 speciation monitoring is required at National
Core (NCore) multi-pollutant monitoring stations by January 1, 2011. 
PM10-2.5 speciation monitoring may also be useful in other locations
where characterization of thoracic coarse particle speciation would be
of high value.  This paper describes the PM10-2.5 speciation monitoring
requirements specified in the ambient air monitoring rule and provides
an overview of the monitoring issues, discussion of the potential use of
existing PM10-2.5 speciation sampling and analysis techniques in a pilot
study to inform the implementation and decision-making process, and
related research questions to inform the planning and implementation
process.  This document also serves as a discussion piece for obtaining
feedback and comments on the development and implementation of a
long-term PM10-2.5 speciation monitoring program.

INTRODUCTION

The EPA issued revisions to the Ambient Air Monitoring Regulations (40
CFR Parts 53 and 58) on October 17, 2006 (U.S. EPA 2006a).  The final
rule establishes ambient air monitoring requirements for an indicator of
thoracic coarse particles (PM10-2.5) to support continued research on
particle distribution, sources, and health effects. At the same time,
EPA also promulgated a new Federal Reference Method (FRM) for measuring
the mass concentration of PM10-2.5 in ambient air.  Although EPA is not
adopting a National Ambient Air Quality Standard (NAAQS) for PM10-2.5 at
this time, the FRM for PM10-2.5 is of value to aid in a variety of
research studies and the development of speciation samplers capable of
providing improved characterization and understanding of the composition
of thoracic coarse particles.  

The final monitoring rule contains a requirement for PM10-2.5 speciation
at NCore multi-pollutant monitoring sites by January 1, 2011, with state
monitoring implementation plans due July 1, 2009.  As compared to the
proposed rule, the final rule increases the number of monitoring sites
from ~20 to ~75 and shifts the focus from urban monitoring to both urban
and rural monitoring locations.  Manually-operated PM10-2.5 speciation
samplers must operate on at least a 1-in-3 day schedule and be
collocated with PM2.5 speciation at NCore stations.  Since EPA is
requiring PM10-2.5 speciation monitoring primarily for scientific
purposes, it is appropriate to have monitoring at a variety of urban and
rural locations to increase the diversity of areas that will have
available chemical species data.  NCore will have about 75 sites mostly
in urban areas, with a subset of about 20 rural sites. For more
information on NCore, see:   HYPERLINK
"http://www.epa.gov/ttn/amtic/ncore/index.html" 
http://www.epa.gov/ttn/amtic/ncore/index.html .  

The primary objective for PM10-2.5 speciation data in the monitoring
rule is to support further research in understanding the chemical
composition and sources of PM10, PM10-2.5, and PM2.5.  In addition, more
specific uses for the data can be inferred and include:

Collection of PM10-2.5 composition data to inform health effect studies,
both in terms of the relationship between specific PM10-2.5 species and
health, and between PM emitted from different source types and health.

Advancement of PM10-2.5 speciation monitoring methods in anticipation of
wider use under a PM10-2.5 NAAQS if one is adopted later.

Use of PM10-2.5 speciation data to promote advancement of source
attribution methods.

Determination of spatial and temporal variations in PM10-2.5
concentrations in urban and rural environments.

PM10-2.5 SPECIATION MEASUREMENT ISSUES 

The Criteria Document (CD), prepared for the previous PM NAAQS review,
provides an overview of the current information on coarse particle
formation, sources, composition, and mass measurement issues (U.S. EPA
2004).  No clear recommendations were given for an approach to
collecting PM10-2.5 speciation data.  It was noted that “satisfactory
techniques are available to separate fine particles from coarse
particles and to collect the fine particles on a filter.  However, no
consensus exists as yet on the best technique for collecting a coarse
particle sample for determination of mass and composition. Candidates
include multistage impaction, virtual impaction, and difference
(subtracting PM2.5 mass or composition from PM10 mass or
composition).” 

Since the writing of the CD, the EPA Office of Research and Development
(ORD) has conducted a multi-site field evaluation of candidate
methodologies for PM10-2.5 mass (U.S. EPA 2006b).  Five PM10-2.5
measurement approaches were initially selected for study and included
virtual impaction (dichotomous sampling), difference, and continuous
methods.  In addition to continuous monitoring devices, integrated
filter-based monitors were used to collect filters for subsequent
speciation analysis.  The ORD results from speciation analyses of the
filters are pending and when available, can be used to inform this
planning process.  So far, ORD has found that when reconstructing
PM10-2.5 mass using the speciation results (sum of species), there is a
significant portion (10-50%) of the mass that is unaccounted for or
unidentified in some locations.  It is important to note that this
includes uncertainties associated with the factors used in
reconstructing mass (e.g., the factor used in conversion from OC to OM).
 Although the level of agreement between the reconstructed mass and the
measured mass was not always high, linear regression comparisons between
constructed mass and measured mass did show high correlation.  The
PM2.5, PM10-2.5 and PM10 mass comparisons of the dichotomous (dichot)
sampler with FRMs in Phoenix, AZ showed the dichot to be 10% higher for
PM2.5, 7% lower for PM10-2.5 and 4% lower for PM10. It was hypothesized
that the higher PM2.5 dichot mass was due to coarse particle intrusion
into the fine mode and significant measurement biases will occur only if
the coarse fraction of PM10 appreciably exceeds the PM2.5 size fraction,
as was seen in Phoenix.  Mass comparisons in other locations
(Birmingham, AL and Lindon, UT) showed very good agreement (± 4%)
between the dichots and FRMs for PM2.5, PM10-2.5 and PM10. High
correlations (R > 0.986) were found in all cases, indicating that the
response between the dichot and FRM is very consistent from one sampling
event to another.

 

Limited PM10-2.5 speciation monitoring studies have been conducted in
the U.S. and most of these studies were conducted using the difference
method.  The uncertainties and inconsistencies between the analytical
techniques used in these studies are unknown.  PM10-2.5 speciation has
been studied at both urban locations (Chow et al., 1996; Chow et al.,
1993; Sardar et al., 2005, Edgerton et al., 2005) and rural IMPROVE
monitoring locations (Eldred et al., 1997; Malm et al., 2007; Lee et
al., 2007).  Soil components (e.g., Si, Al, Ti, Ca, Fe, K), and organic
carbon (OC) were consistently found to be dominant components of
PM10-2.5 and the significance of nitrate and sulfate found was dependent
on the location studied.  For coarse mass in rural locations, Malm et
al. (2007) found the soil components (61%) and particulate organic
carbon mass (24%) to be the major components, with nitrate at 8%,
elemental carbon (EC) at 1%, sea salt at only 2%, and sulfate as
negligible.  Particulate organic carbon mass was defined as OM = OC*1.8.
 In southeastern urban locations, Edgerton et al., (2005) found similar
results to Malm et al. (2007), but also showed a significant portion
(26-38%) of the reconstructed PM10-2.5 mass to be unidentified when
accounting for OM as OC*1.4 and on the order of 16-23% when using OM as
OC*2.5.  In the San Joaquin Valley, Chow et al., (1996) also found total
carbon aerosol (TC); ions (e.g., nitrate, sulfate, sodium, chloride);
and soil components to be abundant in the PM10 fraction and in Southern
California, Sardar et al., (2005) found the soil components, OC and
nitrate to be dominant. 

A list of coarse particle constituents was provided in the Criteria
Document and includes suspended soil or dust; fly ash;
nitrates/chlorides/sulfates; soil components (Si, Al, Ti, Ca, Fe); sea
salt; tire/brake/road wear debris; and biological materials.  Not all of
these components can be measured directly through a filter-based
speciation monitoring program (e.g., fly ash, tire/brake/road wear
debris); however, some components or species may be represented by
components that can be measured with existing techniques used for PM2.5
speciation (e.g., sodium and chloride ion for sea salt).

The current PM10-2.5 FRM difference method, dichotomous sampler, and
current speciation filter-based samplers serve as logical choices for
the basis of a PM10-2.5 speciation sampler design.  A combination of
filter types and analytical methods are currently being used in both the
PM2.5 Chemical Speciation Network (CSN) and IMPROVE monitoring programs
to collect components of PM2.5.  These existing techniques can also be
applied to a PM10-2.5 speciation monitoring program, but not without
some complication.  

The FRMs for PM2.5 and PM10 (low-volume sampling at 16.7 Lpm) provide
relatively precise (within ± 10%) methods for determining the mass on a
Teflon filter.  However, uncertainties remain about the relationship
between the mass and composition of material remaining on the filter as
determined by the FRM and the mass and composition of material that
existed in the atmosphere.  Measurement errors of concern for PM10
sampling include uncertainty in cut point tolerances, particle bounce
and re-entrainment, impaction surface overloading, and losses to sampler
internal surfaces (U.S.EPA 2004).  Another measurement uncertainty for
PM2.5 sampling is the potential for inclusion of a small fraction of
coarse particles in the fine mass fraction under some circumstances.

Modification of the PM2.5 speciation sampler inlets to PM10 was
suggested by CASAC (EPA-SAB-CASAC-CON-04-005) as an option for PM10-2.5
speciation by difference.  This may be a viable alternative as long as
both speciation samplers have identical flow rates, filter sizes, and
filter handling procedures.  One limitation of the most widely used
PM2.5 speciation sampler (MetOne SASS) is the difference in flow rate
(6.7 Lpm) from the PM2.5 and PM10 FRMs (16.7 Lpm).  Differences in flow
rates result in differences in filter face velocity and pressure drop
across the filters, which may adversely affect the volatile species and
subsequent comparison of mass closure or reconstructed mass with the FRM
total mass; however, volatility issues are most likely less important
for PM10-2.5 particles than for PM2.5. 

PM10-2.5 FRM Difference Method

As is the case with all PM measurement methods, uncertainties exist with
the PM10-2.5 difference method. These include data loss if either the
PM2.5 or the PM10 sampler fails; uncertainties in flow rate and filter
weights (both before use and after collection and equilibration of
particles); and uncertainties due to the loss of semi-volatile
components which may occur for each size cut.  Allen et al. (1999) have
suggested ways to improve coarse particle difference measurements by
instituting careful control of sampling aspects (e.g., flow rate
control), management of gravimetric analysis issues, and proper
implementation of field blanks.  The viability of PM10-2.5 speciation by
a difference method requires further evaluation.  However, preliminary
regression comparisons for speciation by difference and the dichot
method have shown high levels of agreement and high correlation for
predominant species.  While there is currently no consensus on whether
the mixing of PM2.5 and PM10-2.5 aerosols causes a bias in either
measurement, CASAC mentioned the need for sampling separation and
collection of filters with only coarse particles to avoid mixing of
PM2.5 and coarse particles and the potential for subsequent chemical
interaction.  Allen et al. (1999) also mentions the importance of
maintaining filter flow rates greater than 10 Lpm, preferably 16.7 Lpm,
to avoid degraded precision.  As mentioned above, the most widely used
speciation sampler has a flow rate of 6.7 Lpm. per channel.  The impact
of this low flow rate should be evaluated if the low-flow speciation
samplers are used for PM10-2.5 by difference. 

Dichotomous Samplers (Dichots)

continues to evaluate and characterized virtual impactors.  As mentioned
previously, results from the multi-site evaluation of PM2.5, PM10-2.5
and PM10 mass comparisons of the dichotomous sampler with FRMs showed
very good agreement (except in Phoenix) and very good correlations (U.S.
EPA 2006b). 

PM10-2.5 Species or Components

Table 1 provides a list of candidate or potential PM10-2.5 species that
can be measured with the existing PM2.5 speciation methods.  The
specific species that need to be measured for PM10-2.5 must be
identified in order to design a monitoring program.  For example, ions
(e.g., nitrate and sulfate) have been identified as only minor
components of PM10-2.5 in some locations.  It is not clear whether the
resources to measure ions are needed to support research needs for
PM10-2.5 speciation.  Elemental analysis methods (e.g., X-Ray
Fluorescence) can provide sulfur, potassium, chloride, and sodium
elements; therefore, it needs to be determined if these elements are
sufficient surrogates for the information needed for PM10-2.5.  If ions
are not needed, then it would eliminate the need to collect an
additional filter (nylon) and resources for an additional lab analysis. 
In addition, the need for elements by XRF versus extractable or water
soluble elements by ICP/MS should be determined.

Table 1. List of Candidate PM10-2.5 Species

Species	Filter Type	Denuder	Analysis Method

PM10-2.5 Gravimetric Mass	Teflon	None	Filter weighing

Elements:

Crustal or soil (Si, Al, Ti, Ca, Fe)

Several other elements currently measured routinely for PM2.5,
including: K, Cl, P, Mg, Cr, etc.	Teflon	None	EDXRF (Energy Dispersive
X-Ray Fluorescence) or alternative extraction method



Soluble Ions

Nitrate, sulfate, sodium, potassium, chloride, ammonium	Nylon or Teflon
MgO	Ion Chromatography

Carbon

Organic and Elemental Carbon

Carbonate Carbon	Quartz	None	Thermal Optical Reflectance (TOR) and
transmittance (TOT) by IMPROVE_A

Separate acidification and analysis

Biological Material (Bioaerosols)

	Teflon or Quartz	None	Scanning Electron Microscopy (SEM) or 

Total Protein Assay as indicator

Fly ash	Teflon	None	Scanning Electron Microscopy (SEM)



Potential issues with XRF measurement of particles have been identified.
 XRF is typically done under vacuum to improve performance, enhance
detection limits, and reduce contamination of detector sources.  Use of
XRF under vacuum and the loss of volatile nitrate (as much as 30%) have
been demonstrated (U.S. EPA 2001).  In the PM2.5 speciation program, the
effects of vacuum are eliminated by analyzing the filter for mass prior
to analysis of elements by XRF.  A separate, denuded nylon filter is
used for nitrate and other ionic species.  Large or coarse particle size
effects may also be problematic for XRF.  Larger particles (greater than
3 micrometers) may absorb some of the incident and emitted x-rays for
light elements such as sodium, magnesium, aluminum, silicon, phosphorus,
sulfur, chlorine, and potassium (Chow 1995).  Absorption corrections
procedures for particle size effects on XRF results can be applied (Van
Dyck et al., 1985) and these factors will have to be optimized for
PM10-2.5 element analysis by XRF.  Another issue to consider is the
sensitivity of XRF and the sampling method or sampler chosen. 

Alternative techniques like Inductively Coupled Plasma/Mass Spectrometry
(ICP/MS).  ICP/MS have some advantages (e.g., improved detection limits
for many species but lower for some), but also some disadvantages which
include increased cost, labor intensive sample preparation, the need for
strong acid extraction, incomplete extraction efficiencies, and sample
filter destruction (XRF is a non-destructive analysis).  If the particle
size effects are addressed with XRF and the method sensitivity is
adequate for dichotomous sampling, then XRF may be a more appropriate
choice for elemental analysis. 

PM10-2.5 organic and elemental carbon (OC and EC) species can be
measured using the same thermal-optical analysis (TOA) method that is
used for PM2.5 speciation.  It is well known that both positive and
negative OC sampling artifacts exist (Eatough et al., 1990; Turpin et
al., 1994, Mader et al., 2001).  Some of the positive artifact can be
addressed by the use of backup quartz filter collection and subsequent
subtraction.  The artifact correction method to be applied to the urban
PM2.5 CSN is currently being evaluated and developed.  Once developed,
it will need to be evaluated for use in the PM10-2.5 program.  Organic
vapor denuders are not currently being used for either the PM2.5 CSN or
IMPROVE programs.  Although denuders may be appropriate, they are still
not ready for “prime time” and may introduce negative OC artifacts
due to the disruption of the gas-particle equilibrium during sampling. 
Han et al., (2007) mention an interference with metal oxides (e.g., iron
oxides) and TOA analysis; whereby certain metal oxides can serve as a
source of O2 in the helium atmosphere.  Since the soil component of
PM10-2.5 is expected to be significant, any effects of metal oxides on
the OC and EC results will need to be explored.  Carbonate carbon may
also be a significant constituent of PM10-2.5 and a separate punch from
the quartz filter will have to be analyzed to quantify it. 

Biological materials (bioaerosols) are collected with the filter-based
particle sampling techniques used for PM10-2.5 or PM2.5 monitoring and
included in the OC measurement, but are not quantified separately from
other components.  If bioaerosol species (e.g., pollens and molds) need
to be qualitatively or quantitatively identified for the PM10-2.5
speciation program, an appropriate measurement technique will need to be
identified (or developed) and evaluated.  Some biological materials can
be identified using the scanning electron microscopy (SEM) technique
(U.S. EPA 2002).  Existing bioaerosol monitoring programs (e.g.,
BioWatch) collect particles using filters and qualitatively test for
biological pathogens (e.g., anthrax). Total protein has been measured
from filters with an assay technique and used as an indicator of total
biological material (Menetrez et al, 2007). The specific bioaerosol
species of interest (or indicator/surrogate species) need to be
specified in order to explore appropriate collection and analysis
techniques. 

Fly ash is also included in the list of PM10-2.5 constituents of
interest in the CD.  Like bioaerosols, fly ash is collected with the
filter-based particle sampling techniques used for PM10-2.5 or PM2.5,
but not quantified separately from other components.  If fly ash is
needed for the PM10-2.5 speciation program, an appropriate measurement
technique will need to be identified or developed and evaluated.  Some
fly ashes can be identified using the SEM analytical technique (U.S. EPA
2002). 

NETWORK DESIGN ISSUES

The final monitoring rule contains a requirement for PM10-2.5 speciation
at NCore multi-pollutant monitoring sites by January 1, 2011.  As
mentioned previously, this was revised from the proposed rule.  The
NCore will have about 75 sites mostly in urban areas, with a subset of
about 20 rural sites.  Spatially, the candidate NCore locations (see
map) may not be the best choice for PM10-2.5 speciation given that the
highest PM10-2.5 mass concentrations are in the Southwest and Southern
CA (summer peak).  The NCore site selection is based on representative
monitoring to provide community-wide characterization of exposure and
sites leveraged with other measurement systems (e.g., PAMS, NATTS). 
Since PM10-2.5 mass is more likely to be influenced by local sources or
wind-blown dust in areas with little vegetation, NCore is not
necessarily the optimal design for PM10-2.5 speciation.  The spatial
adequacy and representativeness of the NCore sites for long-term
PM10-2.5 speciation monitoring will need to be evaluated.  

PM10-2.5 MONITORING PLAN AND METHOD RESEARCH NEEDS

There are still many unanswered questions regarding PM10-2.5 speciation
monitoring; however, a few PM10-2.5 by difference and dichotomous
monitoring sites (3 to 5 locations) should be used for pilot monitoring
to answer some of the questions outlined below and to fine tune the
final monitoring approach. The CASAC (EPA-SAB-CASAC-CON-04-005)
expressed concerns about speciation by difference due to the cumulative
effects of the imprecision in both PM2.5 and PM10 measurements.  White
(1998) showed the uncertainty in coarse mass by difference to be about 3
times the uncertainty in the fine measurement.  An evaluation of the
practicality and validity of PM10-2.5 speciation by difference method is
needed.  The CASAC comments also leaned toward virtual impaction and
felt it had significant advantages (e.g., collects the PM10-2.5 size
fraction directly).  For those PM10-2.5 species that are in common with
the PM2.5 speciation program, the existing National Laboratory Contract
for PM2.5 speciation can be used to analyze the filters.  In order to
sample for the current suite of species in the PM2.5 speciation program
(including ions) by difference, identical PM10 versions of the PM2.5
speciation sampling devices can be collocated with existing PM2.5
samplers.  PM10 sharp cut cyclones would need to be commercially
available at a flow rate of 6.7 Lpm for the MetOne SASS speciation
samplers and the vendor has been contacted about developing and
characterizing such cyclones.  A PM10 cyclone is already available for
the 22 Lpm speciation carbon sampling device.  The existing SASS sampler
could be retrofitted to have two PM2.5 channels with Teflon and nylon
filters and two PM10 channels with Teflon and nylon filters if needed. 
One additional PM10 speciation carbon sampler would need to be
collocated with the existing PM2.5 sampler for OC/EC aerosol species. 
If PM10-2.5 speciation was based on the dichotomous sampler, it would
require three samplers at each location to collect the Teflon, nylon,
and quartz filters for speciation if elements, ions, and carbon species
are needed. 

Some of the questions relevant to PM10-2.5 speciation monitoring are
outlined below:

What are the important PM10-2.5 species to measure?

Are ions important PM10-2.5 species? If so, what ions should be on the
target species list? 

What are the PM10-2.5 speciation sampling artifacts that may be
encountered and how should they be addressed in the monitoring program?
Is speciation by the difference method problematic for PM10-2.5
speciation and if so what specific issues make it problematic? 

The current and most widely used PM2.5 speciation sampler is the MetOne
SASS and it has a flow rate of 6.7 Lpm which is significantly lower than
either the FRM for PM10-2.5 mass or the dichotomous sampler (16.7 Lpm).
If this sampler was configured to do PM10-2.5 by difference, would the
6.7 Lpm flow rate be problematic, especially with our knowledge of
concerns about low-flow cut-points and particle intrusion, and the need
to compare to what is collected by the PM10-2.5 FRM for mass?

What analysis methods should be used? What PM2.5 speciation analysis
methods are appropriate to also use in the PM10-2.5 speciation program?

Is XRF the most appropriate method for PM10-2.5 speciation? Can the
complication of particle size and absorption effects in XRF be
adequately resolved using absorption correction factors? Does XRF
provide adequate sensitivity (detection limits)?  If XRF sensitivity is
not adequate, should some other more sensitive and potentially more
expensive and destructive analytical technique be considered? If an
elemental method that has more sensitivity and applies acid extraction
is needed, is the recovery of extractable metals adequate versus total
metals by XRF?  

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 biological particles and fly ash need to be characterized, what
specific types of biological materials and fly ashes should be measured?
What analysis methods should be used to identify and quantify these
species? Is scanning electron microscopy (SEM) on Teflon filters
adequate to quantify and identify the biological material present? Is
the use of other assay techniques necessary to adequately obtain a
quantitative indicator (e.g., total protein) of the total biological
material present?

When reconstructing PM10-2.5 mass using the sum of PM10-2.5 species
concentrations, a significant portion of unidentified mass has been
identified. Are there other PM10-2.5 target species or analysis methods
that can be used to help identify the source of this mass in order to
reduce the amount of unidentified mass and obtain better mass closure? 

What factors should be considered in the selection of pilot monitoring
site locations or areas? One key issue in the proposed PM rule (January
2006) was the need to distinguish urban from rural coarse particles.
What pilot site selection criteria would help in selecting urban and
rural sites for collecting data to address this issue? What analysis
methods or target species are particularly important to inform this
issue? 

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 PM10-2.5 speciation. Which of
these sampler types should be included or excluded from the pilot
network design?

Additional information may be provided to inform some of the questions
above when ORD publishes results on PM10-2.5 speciation from the
multi-city field evaluation.  In addition, any pilot monitoring program
that is developed and implemented for PM10-2.5 speciation may also
provide information to resolve some of these issues. 

QUALITY ASSURANCE (QA) 

The PM10-2.5 speciation monitoring program is expected to follow similar
requirements as specified for the PM2.5 speciation monitoring program,
which include collocation for precision estimates and sampler flow-rate
audits.  PM10-2.5 species-specific goals for bias and precision have not
been specified.  Additional QA procedures and possibly DQOs will need to
be developed for PM10-2.5 speciation. 

REFERENCES

Allen, G.A., Oh, J.A., Koutrakis, P., Sioutas, C. (1999). Techniques for
High-Quality Ambient Coarse Particle Mass Measurements, Journal of the
Air & Waste Management Association 49, PM-133-141. 

Chow, J.C., Watson, J.G., Lowenthal, D.H., Solomon, P.A., Magliano,
K.L., Ziman, S.D., Richards, L.W. (1993). PM10 and PM2.5 Compositions in
California’s San Joaquin Valley, Aerosol Science and Technology 18,
105-128.

Chow, J.C. (1995). Measurement Methods to Determine Compliance with
Ambient Air Quality Standards for Suspended Particles, Journal of the
Air & Waste Management Association 45, 320-382.

Chow, J.C., Watson, J.G., Lu, Z., Lowenthal, D.H., Frazier, C.A.,
Solomon, P.A., Thuillier, R.H., Magliano, K. (1996). Descriptive
analysis of PM2.5 and PM10 at Regionally Representative Locations during
SJCAQS/AUSPEX, Atmospheric Environment: 30 (12), 2079-2112.

Eatough, D.J., Aghdaie, N., Cottam, M., Gammon, T., Hansen, L.D., Lewis,
E.A., Farber, R.J. (1990). Loss of Semi-volatile Organic Compounds from
Particles during Sampling on Filters, In: Mathai, C.V. (Ed.),
Transaction of Visibility and Fine Particles, Air and Waste Management
Association, Pittsburgh, PA, pp. 146-156.

Edgerton, E.S., Hartsell, B.E., Saylor, R.D., Jansen, J.J., Hansen,
D.A., Hidy, G.M. (2005). The Southeastern Aerosol Research and
Characterization Study: Part II. Filter-based Measurements of Fine and
Coarse Particulate Matter Mass and Composition, Journal of the Air &
Waste Management Association 55, 1527-1542.

Eldred, R.A., Cahill, T.A., Focchini, R.G. (1997). Composition of PM2.5
and PM10 Aerosols in the IMPROVE Network, Journal of the Air & Waste
Management Association 47, 194-203.

EPA-SAB-CASAC-CON-04-005. Clean Air Scientific Advisory Committee
(CASAC) Ambient Air Monitoring and Methods (AAMM) Subcommittee
Consultation on Methods for Measuring Coarse-Fraction Particulate Matter
(PMc) in Ambient Air (July 2004).

Han, Y., Cao, J., An, Z., Chow, J.C., Watson, J.G., Jin, Z., Fung, K.,
Liu, S. (2007). Evaluation of the thermal/optical reflectance method for
quantification of elemental carbon in sediments, Chemosphere (69),
526-533.

Lee, T., Yu, X-Y., Ayres, B., Kreidenweis, S.M., Malm, W.C., Collett, J.
L., (2007). Observations of fine and coarse particle nitrate at several
rural locations in the United States. Atmospheric Environment (2007),
doi:10.1016/j.atmosenv.2007.05.016.

Loo, B.W. and Cork C.P. (1988). Development of high efficiency virtual
impactors, Aerosol Science & Technology (9), 167–176.

Mader, B.T. and Pankow, J.F. (2001). Gas/solid Partitioning of
Semivolatile Organic Compounds (SOCs) to Air Filters. 3. An Analysis of
Gas Adsorption Artifacts in Measurements of Atmospheric SOCs When Using
Teflon Membrane Filters and Quartz Fiber Filters, Environ. Sci. Technol.
35:3422-3432.

Malm, W.C., Pitchford, M.L., McDade, C., Ashbaugh, L.L. (2007). Coarse
Particle Speciation at Selected Locations in the Rural Continental
United States, Atmospheric Environment 41, 2225-2239.

Menetrez, M.Y., Foarde, K.K., Esch, R.K., Dean, T.R., Betancourt, D.A.,
Moore, S.A., Svendsen, E.R., and Yeatts, K. (2007). The Measurement of
Ambient Bioaerosol Exposure, Aerosol Science and Technology, 41:
884-893.

Querol, X., Minguillon, M.C., Perez, N., Alastuey, A., Viana, M.,
Morenao, T., Bernabe, R.M., Blanco, S., Cardenas, B., Vega, E., Sosa,
G., Escalona, S., Ruiz, H., and Artinano, B. (2007). PM Speciation and
Sources in Mexico During the MILAGRO-2006 Campaign, Atmospheric
Chemistry & Physics Discussion, 7, 10589-10629.

Sardar, B.S., Fine, P.M., Sioutas, C. (2005). Seasonal and spatial
variability of the size-resolved chemical composition of particulate
matter (PM10) in the Los Angeles Basin, Journal of  Geophysical
Research, Vol. 110, D07S08.

  SEQ CHAPTER \h \r 1 Turpin, B.J., Huntzicker, J.J., and Hering, S.V.
(1994). Investigation of Organic Aerosol Sampling Artifacts in the Los
Angeles Basin, Atmos. Environ. 28(19):3061-3071.

U.S. EPA 2001. Evaluation of PM2.5 Chemical Speciation Samplers for Use
in the EPA  National PM2.5 Chemical Speciation Network, Office of Air
Quality Planning and Standards, May 2001; EPA-454/R-01-005.

U.S. EPA 2002. Guidelines for the Application of SEM/EDX Analytical
Techniques to Particulate Matter Samples, Office of Research and
Development, September 2002; EPA-600/R-02/070.

U.S. EPA 2004. Air Quality Criteria for Particulate Matter, Volume I of
II; October 2004; EPA-600/P-99/002aF.

U.S. EPA 2006a. Part 53 and 58 – Revisions to the Ambient Air
Monitoring Regulations, Federal Register, Vol. 71, No. 200, October 17,
2006.

U.S. EPA 2006b. Multi-site Evaluations of Candidate Methodologies for
Determining Coarse Particulate Matter (PM10-2.5) Concentrations: August
2005 Updated Report Regarding Second-Generation and New PM10-2.5
Samplers; September 2006; EPA-600/R-06/093.

Van Dyck, P., Markowicz, A., and Van Grieken, R. (1985). Influence of
Sample Thickness, Excitation Energy and Geometry on Particle Size
Effects in XRF, X-Ray Spectrometry, Vol. 14 (4), 183-187.

White, W. (1998). Statistical Considerations in the Interpretation of
Size-Resolved Particulate Mass Data,  Journal of the Air & Waste
Management Association, 48:454-458.

 PAGE   

 PAGE   1 

Figure taken from EPA 2006b

Figure taken from EPA 2006b

