UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
° WASHINGTON D.C. 20460
August 26, 2008
EPA-CASAC-08-020
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
The Honorable Stephen L. Johnson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, D.C. 20460
Subject: CAS AC Ambient Air Monitoring & Methods (AAMM) Subcommittee Peer
Review of the Draft Federal Reference Method (FRM) for Lead in PMio
Dear Administrator Johnson:
EPA's Office of Air Quality Planning and Standards (OAQPS), within the Office of Air
and Radiation, requested that the Agency's Clean Air Scientific Advisory Committee (CASAC)
Ambient Air Monitoring & Methods (AAMM) Subcommittee (CASAC Subcommittee) conduct
a peer review of EPA's "Draft Federal Reference Method (FRM) Lead in PMi0 (Pb-PMi0)" —
that is, lead (Pb) in paniculate matter less than 10 micrometers in diameter (PMio) — dated June
15, 2008. The CASAC Subcommittee roster is enclosed as Enclosure A to this letter,
Subcommittee members' individual written comments are found in Enclosure B, and the
Agency's background and charge memorandum to the Subcommittee is provided in Enclosure C.
The Agency solicited CASAC's advice on this topic as part of EPA's current review of
the National Ambient Air Quality Standards (NAAQS) for Lead. Several options being
considered for the final Lead NAAQS would require the Agency to develop FRM and Federal
Equivalent Method (FEM) criteria for the collection of Pb-PMio in ambient air, in order for
monitoring data to be used in determining attainment with the NAAQS. The Agency has
proposed a FRM for Pb-PMio based on the existing, low-volume PMioc sampler (i.e., a PMio
sampler that meets special requirements that are part of a PMio-2.s reference method sampler, as
specified by Federal regulation), coupled with analysis by the x-ray fluorescence (XRF)
analytical method.
The Agency released its Proposed Rule for the Revision of the NAAQS for Lead (40
CFR Parts 50, 51, 53 & 58) on May 1, 2008, and this was subsequently published in the Federal
Register on May 20, 2008 (73 FR 29184-29291) as a Notice of Proposed Rulemaking (NPR).
On July 14, 2008, the CASAC Subcommittee conducted its review of the proposed FRM for the
measurement of Pb-PMio via a public advisory teleconference. The statutory (chartered)
-------
CASAC held a subsequent public teleconference meeting on August 18, 2008 to discuss and
approve the draft letter (dated August 11, 2008) containing our comments and recommendations.
EPA specifically requested the Subcommittee's comments both regarding the type of
sampler to be used and the choice of the multi-elemental analytical method for the Pb-PMi0
FRM — and, in particular, the low-volume PMi0c FRM sampler and the XRF analysis method,
respectively. The CASAC Subcommittee notes that the range of the level for the revised Lead
NAAQS under consideration in this NPR is quite broad, extending from 0.1 to 0.5 ug/m3. The
Subcommittee was therefore challenged in this peer review by not having a narrower "target"
range for the final NAAQS for Lead, since the level and averaging time of the revised Lead
standard significantly impact the suitability of candidate sampling and analytical methods.
Without more guidance on EPA's data quality objectives (DQOs) for Lead monitoring, the
members of the CASAC Subcommittee are unable to provide definitive responses to Questions 2
and 4 that Agency staff posed to the Subcommittee as part of its review.
Nevertheless, overall — and subject to addressing the CASAC's previously-expressed
concerns with transitioning to a Pb-PMio sampling indicator (reiterated below) — the CASAC
Subcommittee unanimously supports the use of the PM10c FRM sampler. In addition, it is the
consensus recommendation of the Subcommittee in this peer review that EPA consider selecting
inductively coupled plasma-mass spectroscopy (ICP-MS) as the Pb-PMwFRM analytical
method and using XRF as an FEM.
The five charge questions from the Agency, along with a synthesis of the CASAC
Subcommittee's responses, are found immediately below:
1. What are your comments on the use of the low-volume PMwc FRM sampler as the
Pb-PM10 FRM sampler?
If the EPA chooses to transition from a lead in total suspended paniculate (Pb-TSP)
sampling indicator to a Pb-PMio indicator, the CASAC Subcommittee is generally supportive of
using the PMioc FRM sampler. The rationale for such a selection is well laid-out in the draft
FRM document that the Agency presented to the CASAC Subcommittee for its peer review, as
well as in the individual written comments from Subcommittee members. However, as
discussed below, the CASAC has previously noted that the choice of Pb-PMio as a sampling
indicator should be conditional on a considerable tightening of the final Lead standard.
2. What are your comments on the use of XRF as the Pb-PMi0 FRM analysis method?
By way of background, on March 25, 2008, the CASAC Subcommittee held a public
advisory teleconference meeting to conduct a consultation with OAQPS on several ambient air
monitoring issues related to the Lead NAAQS, including issues associated with alternative lead
indicators. As is the CASAC's customary practice, there was no consensus report from the
CASAC as a result of that consultative meeting. However, Subcommittee members' individual
written comments were attached in Appendix B of the CASAC's letter to the Agency (EPA-
CASAC-08-010, dated April 14, 2008). A majority of the Subcommittee members who
submitted individual written comments pursuant to this March 25 consultation generally
indicated that XRF was an appropriate Pb-PMio FRM analysis method, although several
-------
members commented that the Agency should consider using ICP-MS (or, in one case, atomic-
absorption (AA) spectroscopy) as an alternate analytical method for the FRM.
In addition, as mentioned above in this letter, this question is particularly difficult to
answer without a clearer sense of the level of the revised NAAQS for Lead. Moreover, at
present, the CASAC Subcommittee is unsure as to what the analytical requirements are for this
method, as well as the Agency's associated data quality objectives. The Subcommittee
understands that an analysis of the DQOs is underway — albeit in the face of uncertainty
concerning both the level and the averaging time of the revised Lead NAAQS.
That having been said, the CASAC Subcommittee considers XRF as possessing a number
of potential benefits over competing approaches, although it also has some weaknesses.
Specifically, XRF is, overall, viewed positively by the Subcommittee, in that it: is reasonably
cost-effective; is currently being used for analysis of the Speciation Trends Network (STN)
filters; provides concentrations of elements other than lead; avoids the extraction procedures
required by methods such as ICP-MS and AA spectroscopy; and is non-destructive. On the other
hand, in comparison with XRF, ICP-MS offers lower detection limits, more direct calibration
against NIST-traceable references, transparent interpretation of results, and compatibility with
the fiber filters used for high-volume TSP or PMio measurements. Importantly, the uniformity of
sample deposits across the face of low-volume PMi0c filters would need to be more carefully
investigated prior to selection of an XRF FRM, because XRF analyzes only a portion of the
filter. The use of in-line filter holders appears to exacerbate this problem, and it is noted that
some XRF methods (e.g., the PANalytical instrument) slowly rotate the analysis holder during
analysis, which allows the oval-shaped x-ray beam to scan over a much larger area, thus
minimizing bias if there is any inhomogeneity of the filter deposit. Nonetheless, with whole-
filter extraction methods such as ICP-MS, the uniformity of the deposit is no longer an issue,
although completeness of recovery of lead from the filter for ICP-MS analysis must still be
confirmed.
On balance, therefore, the issues of deposit uniformity and calibration standards
associated with XRF raise analytical concerns not found with ICP-MS with respect to EPA's
accuracy goal of "an upper 95 percent confidence limit for the absolute bias of 10 percent."
Accordingly, the CASAC Subcommittee recommends that the Agency consider selecting ICP-MS
as the FRM and using XRF as an FEM.
3. What are your comments on the specific analysis details of the XRF analysis method
contained in the proposedPb-PMw FRM analysis method description?
Whether XRF is used as the FRM or as an FEM, there are a number of issues that need to
be addressed more thoroughly than currently appears in the Agency's draft FRM for Pb-PMio, in
that the appropriate section of EPA's review document should be oriented more towards
providing specific details for the analysis of airborne lead. Whatever analytical method is used
laboratory and field blanks should be used to detect possible contamination in the filters and the
overall system. It is recognized that different lots of filters may have dissimilar blank levels, so
filters should be matched with blanks from the same lots. A relatively large number of
laboratory blanks should be analyzed in each lot. Furthermore, field blanks, comprising
approximately ten percent (10%) of all sample filters, should be deployed and analyzed as well.
-------
4. Do you think the precision, bias andMDL of the XRF method for the proposed Pb
range will be adequate?
Again, answering this question is sensitive to the choice of the form and level of the
revised NAAQS for Lead, as implied by the question itself, and is made more difficult by not
having DQOs available for the CASAC Subcommittee's review. The Subcommittee understands
that there has been limited time to develop the DQOs, and that the Agency is also disadvantaged
by having such a broad range for the level of the final Lead NAAQS under consideration, as well
as the possible change in the averaging time of the standard. As discussed above, XRF is a
viable method, and should be able to meet the bias and method detection limit (MDL)
requirements. However, the lack of uniformity in the deposition of lead on filters could pose
issues with meeting the precision requirements. Accordingly, EPA should confirm that for the
sampler being used, non-uniformity in the deposition of lead particles does not compromise
meeting the specified DQOs. Given its greater sensitivity, and concerns over the non-uniformity
of deposition impacting XRF results, a number of individuals on the Subcommittee specifically
recommend that ICP-MS be selected as the analysis method for the FRM.
5. Are there any method interferences that we have not considered?
In the judgment of the CASAC Subcommittee the Agency has adequately identified the
potential interferences with XRF.
Finally, the CASAC, in its letter dated January 22, 2008 (EPA-CASAC-08-007), had
recommended transitioning the sampling indicator for lead from TSP to a low-volume ambient
air monitor for Pb-PMi0. This transition to a new indicator was also supported by a majority of
the members of the CASAC AAMM Subcommittee during its March 25, 2008 consultative
teleconference (see EPA-CASAC-08-010). Notwithstanding, as discussed in the most recent
letter from the CASAC (EPA-CASAC-08-016, dated July 18, 2008), the discussion leading up to
this recommendation assumed a significant tightening of the lead NAAQS. In particular, as the
CASAC noted in its July 18 letter, a Lead NAAQS set at a level as high as 0.5 ug/m3, using a Pb-
PMio sampling indicator, could potentially allow TSP Pb levels as high as 1 ug/m3 at sites near
large sources with coarse-mode particulate lead emissions. Therefore, the CASAC clarified its
recommendation by stating that, if the level of the revised lead NAAQS approaches this upper
end of the range, the current TSP indicator should not be changed — adding that, while
transitioning from a Pb-TSP to a Pb-PMio sampling indicator would indeed be "preferable," this
change should only be effected if the level of the final NAAQS for Lead is established
"conservatively below an upper bound of 0.2 ug/m3 or lower."
In closing, the CASAC Subcommittee welcomes the opportunity to review EPA's
proposed FRM for Pb-PMio, and reiterates that the choice of an appropriate FRM is crucial with
respect to the timely development of a more health-protective Lead NAAQS. The Subcommittee
stands ready to provide additional advice and recommendations with respect to any air-quality
-------
monitoring issues, especially those related to the NAAQS. As always, we wish the Agency well
in these important efforts to protect both human health and the environment.
Sincerely,
Dr. Armi stead (Ted) Russell, Chair
CASAC AAMM Subcommittee
Dr. Rogene F. Henderson, Chair
Clean Air Scientific Advisory Committee
Enclosures
-------
NOTICE
This report has been written as part of the activities of the EPA's Clean Air
Scientific Advisory Committee (CASAC), a Federal advisory committee independently
chartered to provide extramural scientific information and advice to the Administrator
and other officials of the EPA. The CASAC provides balanced, expert assessment of
scientific matters related to issues and problems facing the Agency. This report has not
been reviewed for approval by the Agency and, hence, the contents of this report do not
necessarily represent the views and policies of the EPA, nor of other agencies within the
Executive Branch of the Federal government. In addition, any mention of trade names or
commercial products does not constitute a recommendation for use. CASAC reports are
posted on the EPA Web site at: http://www.epa.gov/casac.
-------
Enclosure A - Roster of the CASAC Ambient Air Monitoring & Methods
(AAMM) Subcommittee
U.S. Environmental Protection Agency
Clean Air Scientific Advisory Committee (CASAC)
CASAC Ambient Air Monitoring & Methods (AAMM) Subcommittee
CASAC MEMBERS
Dr. Armistead (Ted) Russell (Chair), Georgia Power Distinguished Professor of Environmental
Engineering, Environmental Engineering Group, School of Civil and Environmental Engineering,
Georgia Institute of Technology, Atlanta, GA
Dr. Ellis Cowling, University Distinguished Professor At-Large, Emeritus, Colleges of Natural
Resources and Agriculture and Life Sciences, North Carolina State University, Raleigh, NC
Dr. Donna Kenski, Director of Data Analysis, Lake Michigan Air Directors Consortium (LADCO),
Rosemont, IL
SUBCOMMITTEE MEMBERS
Mr. George 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, Emeritus, Chemistry and Biochemistry Department,
Brigham Young University, Provo, UT
Mr. 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
A-l
-------
Dr. Kazuhiko Ito, Assistant Professor, 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 McMurry, Professor, Department of Mechanical Engineering, Institute of Technology,
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 Prather,* Professor, Department of Chemistry and Biochemistry, University of
California, San Diego, La Jolla, CA
Dr. Jay Turner, Visiting Professor, Crocker Nuclear Laboratory, University of California - Davis, 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, Providence Engineering and Environmental Group LLC, Baton Rouge, LA
Dr. Barbara Zielinska, Research Professor, Division of Atmospheric Science, Desert Research Institute,
Reno, NV
SCIENCE ADVISORY BOARD STAFF
Mr. Fred Butterfield, Designated Federal Officer, 1200 Pennsylvania Avenue, N.W., Washington, DC,
20460, Phone: 202-343-9994, Fax: 202-233-0643 (butterfield.fred@,epa.gov) (Physical/Courier/FedEx
Address: Fred A. Butterfield, III, EPA Science Advisory Board Staff Office (Mail Code 1400F), Woodies
Building, 1025 F Street, N.W., Room 3604, Washington, DC 20004, Telephone: 202-343-9994)
*Dr. Lumley and Dr. Prather did not participate in this CASAC AAMM Subcommittee activity.
A-2
-------
Enclosure B - Comments from Individual CASAC
AAMM Subcommittee Members
This appendix contains the written comments of individual members of the Clean
Air Scientific Advisory Committee (CASAC) Ambient Air Monitoring & Methods
(AAMM) Subcommittee. The comments are included here to provide both a full
perspective and a range of individual views expressed by Subcommittee members during
the review process. These comments do not represent the views of the CASAC AAMM
Subcommittee, the CASAC, the EPA Science Advisory Board, or the EPA itself.
Subcommittee members providing written comments are listed on the next page, and
their individual comments follow.
B-l
-------
Panelist Page#
Mr. George Allen B-3
Dr. Judith Chow B-5
Mr. Bart Croes B-21
Dr. Kenneth Demerjian B-23
Dr. Delbert Eatough B-25
Mr. Dirk Felton B-26
Dr. Philip Hopke B-29
Dr. Kazuhiko Ito B-30
Dr. Donna Kenski B-32
Dr. Peter McMurry B-33
Mr. Richard Poirot B-35
Dr. Jay Turner B-38
Dr. Warren H. White B-40
Dr. Yousheng Zeng B-44
Dr. Barbara Zielinska B-45
B-2
-------
Mr. George Allen
The following are written comments based on the Charge Questions in the EPA OAQPS memo
to the SAB dated June 15, 2008. These comments also reflect discussion during the July 14
teleconference AAMM meeting on a peer review of the Draft Federal Reference Method (FRM)
Lead in PM10 (Pb-PMlO). A copy of these comments is also being sent to Dr. Ted Russell,
CASAC AAMM Subcommittee Chair.
Peer Review Charge Questions in Bold:
1. What are your comments on the use of the low-volume PMlOc FRM sampler as the Pb-
PM10 FRM sampler?
The existing PMlOc sampler is an obvious choice for a sampler since it is well characterized and
commercially available from several vendors. Sequential (automated) PM10 samplers should
also be allowed, either as FRM or FEM samplers. The dichotomous sampler is another obvious
candidate for an FRM or FEM sampler for PM-10 lead.
2. What are your comments on the use of XRF as the Pb-PMlO FRM analysis method?
XRF is sufficient for routine analysis, but for the FRM, a more sensitive and specific technique
should be used, such as ICPMS. If XRF is used, the method should be an FEM. If XRF remains
the FRM analysis method, there are concerns of uniform deposit on the filter that may differ with
different sizes (coarse vs. fine mode) of particles. Appropriate filter deposition testing would
have to be done prior to promulgation of XRF as the FRM analysis method. There are also
concerns regarding different XRF analytical methods and calibration techniques across different
laboratories, the lack of a NIST thin-film XRF Pb reference standard, possible issues with heavy
filter loading, the difficulty of generating spiked samples, and the possibility of interferences.
ICPMS does not have any of these concerns.
3. What are your comments on the specific analysis details of the XRF analysis method
contained in the proposed Pb-PMlO FRM analysis method description?
The XRF analysis method description proposed here is well written and takes into account most
of the issues raised above. It does not resolve the issues of non-uniform deposition or the lack of
a NIST thin-film standard for Pb. A PM10 filter can appear visually to have a uniform deposit,
but in urban areas the visual appearance is often driven by fine-mode aerosol which may not
reflect the deposition pattern of coarse mode Pb. Thus, visual inspection is only a crude first test
for uniform deposition of PM-10 Pb. The issue of filter blanks needs more attention; blank
values can vary by manufacturing lot. Thus, the blanks used for a set of samples must be from
the same lot. The method description does not use field blanks; it is important to have 5% of
filters used as field blanks. The method needs to include a section on how levels below the
method's LOD or LOQ will be handled. I suggest reporting the blank-corrected data as
measured (even if it is slightly negative), but flagging it as below the LOD.
-------
4. Do you think the precision, bias and MDL of the XRF method for the proposed Pb
range will be adequate?
The XRF MDL for Pb will be a function of XRF method and blank levels and variability.
Although the MDL noted in this method description (1 ng/m3 one-sigma) is adequate, it may or
may not be achieved in the real world, since the MDL is a function of many things, including the
number and stability of lab and field blank filters and the length of XRF analysis time. The bias
and method detection limits in this draft are appropriate. I would suggest that the FEM precision
be tightened from 15% to 10%.
5. Are there any method interferences that we have not considered?
Not that I am aware of.
B-4
-------
Dr. Judith Chow
This memo addresses the twelve questions on which the Subcommittee members were asked to
comment regarding Attachment 1, "Draft Federal Reference Method (FRM) for Lead in PMio
(Pb-PMio), and Attachment 2, "Approaches for the Development of a Low-Volume Ambient Air
Monitor for Lead in Total Suspended Particulate (Pb-TSP) Sampler." This supplements prior
comments to the first set of questions that was appended to the April 14, 2008 letter from Dr.
Russell to Administrator Johnson.
Questions for Attachment 1 [Draft Federal Reference Method (FRM) for Lead in
PMIO (Pb-PM10)]
Question 1: What are your comments on the use of the low-volume PMi0c FRM sampler as
the Pb-PM10 FRM sampler?
My prior comments in the April 14 letter recommended that EPA move toward Pb-PMio. These
comments pointed out the lack of specificity and variability of inlet characteristics for the high-
volume TSP sampler (Code of Federal Regulations, 2007a). High-volume TSP is a poor
surrogate for inhalable particles and a poor surrogate for deposited particles. A true "Total
Suspended Particulate" sampler that collects all of particles that remain in the air is of such large
dimensions that it requires a small trailer and a large power supply to operate (Burton and
Lundgren, 1987; Lundgren et al., 1984). The argument given in favor of retaining TSP in the
April 14 letter was that large particles could contaminate surface areas and soils that might be
ingested or resuspended. If toxic soils and house dust are of concern in addition to inhalable
PMio, then these should be sampled and analyzed directly (Egami et al., 1989; Adgate et al.,
1998; Farfel et al., 2001; Bai et al., 2003).
FRM sampler inlets have been wind-tunnel tested and have well-defined cut-points and slopes
(10.2 ± 1.41 |im for SA-246B inlet; Watson and Chow, 1993; 2001). Sampling systems coupled
with these inlets provide accurate flow control, use low trace metal background PTFE Teflon-
membrane filters, and yield precise mass measurements when coupled with appropriate
laboratory weighing procedures. Low-volume PMi0cFRMs (Appendix O to Part 50) are similar
to PM2.5 FRMs, which use the same PMio impactor inlet with the addition of WINS or very
sharp-cut cyclone inlets (Kenny et al., 2000; 2004; Peters et al., 2001a; 2001b; 2001c). The low-
volume PMiocFRM sampler is consistent with EPA's proposed difference method for PMio-2.5
(U.S.EPA, 2006) that uses identical filter media, sample collection, gravimetric analysis, and
quality assurance [QA]/quality control [QC]) procedures for the side-by-side samplers.
Low-volume PMio and PM2.5 samplers are commercially available, are widely deployed in many
urban networks, and network operators are familiar with them. Costs for additional sampling
and analysis should be reasonable. There is no need for a separate Pb-PMio network, although the
existing low-volume PMio network might be expanded to suspected Pb hot-spots, as
recommended by several committee members in the April 14 letter. The Pb-PMio network
should be considered within the context of EPA's integrated air monitoring strategy (Scheffe et
B-5
-------
al., 2007; U.S.EPA, 2005) that intends to re-design the national monitoring system to attain
multiple objectives beyond compliance (Chow and Watson, 2008).
Question 2: What are your comments on the use of XRF as the Pb-PMi0 FRM analysis
method?
Energy dispersive x-ray fluorescence spectroscopy (XRF; NIOSH, 1998; U.S.EPA, 1999a;
Watson et al., 1999a; RTI, 2004; DRI, 2007) is the most commonly used analytical method for
multi-elemental analysis on Teflon-membrane filter samples, and the protocols always include
Pb. XRF does not destroy the sample, so it can be archived and re-examined for stable particles
by other methods (volatile aerosol components such as ammonium nitrate evaporate in XRF's
evacuated sample chamber). XRF is currently used for PM2.5 elemental analysis at urban
locations in the Chemical Speciation Network (CSN), at non-urban locations in the Interagency
Monitoring of PROtected Visual Environments (IMPROVE) network, and in many special
studies.
Other methods have been proven to be equally sensitive, accurate, and precise for Pb
measurements, including Proton Induced X-ray Emission Spectroscopy (U.S.EPA, 1999b),
Atomic Absorption Spectroscopy (AAS) Code (Fernandez, 1989; NIOSH, 1994a; 1994b;
U.S.EPA, 1999c; Code of Federal Regulations, 2007b), Inductively-Coupled Plasma Atomic
Emission Spectroscopy (ICP-AES; U.S.EPA, 1999d; NIOSH, 2003a; 2003b; 2003c),
Inductively-Coupled Plasma Mass Spectrometry (ICP-MS), and Anodic Stripping Voltametry
(ASV; NIOSH, 2003d). These methods are commonly applied to air filters for Pb, especially in
workplace environments and Hazardous Air Pollutant (HAPs) networks.
With adequate standard operating procedures (SOPs; such as those cited above), these methods
give comparable results for a wide range of sample types and environments (Keppler et al., 1970;
Gilfrich et al., 1973; Camp et al., 1974; 1978; Ahlberg and Adams, 1978; Nottrodt et al., 1978;
Witz et al., 1982; Lin et al., 1993; Walder and Furuta, 1993; Pyle et al., 1996; Bettinelli et al.,
1997; Reynolds et al., 1997; Watson et al., 1997; 1999b; 2000; Lemieux et al., 1998; Ashley et
al., 1999; Rich et al., 1999; VanCott et al., 1999; Gigante and Gonsior, 2000; Sterling et al.,
2000; Farfel et al., 2001; Harper et al., 2002; 2004; 2005; 2006; 2007; Menzel et al., 2002;
Sussell and Ashley, 2002; Bai et al., 2003; Drake et al., 2003; Moreira et al., 2005; Ariola et al.,
2006; Harper and Pacolay, 2006; Harris et al., 2006; Herner et al., 2006; Kilbride et al., 2006;
Kim et al., 2007).
Figure 1 shows an example comparing Pb concentrations measured by AAS on high-volume
PMio quartz-fiber filter analyzed by the Illinois Department of Environmental Quality with Pb by
XRF on the summed fine and coarse Teflon-membrane filters from a collocated dichotomous
sampler analyzed by DRI. The results are comparable, with a few outliers. These monitors from
South Chicago were in a highly industrialized area with relatively high levels of arsenic (As),
selenium (Se) and other potentially toxic elements. Refined Pb is amenable to a common acid
extraction method, such as nitric acid and aqua regia, which is not the case for most minerals
(and possibly not for Pb in its native ore prior to refining). The comparisons for other toxic
elements in Watson et al. (2000) are not as good as those for Pb.
B-6
-------
Although a method may be shown to yield quantities comparable with reference materials and
analyses by other methods, it may be inadequate if the equipment and procedures are not up to
the task. Each SOP should state its assumptions and include tests to indicate when deviations
from those assumptions are excessive. The procedure should attempt to minimize the effects of
interferences or sample deviations from the ideal. The ability of an XRF procedure to attain a 1
ng/m3 Pb detection limit depends on the filter mass (which affects the background count), sample
volume, sample duration, and deposit area. It also depends on the Pb excitation radiation energy,
intensity, beam area, and analysis time. The sensitivity and resolution of the SiLi detector is an
important consideration, as well as peak overlap that will raise the background (which decreases
the analysis precision). There are different, but analogous, considerations for the other methods
cited above.
HI lip
la ns
Slope- irEE±n.iis
her-IP I- DUE ± DUE m/m'
ipp- D.1.3:!1 ± nrai
DID DDE
nCHQT FH.. Load 0
CICHOT FH „ Lnad ij
i
5 DH4
pl- Dm+iDHJI pnitii1
s
£ DUt
- 1.13T±D£HJ
pl - ninz ±Qtm p
t O.I h in^lrtrccpt -
OD6
aOHOT FH,. Had OJn.(n':i
Offi
aOHOT FMnLi
Figure 1. Comparison of PM10 lead concentrations from an Andersen high-volume PMi0 on QMA
quartz-fiber filters analyzed by AAS and a Sierra 241 dichotomous PM10/PM2.5 sampler with Teflon-
membrane filters analyzed by XRF (Watson et al., 1999a) at the four sites during the third year of the
Robbins Paniculate Study in South Chicago between 10/01/97 and 09/26/98 (Watson et al., 2000).
As long as the minimum detectable limits (MDLs; 1.5 ng/cm2), precision (±15% at 90%
confidence level), and accuracy (±5%) are within the EPA's specified levels, any of the methods
cited above should be adequate. That said, XRF and/or proton induced x-ray emission (PIXE)
can simultaneously acquire 40-50 elements without much additional cost (except for the cost of
acquiring additional standards, performing instrument calibration, and data processing). If only
Pb is desired, most of the multi-element excitation conditions can be dropped, thereby increasing
B-7
-------
throughput and further lowering costs. The issues of extraction efficiency, use of different acid
mixtures for extraction, matrix interferences, potential contamination, and sample destruction
inherent in AAS, ICP-AES, TCP-MS, and ASV result in these methods being more costly, but
they may be of use in some instances. For example, ICP-MS can quantify Pb isotopic
abundances that might be of use in quantifying source contributions (Munksgaard and Parry,
1998).
Question 3. What are your comments on the specific analysis details of the XRF analysis
method contained in the proposed Pb-PMi0 FRM analysis method description?
While the method description in Appendix Q to Part 50, "Reference Method for the
Determination of Lead in Particulate Matter as PMio Collected From Ambient Air" covers many
details, there are several points that need clarification:
Section 1.1 (Line 2). PMio should be collected on an "acceptance tested" 46.2 mm diameter
polytetrafluoroethylene (PTFE) filter. Acceptance testing is performed to verify blank levels
for Teflon-membrane filters. In the early 1970s, one batch of Teflon-membrane filters was
contaminated with Pb from the manufacturer, and this compromised the study results (Chow,
1995a).
Section 1.1 (Lines 7 and 8). The definition of PMio should include a specific inlet efficiency
curve with a 50% cut-point and slope, similar to the PMio FRM specification (U.S.EPA,
1987).
Section 1.4 (Line 1). Is it necessary to specify "electrically powered"? I don't see any problem
with other vacuum assisted suction methods as long as the flow rate specifications are
attained. Photovoltaic cells and batteries are also sources of electricity.
Section 1.4 (Line 8). Change "Line intensity" to "photon energy".
Section 2.1 (Line 3). The deposit area on ringed Teflon-membrane filters varies slightly from
different speciation samplers (e.g., 11.76 - 11.78 cm2), and it is smaller than the 11.86 cm2
estimated for the Pb-PMi0 FRM sampler. It would be better if the deposit area is measured
from several samplers and sample batches to assure that the correct value is being used.
Section 2.2 (Lines 4-5). "The one-sigma detection limit for Pb is calculated as the average
overall uncertainty or propagated error for Pb, determined from measurements on a series of
blank filters." This should be more explicit, i.e., translate the square root of the number of
counts from a series of blank filters near the Pb analysis energies into |ig/m3 using the XRF
calibration factor (jig/count), sample volume, and deposit area. The one-sigma detection
limit is best based on each batch of unexposed blank filters to account for batch-to-batch
variations. Even though these variations are expected to be small, it is a better practice to
ensure consistency among different batches of filters. This might be incorporated into the
acceptance testing criteria.
-------
Section 3.1 (Lines 1 and 3). Define "too much deposit" (Line 1) and "heavy deposit" (Line 3).
This shouldn't be a problem with XRF, because Pb has a strong energy and is not much
affected by particle size or sample deposit (Criss and Birks, 1968; Hunter and Rhodes, 1972;
Rhodes and Hunter, 1972; Dzubay and Nelson, 1975; Adams and Billiet, 1976). One could
require a calculation of self-adsorption and the loading at which it might exceed the
measurement tolerances using one or more of the cited methods.
Section 3.1 (Line 5). While an optimum PMio filter loading of 150 jig/cm2 or 1.6 mg/filter is
reasonable for a 46.2 mm filter with a low-volume (16.7 L/min) sampler, this value needs to
be justified with a citation. The same is true for the minimum deposit of 15 jig/cm2 (Line 7).
An optimal loading estimate might be required to be part of the procedures, again using
published formulae.
Section 3.1 (Lines 8-10). Deposit non-uniformity may occur if an in-line filter holder is used, but
in-line filter holders are not part of PMio low-volume samplers. The deposits are very
uniform with these samplers, as evidenced by their appearance. Modern XRF equipment
also rotates the sample, and the incident beam is at an off-center angle, thereby lessening the
effects of a non-uniform deposit. Deposit uniformity might be defined by a performance
specification of some kind and be addressed in the SOP.
Section 3.2 (Line 11). "Energy resolution" should be defined as < 155-160 eV full-width at half
maximum..
Section 4.1 (Line 4). A CV of 15% is high. Typically, precision can be much better than ±10%.
Section 6.1.2. (Lines 4-5). Selecting 50 out of 500 filters, or 10% of blank filters, for acceptance
testing is more than is needed. Two filters out of a hundred are more reasonable and cost-
effective.
Section 6.1.2 (Line 8). Where did 4.8 ng Pb/cm2 come from? Based on the past records, 1-3 ng
Pb/cm2 seems to be a more adequate acceptance level, but this needs to be validated with a
citation.
Section 6.2.1 (Lines 2 and 3). The method should not imply that Thermo and PANalytical, are the
only units. UC Davis designed and operates its own system for IMPROVE samples, and it
seems to work fine. I believe EPA is still using the old LBL workhorse in its RTF labs.
Xenemetrix (new owners of Jordan Valley, www.xenemetrix.com/index.htm) and Spectro
(www.spectro.com/pages/e/index.htm) also have XRF units adaptable to this purpose.
Section 6.2.2 (Lines 1-4). Both 15 and 50 jig/cm2 Pb thin film standards can be obtained from
Micromatter Inc. (Arlington, WA). NIST (2008) also has a Pb standard solution, standard
reference material (SRM) 3128 with certified Pb value of 9.987 ± 0.018 mg/g, or other
SRMs in different matrices that might be applicable to assessing accuracy and precision.
Section 6.2.4 (Line 17). "Calibration is performed only when significant repairs occur or when a
change in fluorescers, X-ray tubes, or detector is made." Most XRFs are robust and may not
B-9
-------
need repairs for years. QA standards with each run monitor intensities and peak separations.
Nevertheless, it's a good idea to perform base calibrations at least once per year, and to use
the Lp line as a secondary peak to verify Pb by the La line.
Section 6.2.4.2 (Lines 3-4). Rather than keeping 20-30 filters as clean blank filters, it is better to
retain 2% of every new batch of filters (i.e., 100 per batch) for acceptance testing (see above
comment on Section 6.1.2).
Question 4. Do you think the precision, bias and MDL of the XRF method for the proposed
Pb range will be adequate?
Yes, with the appropriate samples and procedures. Arsenic (As) and other spectral interferences
can be estimated and corrected, and this is commonly done using the Pb Lp as well as the Pb La
to quantify Pb levels. A quick calculation shows that if As levels were so high as to overwhelm
the Pb lines, then Pb exposure would not be the biggest problem. The deposit inhomogeneity
reported by Bandhu et al. (2000) was caused by their use of in-line filter holders. Chow (1995b)
shows pictures of samples from in-line filter holders, demonstrating that you don't need a lot of
analysis to know when the deposit is non-uniform. The aerosol sampler (Fitz et al., 1989) used
in the Southern California Air Quality Study (SCAQS) used in-line filter holders and required
some extra effort to adjust the elemental data (Cahill et al., 1989; Chow et al., 1994; Matsumura
and Cahill, 1991) for subsequent interpretation. None of the samplers under consideration use
in-line filter holders, and all of them have a long-enough transition zone to assure a uniform
deposit. The good comparability reported in most of the studies cited above could not be
achieved if this were not the case.
Question 5. Are there any method interferences that we have not considered?
XRF spectrum processing methods are well-established for thin samples, and most of the newer
analyzers have software that can implement several of the most common approaches to
background subtraction, peak overlap correction, self-absorption (not really needed for Pb),
coincidence counting, and deadtime corrections. The software implements well-established and
non-proprietary methods (Bonner et al., 1973; Dzubay et al., 1977; Giauque et al., 1977;
Grennfelt et al., 1971; Lubecki, 1969; Parkes et al., 1974; Russ, 1977; Statham, 1976; Statham,
1977) that can be applied to any digitized spectrum.
B-10
-------
References
Adams, F.C.; and Billiet, J. (1976). Experimental verification of the X-ray absorption correction
in aerosol loaded filters. X-Ray Spectrometry, 5(4): 188.
Adgate, J.L.; Willis, R.D.; Buckley, T.J.; Chow, J.C.; Watson, J.G.; Rhoads, G.G.; and Lioy, P.J.
(1998). Chemical mass balance source apportionment of lead in house dust. Enivron.
Sci. Technol, 32(1): 108-114.
Ahlberg, M.S.; and Adams, F.C. (1978). Experimental comparison of photon and particle-
induced X-ray emission analysis of air particulate matter. X-Ray Spectrometry, 7(2):73.
Anthony, T.R.; and Flynn, M.R. (2006). Computational fluid dynamics investigation of particle
inhalability. J. Aerosol Sci., 37(6):750-765.
Ariola, V.; D'Alessandro, A.; Lucarelli, F.; Marcazzan, G.; Mazzei, F.; Nava, S.; Garcia-
Orellana, I; Prati, P.; Valli, G.; Vecchi, R.; and Zucchiatti, A. (2006). Elemental
characterization of PMio, PM2.5 and PMi in the town of Genoa (Italy). Chemosphere,
62(2):226-232.
Ashley, K.; Song, R.G; Esche, C.A.; Schlecht, P.C.; Baron, P.A.; and Wise, T.J. (1999).
Ultrasonic extraction and portable anodic stripping voltammetric measurement of lead in
paint, dust wipes, soil, and air: An interlaboratory evaluation. J. Environ. Monit.,
l(5):459-464.
Bai, Z.P.; Yiin, L.M.; Rich, D.Q.; Adgate, J.L.; Ashley, P.J.; Lioy, P.J.; Rhoads, G.G.; and
Zhang, J.F. (2003). Field evaluation and comparison of five methods of sampling lead
dust on carpets. AIHA Journal, 64(4):528-532.
Bandhu, H.K.; Puri, S.; Garg, M.L.; Singh, B.; Shahi, J.S.; Mehta, D.; Swietlicki, E.; Dhawan,
K.D.; Mangal, P.C.; and Singh, N. (2000). Elemental composition and sources of air
pollution in the city of Chandigarh, India, using EDXRF and PIXE techniques. Nuclear
Instruments and Methods in Physics Research, 160( 1): 126-138.
Benarie, M.M. (1977). Continuous sampling of urban suspended matter over large grain size
spectrum - Validity of the Hivol sampler and other results. Atmos. Environ., 11:527-529.
Bettinelli, M.; Baroni, U.; Bilei, F.; and Bizzarri, G. (1997). Characterization of SCR-DENOx
materials by ICP-MS and ICP-AES: Comparison with XRF and NAA. Atomic
Spectroscopy, 18(3):71-79.
Blanchard, G.E.; and Romano, DJ. (1978). High volume sampling: Evaluation of an inverted
sampler for ambient TSP measurement. J. Air Poll. Control Assoc., 28(11): 1142-1145.
B-ll
-------
Bonner, N.A.; Bazan, F.; and Camp, D.C. (1973). Elemental analysis of air filter samples using
x-ray fluorescence. Report No. UCRL-51388. Prepared for U.S. Atomic Energy
Commission, by Univ. of Calif, Lawrence Livermore Laboratory, Livermore, CA.
Burton, R.M.; and Lundgren, D.A. (1987). Wide-Range Aerosol Classifier - A size selective
sampler for large particles. AerosolSci. Technol, 6(3):289-301.
Cahill, T.A.; Matsumura, R.T.; Perley, B.; Eldred, R.A.; Wilkinson, K.; and Kusko, B.H.(1989).
Total elemental analysis of filters in SCAQS, Anaheim, CA. Proceedings, 82nd Annual
Meeting of the Air & Waste Management Association, Pittsburgh, PA.
Camp, D.C.; Cooper, J.A.; and Rhodes, J.R. (1974). X-ray fluorescence analysis results of a first
round intercomparison study. X-Ray Spectrometry, 3:47.
Camp, D.C.; van Lehn, A.L.; and Rhodes, J.R. (1978). Intercomparison of trace element
determinations in simulated and real air particulate samples. X-Ray Spectrometry, 4:123.
Chahal, H.S.; and Hunter, D.C. (1976). High volume air sampler: An orifice meter as a
substitute for a rotameter. J. Air Poll. Control Assoc., 26(12): 1171-1172.
Chow, J.C.; Fujita, E.M.; Watson, J.G.; Lu, Z.; Lawson, D.R.; and Ashbaugh, L.L. (1994).
Evaluation of filter-based aerosol measurements during the 1987 Southern California Air
Quality Study. Environ. Mon. Assess., 30(1):49-80.
Chow, J.C. (1995a). 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. (1995b). Summary of the 1995 A&WMA critical review: Measurement methods to
determine compliance with ambient air quality standards for suspended particles. EM,
l(May): 12-15.
Chow, J.C.; and Watson, J.G. (2008). New directions: Beyond compliance air quality
measurements. Atmos. Environ., 42(21):5166-5168.
Clements, H.A.; McMullen, T.B.; Thompson, R.J.; and Akland, G.G. (1972). Reproducibility of
the hivol sampling method under field conditions. J. Air Poll. Control Assoc.,
22(12):955-958.
Code of Federal Regulations (2007a). Appendix B to Part 50-Reference method for the
determination of suspended particulate matter in the atmosphere (high volume method).
CFR, 40(50):26-38. http://frwebgatel.access.gpo.gov/cgi-
bin/PDFgate.cgi?WAISdocID=05434191166+l+l+0&WAISaction=retrieve.
B-12
-------
Code of Federal Regulations (2007b). Appendix G to Part 50-Reference Method of the
determination of lead in suspended particulate matter collected from ambient air. CFR,
40(50):57-63. http://frwebgatel.access.gpo.gov/cgi-
bin/PDFgate.cgi?WAISdocID=05434191166+l+l+0&WAISaction=retrieve.
Criss, J.W.; and Birks, L.S. (1968). Calculation methods for fluorescent X-ray spectrometry.
Empirical coefficients vs. fundamental parameters. Anal. Chem., 40(7): 1080-1086.
Drake, P.L.; Lawryk, N.J.; Ashley, K.; Sussell, A.L.; Hazelwood, K.J.; and Song, R.G. (2003).
Evaluation of two portable lead-monitoring methods at mining sites. Journal of
Hazardous Materials, 102( 1): 29-3 8.
DRI (2007). DRI SOP 2-209r3: X-ray fluorescence (XRF) analysis of aerosol filter samples
(PANalytical Epsilon 5). Prepared by Desert Research Institute, Reno, NV.
Dzubay, T.G.; Lamothe, P.J.; and Yoshuda, H. (1977). Polymer films as calibration standards
for X-ray fluorescence analysis. Adv. X-Ray Anal., 20:411.
Dzubay, T.G.; and Nelson, R.O. (1975). Self absorption corrections for X-ray fluorescence
analysis of aerosols. In Advances in X-Ray Analysis, Vol. 18, W.L. Pickles, C.S. Barrett,
J.B. Newkirk, and C.O. Rund, Eds. Plenum Publishing Corporation, New York, NY, pp.
619-631.
Egami, R.T.; Watson, J.G.; Rogers, C.F.; Ruby, M.G.; Rood, M.J.; and Chow, J.C. (1989).
Particle fallout container measurement of dustfall from the atmosphere. In Methods of
Air Sampling and Analysis, 3 ed., J.P. Lodge, Ed. Lewis Publishers, Chelsea, MI, pp.
440-445.
Farfel, M.R.; Orlova, A.O.; Lees, P.S.J.; Bowen, C.; Elias, R.; Ashley, P.J.; and Chisolm, J.J., Jr.
(2001). Comparison of two floor mat lead dust collection methods and their application
in pre-1950 and new urban houses. Environ. Sci. Technol, 35(10):2078-2083.
Fernandez, FJ. (1989). Atomic absorption spectroscopy. InMethods of Air Sampling and
Analysis, 3 ed., J.P. Lodge, Jr., Ed. Lewis Publishers, Chelsea, MI, pp. 83-89.
Fitz, D.R.; Chan, M.; Cass, G.R.; Lawson, D.R.; and Ashbaugh, L.L.(1989). A multi-component
size-classifying aerosol and gas sampler for ambient air monitoring, Anaheim, CA.
Proceedings, 82nd Annual Meeting of the Air & Waste Management Association,
Pittsburgh, PA.
Giauque, R.D.; Garrett, R.B.; and Goda, L.Y. (1977). Calibration of energy-dispersive X-ray
spectrometers for analysis of thin environmental samples. In X-Ray Fluorescence
Analysis of Environmental Samples, T.G. Dzubay, Ed. Ann Arbor Science Publishers,
Ann Arbor, MI, pp. 153-181.
B-13
-------
Gigante, G.E.; and Gonsior, B. (2000). Comparison of different excitation methods for X-ray
spectral analysis: the case of synchrotron radiation. Fresenius Journal of Analytical
Chemistry, 368(7):644-648.
Gilfrich, J.V.; Burkhalter, P.G.; and Birks, L.S. (1973). X-ray spectrometry for participate air
pollution - A quantitative comparison of techniques. Anal. Chem., 4:2002.
Grennfelt, P.; Akerstrom, A.; and Brosset, C. (1971). Determination of filter-collected airborne
matter by X-ray fluorescence. Atmos. Environ., 5(1).
Harper, M.; Hallmark, T.S.; and Bartolucci, A.A. (2002). A comparison of methods and
materials for the analysis of leaded wipes. J. Environ. Monit., 4(6):1025-1033.
Harper, M.; Hallmark, T.S.; Andrew, M.E.; and Bird, AJ. (2004). A comparison of X-ray
fluorescence and wet chemical analysis of air filter samples from a scrap lead smelting
operation. J. Environ. Monit., 6(10):819-826.
Harper, M.; Pacolay, B.; and Andrew, M.E. (2005). A comparison of X-ray fluorescence and
wet chemical analysis for lead on air filters from different personal samplers used in a
bronze foundry. J. Environ. Monit., 7(6):592-597.
Harper, M.; and Pacolay, B. (2006). A comparison of X-ray fluorescence and wet chemical
analysis for lead on air filters from different personal samplers used in a secondary lead
smelter/solder manufacturer. J. Environ. Monit., 8(1): 140-146.
Harper, M.; Pacolay, B.; Hintz, P.; and Andrew, M.E. (2006). A comparison of portable XRF
and ICP-OES analysis for lead on air filter samples from a lead ore concentrator mill and
a lead-acid battery recycler. J. Environ. Monit., 8(3):384-392.
Harper, M.; Pacolay, B.; Hintz, P.; Barfly, D.L.; Slaven, I.E.; and Andrew, M.E. (2007).
Portable XRF analysis of occupational air filter samples from different workplaces using
different sampler: final results, summary and conclusion. J. Environ. Monit., 9:1263-
1270.
Harris, A.R.; Fifarek, B.J.; Davidson, C.I.; and Blackmon, R.L. (2006). Stationary sources of
airborne lead: A comparison of emissions data for southern California. J. Air Waste
Manage. Assoc., 56(4):512-517.
Herner, J.D.; Green, P.G.; and Kleeman, MJ. (2006). Measuring the trace elemental
composition of size-resolved airborne particles. Environ. Sci. Technol, 40(6): 1925-1933.
Hu, S.; Seshadri, S.; and McFarland, A.R. (2007). CFD study on compound impaction in a jet-
in-well impactor. Aerosol Sci. Technol., 41(12): 1102-1109.
Hunter, C.B.; and Rhodes, J.R. (1972). Particle size effects in X-ray emission analysis:
Formulae for continuous size distribution. X-Ray Spectrometry, 1:107-111.
B-14
-------
Kenny, L.C.; Gussman, R.A.; and Meyer, M.B. (2000). Development of a sharp-cut cyclone for
ambient aerosol monitoring applications. AerosolSci. Technol, 32(4):338-358.
Kenny, L.C.; Merrifield, T.; Mark, D.; Gussman, R.; and Thorpe, A. (2004). The development
and designation testing of a new USEPA-approved fine particle inlet: A study of the
USEPA designation process. Aerosol Sci. Technol., 38(Suppl.2):15-22.
Keppler, J.F.; Maxfield, M.E.; Moss, W.D.; Tietjen, G.; and Linch, A.L. (1970). Interlaboratory
evaluation of the reliability of blood lead analysis. J. Am. Ind. Hyg. Assoc., 31:412.
Kilbride, C.; Poole, J.; and Hutchings, T.R. (2006). A comparison of Cu, Pb, As, Cd, Zn, Fe, Ni
and Mn determined by acid extraction/ICP-OES and ex situ field portable X-ray
fluorescence analyses. Environ. Poll., 143(1): 16-23.
Kim, N.S.; Kim, J.H.; Ahn, K.D.; and Lee, B.K. (2007). Use of field-portable x-ray fluorescence
(FPXRF) Analyzer to measure airborne lead levels in Korean Workplaces. Journal of
Occupational Health, 49:493-498.
Lemieux, P.M.; Ryan, J.V.; French, N.B.; Haas, W.J.; Priebe, S.; and Burns, D.B. (1998).
Results of the September 1997 DOE/EPA demonstration of multimetal continuous
emission monitoring technologies. Waste Management, 18(6-8):385-391.
Lin, J.M.; Fang, G.C.; Holsen, T.M.; and Noll, K.E. (1993). A comparison of dry deposition
modeled from size distribution data and measured with a smooth surface for total particle
mass, lead and calcium in Chicago. Atmos. Environ., 27A(7):1131-1138.
Lubecki, A. (1969). Theoretical discussion of methods of elimination of matrix effects in non-
dispersive X-ray analysis. J. Radioanal. Chem., 2:3-18.
Lundgren, D.A.; Hausknecht, B.J.; and Burton, R.M. (1984). Large particle size distribution in 5
United States cities and the effect on a new ambient particulate matter standard (PMi0).
Aerosol Sci. Techno!., 3(4):467-473.
Matsumura, R.T.; and Cahill, T.A. (1991). Spatial inhomogeneities in SCAQS filters. Prepared
for , Research Division, California Air Resources Board, by Air Quality Group, Crocker
Nuclear Lab, University of California, Davis, CA.
McKee, H.C.; Childers, R.E.; and Saenz, O. (1971). Collaborative study of reference method for
the determination of suspended particulates in the atmosphere (High Volume Method).
Report No. APTD-0904. Prepared by U.S. Environmental Protection Agency, Research
Triangle Park, NC.
Menzel, N.; Schramel, P.; and Wittmaack, K. (2002). Elemental composition of aerosol
particulate matter collected on membrane filters: A comparison of results by PIXE and
ICP-AES. Nuclear Instruments & Methods in Physics Research Section B-Beam
Interactions with Materials and Atoms, 189:94-99.
B-15
-------
Moreira, F.R.; Borges, R.M.; and Oliveira, R.M. (2005). Comparison of two digestion
procedures for the determination of lead in lichens by electrothermal atomic absorption
spectrometry. Spectrochimica Acta Part B-Atomic Spectroscopy, 60(5):755-758.
Munksgaard, N.C.; and Parry, D.L. (1998). Lead isotope ratios determined by ICP-MS:
Monitoring of mining-derived metal particulates in atmospheric fallout, Northern
Territory, Australia. Sci. Total Environ., 217(1-2): 113-125.
NIOSH (1994a). Method 7082: Lead by flame AAS (Issue 2). Report No. 4th. Prepared by
National Institute of Occupational Safety and Health, Cincinnati, OH.
http://www.cdc.gov/niosh/nmam/pdfs/7082.pdf
NIOSH (1994b). Method 7105: Lead by GFAAS (Issue 2). Report No. 4th. Prepared by
National Institute of Occupational Safety and Health, Cincinnati, OH.
http://www.cdc.gOv/niosh/nmam/pdfs/7105 .pdf
NIOSH (1998). Method 7702: Lead by field portable XRF (Issue 1). Report No. 4th. Prepared
by National Institute of Occupational Safety and Health, Cincinnati, OH.
http://www.cdc.gov/niosh/nmam/pdfs/7702.pdf
NIOSH (2003a). Method 7300: Elements by ICP (Aqua regia ashing) (Issue 1). Report No. 4th.
Prepared by National Institute of Occupational Safety and Health, Cincinnati, OH.
http://www.cdc.gov/niosh/nmam/pdfs/7301.pdf
NIOSH (2003b). Method 7300: Elements by ICP (Nitric/perchloric acid ashing) (Issue 3).
Report No. 4th. Prepared by National Institute of Occupational Safety and Health,
Cincinnati, OH. http://www.cdc.gov/niosh/nmam/pdfs/7300.pdf.
NIOSH (2003c). Method 7303: Elements by ICP (Hot Block/HCl/HNO3 Digestion) (Issue 1).
Report No. 4th. Prepared by National Institute of Occupational Safety and Health,
Cincinnati, OH. http://www.cdc.gov/niosh/nmam/pdfs/7303.pdf
NIOSH (2003d). Method 7701: Lead by portable ultrasonic extraction ASV (Issue 2). Report
No. 4th. Prepared by National Institute of Occupational Safety and Health, Cincinnati,
OH. http://www.cdc.gov/niosh/nmam/pdfs/7701.pdf
NIST (2008). Standard reference materials: 105.13 Lead in paint, dust, and soil (powder and
sheet forms). Prepared by National Institute of Standards and Technology, Gaithersburg,
MD. https://srmors.nist.gov/tables/view table.cfm?table=l05-13.htm.
Nottrodt, K.H.; Georgii, H.W.; and Groeneveld, K.O. (1978). Absolute element concentrations
in aerosols analyzed by atomic absorption spectrometry and by proton-induced X-ray
emission, a comparison. J. Aerosol Sci., 9:169.
Parkes, J.; Rabbitt, L.J.; and Hampshire, MJ. (1974). Live peak-stripping during X-ray energy
dispersive analysis. Anal. Chem., 46(12):1830-1831.
B-16
-------
Peters, T.M.; Gussman, R.A.; Kenny, L.C.; and Vanderpool, R.W. (200la). Evaluation of PM2.5
size selectors used in speciation samplers. Aerosol Sci. Technol., 34(5):422-429.
Peters, T.M.; Norris, G.A.; Vanderpool, R.W.; Gemmill, D.B.; Wiener, R.W.; Murdoch, R.W.;
McElroy, F.F.; and Pitchford, M.L. (2001b). Field performance of PM2.5 Federal
Reference Method Samplers. Aerosol Sci. Technol.,34(5)A33-443.
Peters, T.M.; Vanderpool, R.W.; and Wiener, R.W. (2001c). Design and calibration of the EPA
PM2.5 well impactor ninety-six (WINS). Aerosol Sci. Technol., 34(5):389-397.
Pyle, S.M.; Nocerino, J.M.; Deming, S.N.; Palasota, J.A.; Palasota, J.M.; Miller, E.L.; Hillman,
D.C.; Kuharic, C.A.; Cole, W.H.; Fitzpatrick, P.M.; Watson, M.A.; and Nichols, K.D.
(1996). Comparison of AAS, ICP-AES, PSA, and XRF in determining lead and
cadmium in soil. Environ. Sci. Technol., 30(1):204-213.
Reynolds, S.J.; Etre, L.; Thorne, P.S.; Whitten, P.; Selim, M.; and Popendorf, WJ. (1997).
Laboratory comparison of vacuum, OSHA, and HUD sampling methods for lead in
household dusts. J. Am. Ind. Hyg. Assoc., 58(6):439-446.
Rhodes, J.R.; and 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.
Rich, D.Q.; Yiin, L.M.; Rhoads, G.G.; Glueck, D.H.; Weisel, C.; and Lioy, P.J. (1999). A field
comparison of two methods for sampling lead in household dust. J. Expo. Anal. Environ.
Epidemiol.,9(2):106-112.
RTI (2004). Standard operating procedure for the x-ray fluorescence analysis of PM2 5 deposits
on Teflon filters. Prepared by Environ. & Industrial Meas. Div., Research Triangle
Institute, Research Triangle Park, NC.
http://www.epa.gov/ttn/amtic/files/ambient/pm25/spec/xrfsop.pdf
Russ, J.C. (1977). Processing of energy-dispersive X-ray spectra. X-Ray Spectrometry, 6(1):37-
55.
Scheffe, R.; Hubbell, B.J.; Fox, T.D.; Rao, V.B.; and Pennell, W.R. (2007). The Rationale for a
Multipollutant, Multimeda Air Quality Management Framework. EM, May: 14-20.
Smith, F.; and Nelson, A.C., Jr. (1973). Reference method for the determination of suspended
particulates in the atmosphere (High Volume Method). Report No. EPA-R4-73-028b.
Prepared by U.S. Environmental Protection Agency, Research Triangle Park, NC.
Statham, P.J. (1976). A comparative study of techniques for quantitative analysis of the x-ray
spectra obtained with a Si(Li) detector. X-Ray Spectrometry, 5:16-28.
Statham, P.J. (1977). Deconvolution and background subtraction by least-squares fitting with
prefiltering of spectra. Anal. Chem., 49(14):2149-2154.
B-17
-------
Sterling, D.A.; Lewis, R.D.; Luke, D.A.; and Shadel, B.N. (2000). A portable X-ray
fluorescence instrument for analyzing dust wipe samples for lead: Evaluation with field
samples. Environmental Research, 83(2): 174-179.
Sussell, A.; and Ashley, K. (2002). Field measurement of lead in workplace air and paint chip
samples by ultrasonic extraction and portable anodic stripping voltammetry. J. Environ.
Monit.,4(iy.l56-161.
U.S.EPA (1982). Reference Method for the determination of suspended paniculate matter in the
atmosphere (High-Volume Method) - 40 CFR Part 50. Federal Register, 48:54912.
U.S.EPA (1983). Reference Method for the Determination of Suspended Particulate Matter in
the Atmosphere (High-Volume Method) - 40 CFR Part 50 Appendix B. Federal
Register, 48:17355.
U.S.EPA (1987). Reference Method for the Determination of Particle Matter as PMio in the
Atmosphere. Federal Register, 52:29467.
U.S.EPA (1989). Ambient air monitoring reference and equivalent methods; Reference method
designation. Federal Register, 54(56): 12273.
U.S.EPA (1999a). Compendium Method IO-3.3: Determination of metals in ambient parti culate
matter using X-ray fluoescence (XRF) spectroscopy. Report No. EPA/625/R-96/010a.
Prepared by U.S. Environmental Protection Agency, Research Triangle Park, NC.
http ://www. epa. gov/ttn/amtic/files/ambient/inorganic/mthd-3 -3 .pdf.
U.S.EPA (1999b). Compendium Method IO-3.4: Determination of metals in ambient particulate
matter using inductively coupled plasma spectrometry (ICP). Report No. EPA/625/R-
96/0lOa. Prepared by U.S. Environmental Protection Agency, Research Triangle Park,
NC. http://www.epa.gov/ttn/amtic/files/ambient/inorganic/mthd-3-4.pdf.
U.S.EPA (1999c). Compendium Method IO-3.2: Determination of metals in ambient particulate
matter using atomic absorption (AA) spectroscopy. Report No. EPA/625/R-96/010a.
Prepared by U.S. Environmental Protection Agency, Research Triangle Park, NC.
http://www.epa.gov/ttn/amtic/files/ambient/inorganic/mthd-3-2.pdf
U.S.EPA (1999b). Compendium Method IO-3.6: Determination of metals in ambient particulate
matter using proton induced X-ray emission (PIXE) spectroscopy. Report No.
EPA/625/R-96/010a. Prepared by U.S. Environmental Protection Agency, Research
Triangle Park, NC. http://www.epa.gov/ttn/amtic/files/ambient/inorganic/mthd-3-6.pdf
U.S.EPA (2005). National ambient air monitoring strategy. Prepared by U.S. Environmental
Protection Agency, Research Triangle Park, NC.
http://www.epa.gov/ttn/amtic/files/ambient/monitorstrat/naamstrat2005.pdf
B-18
-------
U.S.EPA (2006). National ambient air quality standards for particulate matter: Proposed rule.
Federal Register, 71(10):2620-2708.
van der Meulen, A.; Hofschreuder, P.; van de Vate, J.F.; and Oeseburg, F. (1984). Feasibility of
high volume sampling for determination of total suspended parti culate matter and trace
metals. J. Air Poll. Control Assoc., 34(2): 144-151.
VanCott, R.J.; McDonald, B.J.; and Seelos, A.G. (1999). Standard soil sample preparation error
and comparison of portable XRF to laboratory AA analytical results. Nuclear
Instruments & Methods in Physics Research Section A-Accelerators Spectrometers
Detectors and Associated Equipment, 422(1-3):801-804.
Walder, A.J.; and Furuta, N. (1993). High-precision Pb isotope ratio measurement by
inductively coupled plasma multiple collector mass spectrometry. Anal. Sci., 9:675-681.
Watson, J.G. (1979). Chemical element balance receptor model methodology for assessing the
sources of fine and total suspended particulate matter in Portland, Oregon. Ph.D.
Dissertation, Oregon Graduate Center, Beaverton, OR.
Watson, J.G.; Bowen, J.L.; Chow, J.C.; Rogers, C.F.; Ruby, M.G.; Rood, M.J.; and Egami, R.T.
(1989). High volume measurement of size classified suspended particulate matter. In
Methods of Air Sampling and Analysis, 3rd ed., J.P. Lodge, Ed. Lewis Publishers, Inc.,
Chelsea, MI, pp. 427-439.
Watson, J.G.; and Chow, J.C. (1993). Ambient air sampling. In Aerosol Measurement:
Principles, Techniques and Applications., K. Willeke and P. A. Baron, Eds. Van
Nostrand, Reinhold, New York, NY, pp. 622-639.
Watson, J.G.; Chow, J.C.; Rogers, C.F.; Green, M.C.; Kohl, S.D.; Frazier, C.A.; Robinson, N.F.;
and DuBois, D.W. (1997). Annual report for the Robbins Particulate Study - October
1995 through September 1996. Report No. 7100.3F2. Prepared for VERSARInc.,
Lombard, IL, by Desert Research Institute, Reno, NV.
Watson, J.G.; Chow, J.C.; and Frazier, C.A. (1999a). X-ray fluorescence analysis of ambient air
samples. In Elemental Analysis of Airborne Particles, Vol. 7, S. Landsberger and M.
Creatchman, Eds. Gordon and Breach Science, Amsterdam, pp. 67-96.
Watson, J.G.; Chow, J.C.; Kohl, S.D.; Kuhns, H.D.; Robinson, N.F.; Frazier, C.A.; and
Etyemezian, V. (1999b). Annual report for the Robbins Particulate Study: October 1996
through September 1997. Report No. 7100.4F3. Prepared for VERSAR, Inc., Lombard,
IL, by Desert Research Institute, Reno, NV.
Watson, J.G.; Chow, J.C.; Kohl, S.D.; Kuhns, H.D.; Robinson, N.F.; Frazier, C.A.; and
Etyemezian, V. (2000). Annual report for the Robbins Particulate Study - October 1997
through September 1998. Prepared for Versar Inc., Lombard, IL, by Desert Research
Institute, Reno, NV.
B-19
-------
Watson, J.G.; and Chow, J.C. (2001). Ambient air sampling. In Aerosol Measurement:
Principles, Techniques, and Applications, Second Edition, 2nd ed., P. Baron and K.
Willeke, Eds. John Wiley & Sons, New York, NY, pp. 821-844.
Wedding, J.B.; McFarland, A.R.; and Cermak, I.E. (1977). Large particle collection
characteristics of ambient aerosol samplers. Enivron. Sci. TechnoL, 11(4):387-390.
Witz, S.; Smith, T.B.; Shu, M.; and 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.
B-20
-------
Mr. Bart Croes
Overall, the documents provided to the Subcommittee continue the impressive responsiveness by
U.S. EPA staff to CAS AC and our Subcommittee's comments. Staff should be commended for
taking a systematic approach towards implementation of a likely revised lead (Pb) National
Ambient Air Quality Standard (NAAQS). I appreciate the opportunity to comment during
several stages of the process, and agree with the basic approach taken by U.S. EPA. My
comments address the consultation questions posed by Lewis Weinstock in his June 15, 2008
memo to Fred Butterfield. 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.
Charge Questions:
Attachment 1 - Draft Federal Reference Method (FRM) Lead in PM10 (Pb-PMlO)
1. What are your comments on the use of the low-volume PMioc FRM sampler as the Pb-PMw
FRM sampler?
Replacing the current, high volume FRM with a low volume sampler based on PMlOc and
PM2.5 FRMs is desirable for the following reasons:
> Low volume sampling offers advantages in pressure/temperature flow correction for
sample collection in local (actual) conditions.
> Low volume samplers have solid state electronic controls and data logging while high
volume samplers utilize mechanical timers and have no data logging capability,
> Low volume samplers offer the opportunity for remote operation and data access
where high volume samplers do not.
> Quartz and glass fiber filters used in high volume sampling have far higher
background levels of Pb than Teflon filters used in low volume sampling.
Leaving the door open to potential FEMs is desired. For example, the ARB Toxics network
(Xontech 924, low volume TSP, Teflon filter, ICP-MS) may or may not be equivalent, but
California should have the opportunity to find out.
2. What are your comments on the use ofXRF as the Pb-PMw FRM analysis method?
While it has problems with non-uniform deposits, XRF provides an efficient method of
analysis and requires less sample preparation than other analytical methods. The other
species will also allow source apportionment.
3. What are your comments on the specific analysis details of the XRF analysis method
contained in the proposedPb-PMw FRM analysis method description?
The description as written was adequate.
B-21
-------
4. Do you think the precision, bias andMDL of the XRF method for the proposed Pb range will
be adequate?
The MDL for Pb is well below 0.001 ug/m3 (looking at California data) for a 24-hour
sample, so there should be no problem with determining compliance for the new standard.
Most of the lead samples at the (very clean) IMPROVE sites are valid.
Using IMPROVE XRF as an example, here are data from the Agua Tibia site, north of
Escondido in northern San Diego County. The scatterplot of reported uncertainty vs.
concentration shows good performance across the range of concentrations reported, with
most concentrations in the well-quantified zone.
The vertical lines denote (left to right):
- The mean reported MDL.
- Warren White's "Rule of Thumb" MQL (10 x MDL).
- The proposed new standard (0.2 ug/m3).
- The old standard (1.5 ug/m3).
0.00
1e-005
0.0001 2 3 * 56Q.001 2 -3 * 56 Q .01 2 3 I 56T 0.1 2 3
PbCONCug/m3
2 3 i 56T 10
Based on this quick look, commercial XRF systems are capable of very good quantification
near the proposed new standard.
5. Are there any method interferences that we have not considered?
Not to my knowledge.
B-22
-------
Dr. Kenneth Demerjian
Comments regarding monitoring methods for the measurement of PM lead in the
atmospheric re: Draft Federal Reference Method (FRM) for Lead in Pb-PMio
1. What are your comments on the use of the low-volume PMioc FRM sampler as the Pb-
PMio FRM sampler?
The application of the PMlOc FRM sampler is an acceptable approach for the monitoring of
lead. It leaves open the possibility of missing Pb exposure from PM-Pb > 10 um diameter
particles. Measuring the concentration of PM-Pb as a function of particle size at a select number
of representative monitoring sites would address this size cut issue and the data would likely be
informative to the health community as well.
2. What are your comments on the use of XRF as the Pb-PMio FRM analysis
method?
I do not agree with this choice. I recommend that ICP-MS be the FRM for the analysis of Pb and
that XRF be considered as a FEM. The ICP-MS has better overall quality assurance and quality
control (QA/QC) and traceable standards than the XRF method. In addition, the extraction and
digestion of Pb compounds in ICP-MS analyses has proven to be quite effective and efficient
(Qureshi, et al., 2006).
Among the issues raised regarding XRF, the uniformity of material on the filter collection
surface and the potential role of large particle contributions to this non homogeneity remain of
greatest concern. It would seem prudent to study these issues prior to formally committing to a
decision on sampler type and the performance requirements of the analytical methods. The fact
that the TSP Pb measurement has been of historical poor quality in terms of particle size
sampling, should not be used as a rationalization that any incremental improvement in PM-Pb
monitoring is better than the status quo.
3. What are your comments on the specific analysis & details of the XRF analysis
method contained in the proposed Pb-PMio FRM analysis method description?
An effort should be made to archive and test filter blanks by batch number.
4. Do you think the precision, bias and MDL of the XRF method for the proposed Pb
range will be adequate?
The approach described is adequate for characterizing the performance of the XRF analysis for
Pb under ideal filter sample collection. It is clear from discussions among committee members
that significant uncertainties remain with regard to XRF's quantification. These include
potential effects of sampling inlets, Pb particle size and the uniformity of collected PM on the
filter. D. Felton's comments, present data which indicate the extreme sensitivity in precision
and accuracy with respect to ambient Pb concentration levels and certainly makes the case for
B-23
-------
the need to reconsider the statistical measures for precision and accuracy for the low Pb
concentrations typical observed in urban areas (e.g. figure 1 below)
DOE
DJCW-
DOC -\
-STN PW2 j FB HeaELrernens n Mew Ycrt City (4 manners)
£303 2001 2032 2033 2004 23DE 2006
Figure 1. BoiplotT &T>" PM2.5 Pb Concentration: in New York City
5. Are there any method interferences that we have not considered?
All standard sources of interference have been identified. The low levels ambient PM Pb in the
atmosphere will continue to be a challenge and require maintaining filter blank quality and
monitoring the integrity of sample handling and potential contamination sources within the
sample collection system.
Qureshi, S., V. A. Dutkiewicz, K. Swami, K. X. Yang, L. Husain, J. J. Schwab, and K. L. Demerjian, 2006. Elemental
Composition of PIVhs Aerosols in Queens, New York: Solubility and Temporal trends, Atmospheric Environment, 40,
S238-S251.
B-24
-------
Dr. Delbert Eatough
Comments on Draft Federal Reference Method (FRM) for Lead in Ph-PM/m
What are your comments on the use of the low-volume PMioc FRM sampler as the Pb-PMio
FRM sampler?
I fully support the suggestion. The sampler is well characterized, available and
compatible with other instruments in existing networks. For reasons stated in Section 2., I do not
believe that basing the standard on a low volume TSP sampler is a good idea at this time.
What are your comments on the use of XRF as the Pb-PMi0 FRM analysis method?
I am not an expert on XRF but concur with the points made by others that the
establishment of ICP-MS as the FRM with XRF as a FEM is a reasonable direction to go. The
reasons given for going this direction as discussed in the call included: 1) Availability of the
technique in many states; 2) Ease of extraction and sensitivity of analysis for the techniques; and
3) Avoidance of the issues inherent with XRF if the deposit on the filter is not uniform.
What are your comments on the specific analysis & details of the XRF analysis method
contained in the proposed Pb-PMi0 FRM analysis method description?
Do you think the precision, bias and MDL of the XRF method for the proposed Pb range
will be adequate?
Are there any method interferences that we have not considered?
As I am not an expert on XRF analysis, I defer to the comments made by members of the
committee who are.
B-25
-------
Mr. Dirk Felton
Peer Review of the Draft Federal Reference Method (FRM) for Pb-PMlO: Attachment 1 -
Draft Federal Reference Method (FRM) Lead in PM-10 (Pb-PMlO)
Background and Summary: In order for monitoring data to be used in determination of
attainment with the NAAQS, the data must be collected with a FRM or FEM. A number of
options under consideration for the Pb NAAQS indicator would require the EPA to develop a
FRM and FEM criteria for the measurement of Pb in PMi0. The EPA has proposed language for
a FRM for Pb- PMio based on the existing FRM sampler for low volume PMioc in Appendix O to
Part 50 of the Code of Federal Regulations (CFR) coupled with analysis by x-ray fluorescence
(XRF). The attached document includes the proposed regulatory text for the FRM for Pb in
Charge Questions:
What are your comments on the use of the low-volume PMwc FRM sampler as the Pb-PMw FRM
sampler?
The PMioc sampler is adequate for use as the Pb-PMi0 FRM sampler. Many States
already use this sampler for NATTS PMio metals sampling. The sequential versions of
the samplers should also be designated as FRMs because future Pb PMio FEM
evaluations should use the FRM samplers and protocols most predominantly utilized in
the national network. Future FEM evaluations should be designed with the identical
sample collection interval (midnight to midnight) and filter handling procedures as
followed by the majority of the data providers for the national network.
What are your comments on the use of XRF as the Pb-PMw FRM analysis method?
Specifying XRF would make analytical problems stemming from non-uniform loading,
spectral overlap and non-ideal filter loading densities an inherent part of the FRM.
ICPMS should be the analysis method for the FRM and for the PEP audit samples.
ICPMS is more accurate and it does not require the filter to be uniformly loaded. XRF
should be designated as a cost effective FEM that is routinely compared to ICPMS
through the periodic collocation of the PEP audit program.
It should also be noted that gravimetric mass determination of the sample filter is not
required for Pb analysis.
What are your comments on the specific analysis details of the XRF analysis method contained
in the proposed Pb-PMw FRM analysis method description?
The section on background measurement and correction states that 20-30 clean blank
filters are kept in a sealed container and are used exclusively for background
measurement and correction. These should be replaced with filters that are representative
B-26
-------
of the "batches" of filters that are used for the current measurements. It is likely that
filter qualities such as thickness, density and contaminant concentrations will change over
time.
Do you think the precision, bias andMDL of the XRF method for the proposed Pb range will be
adequate?
XRF is not the most accurate method for use in a Pb FRM and if selected it should be
viewed as a compromise between cost effectiveness and accuracy at concentrations
below about 0.01 ug Pb/m3. For low concentration measurements, it is preferable to use
ICPMS which is more accurate and does not require the filter to be uniformly loaded for
the FRM. XRF should be designated as an FEM and be permitted for use unless
accuracy at very low concentrations is necessary for specific monitoring objectives.
The MDL for Pb XRF as stated in the draft Reference Method is 0.001 ug/m3. At this
concentration the Pb data is not accurate enough to be used reliably for anything other
than to demonstrate that the amount of Pb in the air is low. The EPA should consider
establishing a minimum reporting level for XRF Pb no lower than 0.005 ug/m3. Levels
below this can be reported but flagged as between detection limit and reporting limit or
set to zero if they are below 0.001 ug/m3.
The draft PMi0 method references the procedures in Appendix A, Part 58 for use in
precision calculations. CFR Appendix A, Part 58 (1997 - section 5.3.1.1) states that the
concentrations of both collocated pairs of Pb data must be above 0.15 ug/m3 in order for
the data to be used in precision calculations. This concentration will be too high for most
of the sites in the new Pb monitoring network. A lower value can be selected but the
precision of the measurement will decrease rapidly at lower concentrations. In Figure 1
below, STN PM2.5 Pb is compared to data from a collocated PM2.5 FRM in which the
filters were analyzed for Pb by XRF. This data should emulate what we would expect to
see for the precision calculations for a clean site in the proposed low volume PMio Pb
network. As we can see, the Percent Difference rapidly increases below about 0.02 ug
Pb/m3. This is only one example but it serves to demonstrate that the proposed method's
precision determination will have to account for XRF's increase in error at low
concentrations.
The EPA may have to revise the way statistics are calculated for Pb or other NAAQS
developed in the future for individual components of PM. The typical ambient
concentrations of Pb are of course much lower than those for gravimetric mass and are
closer to instrument and method detection limits. The statistics used to determine
precision and accuracy may have to be specified as a range; looser at low concentrations
where much of the ambient data will be and tighter at higher concentrations closer to the
Pb NAAQS.
B-27
-------
Figure 1: New York City: Low Volume Pb XRF % Difference
2005: STN PM-2.5 Pb and FRM PM-2.5 Pb
180
160
0.01
0.02
0.03
0.04 0.05 0.06
Pb PM-2.5
0.07
0.08 0.09
Are there any method interferences that we have not considered?
The sampler components and the shipping and handling materials for the filter samples
must be free of Pb that can affect the integrity of the sample. The metal used to produce
the sampler inlet is of particular concern and there should be a specified limit for the
amount of trace Pb that is permissible for any component of the sampler including o-rings
and greases. It would also be advisable to restrict the use of brass upstream of the sample
filter or in any part that experiences wear and is exposed to the sampler exhaust such as
in cooling fans and motor brushes.
B-28
-------
Dr. Philip Hopke
A stumbling block for the development of a new sampler from "scratch" is what are the criteria
that would be desired in such a sampler. My suggestion is that if we are concerned about a
combination of inhaled risk including deposition in the head airways that would result in
transport to the GI tract as well as hand-to-mouth behavior, then we should look at developing a
sampler that would meet the "inhalable" curve defined by industrial hygienists. Figure 1
presents the penetration curves for the typical PM size fractions.
100
2.5 4 10 20 50
Aerodynamic Diameter, Da (\im)
100
Clearly such development would take some time so I would suggest as multi-pronged approach.
There are at least three commercially available low-volume TSP heads currently on the market
(Thermo, BGI, and URG). These could be tested by Dr. Kenny in the UK or there are wind
tunnels at universities where sufficient testing is possible even it if does not fully meet 40 CFR
58 requirements. Depending on the outcome of these tests, it might be possible to denote one or
more of these as sufficiently close to the IPM curve to move ahead with these. If none of the
heads provide adequate response characteristics, then an effort can be initiated to design an inlet
that meets the established criteria.
It should be noted that any TSP head is going to be sensitive to wind speed. They are
cylindrically symmetric and thus, wind direction invariant.
B-29
-------
Dr. Kazuhiko Ito
General Comment:
I understand that, because of the schedule for the proposed new FRM for Pb-PMio, we are asked
specific charge questions at this point. However, based on the conversations that took place
during the July 14th conference call, it seems to me that there are some important uncertainties
that need to be investigated or characterized further even after the new method and alternative
low-volume TSP samplers are considered. Specifically, as Dr. Hopke pointed out, it seems
unclear if the Pb-PMi0 (or perhaps even Pb-TSP) is the most appropriate indicator of Pb
exposure if the relevant route of exposure is ingestion of surface deposited Pb.
Charge Questions and comments:
• Attachment 1 Draft Federal Reference Method (FRM) Lead in PMio (Pb-PMi0)
What are your comments on the use of the low-volume PMioc FRM sampler as the Pb-PMio
FRM sampler?
To the extent that we are interested in Pb in PMio size fraction, the PMioc sampler is
acceptable and appropriate for Pb-PMi0 FRM, given the performance shown in the past tests.
What are your comments on the use ofXRF as the Pb-PMio FRM analysis method?
I imagine the information on the issues associated with Pb analysis by XRF is available
from the nationwide PM2.5 speciation data collected since 2000. Analysis of such data would be
informative.
What are your comments on the specific analysis details of the XRF analysis method
contained in the proposed Pb-PM10 FRM analysis method description?
The document describes potential spectral interferences and spectral overlaps, but it does
not give us a sense of the extent of this problem in the real data. It would be helpful if the
document could also describe likely extent of this issue. Again, how serious a problem was this
in the nationwide PM2 5 chemical speciation data?
Do you think the precision, bias andMDL of the XRF method for the proposed Pb range will
be adequate?
I think this answer depends on the extent of spatial variation of Pb-PMio in the locations
of interest as well as the actual NAAQS level for Pb. The goal for a 15% precision for co-
located monitors may be adequate if a coefficient of variation of annual means for multiple
monitors within an area of interest is, say, 50%, but this would vary from city to city. I happened
to look at within-city variation of several PM2.5 chemical species including Pb in 28 MSAs
several years ago for a different reason (I was comparing within-city vs. across-city variation of
B-30
-------
PM components). Figure 1 shows the coefficient of variation (CV) for the across-MSA variation
vs. distributions of CV's of within-MSA variation for the 28 MSAs where there were multiple
monitors for years 2000-2003. For Pb, the CV ranges from nearly zero to 60% with the median
of- 25%. Therefore, the adequacy of precision of 15% may be OK for the cities where high Pb
levels occur (I imagine Pb-PMi0 variation would be larger than that for Pb-PM2.5).
%C.V.
8
C,V, of annuBI means x ross WSA^
Ml V Fti Zn Gr Mfl Fe 3» As 3fl 30-1 NO3 EC OC
Figure 1. Comparison of coefficient of variation (C.V.) of annual (multi-year, '00-'03) means across
MSAs (denoted with bold "-") and distribution of within-MSA C.V. of annual means in the 28 MSAs
where multiple monitors were available, "o" represents extreme value.
Are there any method interferences that we have not considered?
I don't know of any.
B-31
-------
Dr. Donna Kenski
Comments on the Peer Review for Pb NAAQS
1. What are your comments on the use of the low-volume PMioc FRM sampler as the
Pb-PM10 FRM sampler?
The PMioc FRM sampler is the obvious and best choice for a Pb-PMio FRM sampler.
2. What are your comments on the use of XRF as the Pb-PMi0 FRM analysis method?
I don't see the logic in selecting XRF over ICP/MS as the analysis method. ICP/MS is more
sensitive and not subject to the interferences that are documented in Joann Rice's memo. NIST-
traceable standards can be used for calibration and many states have in-house labs that can
perform the analysis. And, it does not require uniform filter loading. XRF is perfectly suitable
for an FEM, but I recommend that ICP/MS be selected as the FRM analytical method.
3. What are your comments on the specific analysis details of the XRF analysis method
contained in the proposed Pb-PMi0 FRM analysis method description?
The description as written was adequate.
4. Do you think the precision, bias and MDL of the XRF method for the proposed Pb
range will be adequate?
The MDL as specified is fine if determination of compliance with the NAAQS is the only issue.
But since health professionals, EPA, and others have a valid interest in determining
concentrations at levels far below the NAAQS, it seems shortsighted not to measure Pb with
higher accuracy at the (more common) low concentrations as well. With the (presumed) lowering
of the Pb NAAQS, and with generally lower ambient concentrations across the country, the
MDL should be lower than the 0.001 jig Pb/m3 that is proposed. As this is easily achievable and
already being accomplished by other national networks, it ought to be part of the FRM method.
5. Are there any method interferences that we have not considered?
Not that I know.
B-32
-------
Dr. Peter McMurry
Comments regarding measurement methods for particulate lead in atmospheric aerosols —
Comments on Draft Federal Reference Method (FRM) for Lead in Ph-PM/m
What are your comments on the use of the low-volume PMioc FRM sampler as the Pb-PMio
FRM sampler ?
I think it is a good idea. This sampler is readily available, is familiar to monitoring agencies, and
has been well-tested. Furthermore, the Pb samples would be sampled in the same way as PMi0
mass, so the fraction of mass that is Pb would be well defined.
What are your comments on the use of XRF as the Pb-PMi0 FRM analysis method?
Chemical analysis is not my primary area of expertise. Therefore, the views expressed here
represent my synthesis of comments from today's telephone conversation.
I think very compelling arguments were made to use an extraction method such as ICP-MS
rather than XRF as the FRM analysis method. These include the (1) the confidence that Pb can
be effectively extracted with efficiencies that approach 100%, (2) the availability of NIST
traceable standards for liquid solutions of Pb that can be used to calibrate analytical instruments
used to analyze dissolved extracts and the corresponding lack of such standards for deposited Pb,
(3) the availability of instruments, such as ICP-MS in states and the corresponding unavailability
of XRF instruments, (4) the sensitivity of XRF to spatial distributions of deposits on filters,
which are unlikely to be uniform (especially for coarse particles) and the corresponding
insensitivity of extraction methods to such non-uniformities, (5) the use of proprietary software
for analyzing XRF data, which shields the public from a clear understanding of how
concentrations of lead are determined, and (6) the superior sensitivity of methods such as ICP-
MS. I question whether XRF would meet the accuracy and precision goals required for a
standard.
Because XRF is inexpensive and nondestructive, I think it makes sense to use it as a FEM.
What are your comments on the specific analysis & details of the XRF analysis method
contained in the proposed Pb-PMi0 FRM analysis method description?
I will defer to other members on the committee on this.
Do you think the precision, bias and MDL of the XRF method for the proposed Pb range
will be adequate?
Again, I will defer to those members of the committee who are more knowledgeable than I on
this topic. I was left with the sense there are compelling arguments for using another analytical
method as the FRM.
B-33
-------
Are there any method interferences that we have not considered?
Not to my knowledge.
B-34
-------
Mr. Richard Poirot
Comments on Proposed Pb PMi0 FRM - Appendix Q Part 50
An important prefacing comment is that the CAS AC Lead Panel has advocated a transition of the
Pb indicator from TSP to PMio Pb if, and only if the level of the standard is set lower than 0.2
|ig/m3. If a level equal to or higher than 0.2 |ig/m3 is selected, the CASAC Pb panel is
unanimously opposed to a reduction in the indicator particle size range from TSP to
What are your comments on the use of the low-volume PMioc FRM sampler as the Pb-PMio
FRM sampler ?
Assuming the level of the Pb standard is set below 0.2 |ig/m3, the PMi0c sampler would be an
appropriate choice for a Pb-PMio FRM sampler.
What are your comments on the use ofXRF as the Pb-PMio FRM analysis method?
XRF should be an adequate analytical method for a Pb NAAQS set toward the middle to upper
end of the range of levels recommended by EPA staff and the CASAC Pb Panel. If the level is
set toward the low end of that recommended range (0.02 |ig/m3), a more accurate analytical
method like ICP-MS, with lower detection limits and smaller analytical errors, would be
preferable. Consideration should also be given to specifying ICP-MS as the FRM and
establishing XRF as a FEM.
What are your comments on the specific analysis & details of the XRF analysis method
contained in the proposed Pb-PMio FRM analysis method description?
I think it should be useful and possible to tighten up some of the specific details relating
specifically to the determination of Pb by EDXRF. Much of the description sounds more like
cautionary guidance rather than prescribed details of specific procedures, and also seems to
pertain to XRF analysis of elemental species other than Pb. This raises a related point that it
would be best to be very clear up-front about analytical and data reporting procedures for
(readily detectable) XRF elements other than Pb that may result at little or no extra cost from the
Pb XRF analysis (and would also represent an important reason in favor of the choice of XRF as
part of the Pb FRM). Arguably this "supplemental data" would have value for quality assurance
and source attribution of the Pb measurements, and if significant additional costs are not
incurred, analytical and data reporting procedures could be specified. Similar considerations
would also apply if (multi-elemental) ICP-MS were selected as the Pb FRM. Along similar
lines, clear guidance should also be provided on whether (or not) there should be PMio mass
measurements conducted on the Pb FRM filters. Possibly the Agency would want to provide for
an optional national analytical contractor, as has proved effective for IMPROVE and STN
networks. Alternatively, some consideration should be also given to coordination with the
evolving NAATS metals sampling program which generally (but not always) utilizes PMi0c
samplers combined with ICP-MS analyses (at most but not all sites), and which would benefit
from greater internal consistency.
B-35
-------
Do you think the precision, bias andMDL of the XRF method for the proposed Pb range will
be adequate?
The precision, bias and MDL of the XRF method should be adequate for a Pb NAAQS in the
currently proposed range of 0.1 to 0.3 |ig/m3, although a PMio sampling method is not
recommended if the level is greater than or equal to 0.2 |ig/m3. The XRF precision, bias and
MDL could pose problems for a NAAQS set at the lower end of the EPA staff-recommended
range of 0.02, and may result in uncertainties in spatial patterns and temporal trends at
population-oriented monitoring sites where levels are likely to fall well below the proposed
NAAQS range. The indicated XRF PMio Pb MDL of 0.001 |ig/m3 would be only 1% of the
lower bound level of the proposed NAAQS and unlikely to have a significant influence on
compliance determinations. I also think it's likely that this MDL could be further reduced. For
example the current MDL for PM2.5 Pb in the IMPROVE network is closer to 0.0001 |ig/m3.
Current Pb precision comparisons are limited to concentrations above 10% of the current
NAAQS (i.e., 0.15 |ig/m3). This limit will need to be lowered to reflect the hopefully much
lower level of the currently revised NAAQS. Also, since it generally appears likely that the
Administrator may select a level (and form) of the standard which are less stringent than are
warranted by the Agency's Risk/Exposure Assessment and Staff Paper, some consideration
should be given to collection of accurate and precise data at levels below and possibly well
below the level of the NAAQS selected in this review cycle.
Are there any method interferences that we have not considered?
None that I'm aware of- related to XRF analysis of Pb on PMioc filters. However, it should be
recognized that XRF is not very well suited for analysis of fiberglass TSP or hi-volume Quartz
PMio filters. ICP-MS would be a better choice for an FRM analytical method that could be used
across all potential filter types, and would provide a better basis for comparative sampling to
develop better information on Pb particle size distributions, sources etc. — especially in the
event that TSP (and/or hi-vol PMio) are retained (or specified as FEM).
Other minor comments on Pb PMio FRM:
p. 3, para 1, line 6: The hyphenated "24-hour sample" is correct here, but all other instances of
the number "24" in this document are also (incorrectly) attached by hyphen to the words that
follow. These include "24-hours" in line 3 and "24-cubic meters" in line 5 of this paragraph and
2 instances of "24-m3" in 2nd and 3rd paragraphs on page 5.
p. 6, para 1: You present optimal (150 jig/cm2) and minimal (15 jig/cm2) PMio mass loading
levels (roughly 75 and 7.5 |ig/m3 respectively) for XRF Pb quantification, but also indicate
potential distortion with "unusually heavy deposits". Why not also provide the PMio mass level
that would be considered unusually heavy (i.e. an upper bound to go along with the ideal and
minimal loading levels). Also, unless mass measurements are required, how will it be known
whether the filter loading is above, below or within the range where distortion-free
measurements are expected?
B-36
-------
p. 6, para 2, line 3: You could add "S/P" to this list of XRF interferences.
p. 7, para 2, line 6: This effect is "especially significant and more complex for
measurements..." than for what?
p. 8, section 6.1.2, last line: Just for curiosity, what is the basis for your selection of this blank
Pb limit of 4.8 ng/cm2? This (x 11.86 cm2 of exposed filter / 24 m3 of air sampled) would yield
an implied ambient Pb concentration of 0.002 |ig/m3 — or about twice the indicated Pb XRF
MDL - or about 1% of a standard of 0.2 |ig/m3 (are you giving us a hint about the intended level
oftheNAAQS?).
p. 9, section 6.2.3, line 2: What do you mean "Filters are typically archived in cold storage"?
For what current analyses is this cold storage procedure "typical"? Will it be required for Pb
sampling? What elements, if any, which are quantifiable by XRF do you expect to see volatized
from filters if they are not kept in cold storage prior to analysis? Certainly you don't expect any
loss ofPb, do you?
B-37
-------
Dr. Jay Turner
Peer Review: Draft Federal Reference Method (FRM) Lead in PMio (Pb-PMi0)
Charge Questions
1. What are your comments on the use of the low-volume PM10c FRM sampler as the Pb-PM10
FRM sampler? The low-volume PMioc FRM sampler is an appropriate choice as the Pb-
PMio FRM sampler. It is an adaptation of the PM2.5 FRM sampler which now has nearly ten
years of use and refinement, including both single-event and sequential configurations.
There are also operational and cost advantages to placing measurements for multiple
NAAQS on the same sampler platform. For sites specified for both PMi0 and Pb-PMi0
compliance monitoring, filter samples collected using the low-volume PMioc FRM sampler
could be subjected to both gravimetric analysis and Pb elemental analysis, providing
compliance data for both standards from a single sample.
2. What are your comments on the use ofXRF as the Pb-PM]0 FRM analysis method? I prefer
the use of ICPMS (or GFAAS) as the FRM with an expectation that XRF would be given
FEM status. While ICPMS does have the added complexity of a sample digestion step, it can
be more easily calibrated than XRF. Our recent experience with ICPMS has demonstrated
high recovery for both coal fly ash and urban particulate matter NIST Standard Reference
Materials (SRM) from quartz filters using a nitric acid and hydrochloric acid extraction
solution (following the NATTS PMio metals sampling and analysis protocol developed by
ERG). The Pb-PMio method would use Teflon filters and ERG has also developed a protocol
for this case which could be used as a starting point for the analysis method specifications.1
3. What are your comments on the specific analysis details of the XRF analysis method
contained in the proposedPb-PM10 FRM analysis method description? I defer to the XRF
experts for a critique of the analysis method details. Given the variations in instrument
hardware and software, all labs reporting compliance data based on XRF should participate
in an audit program which includes analyses of samples with traceability to ICPMS.
4. Do you think the precision, bias andMDL of the XRF method for the proposed Pb range will
be adequate? These questions are best addressed after completion of the DQO process.
Perhaps the required MDL could be relaxed depending on the NAAQS concentration value,
although a detection limit that is much lower than the standard is desirable to simplify the
data handling for concentrations below the MDL. Precision should be determined using data
with Pb concentrations above a defined threshold value since the precision reported as a
percentage CV will degrade as the MDL is approached. In general, we should be prepared
for both ICPMS and XRF data being reported to AQS, and these methods will have very
different detection limits. This will add complexity to certain non-compliance data analyses;
including trends analyses studies on Pb health effects.
1 "Standard Operating Procedure for the Determination of Metals in Ambient Particulate Matter Analyzed
by Inductively Coupled Plasma/Mass Spectrometry (ICP/MS)", prepared by ERG for EPA under Work
Assignment 5-03, ERG No.: 0143.04.005, EPA Contract No.: 68-D-00-264, September 2005.
B-38
-------
5. Are there any method interferences that we have not considered? I defer to the XRF experts
on the issue of method interferences.
B-39
-------
Dr. Warren H. White
COMMENTS ON PROPOSED FRM FOR PM10 LEAD —
Use of low-volume PMuic FRM sampler
As noted in our previous consultation, the uniformity of this sampler's collected deposit needs to
be established if XRF is used as the analytical method. Attachment 1 illustrates the need for
such a determination with an example of a non-uniform sample collected with a different (non
FRM) sampler. The elements Pb and Fe, presumed associated with different particle size
classes, show quite different deposition patterns in this example. As the x-ray beam fluoresces
only the central portion of the filter, the ratio of reported loading to ambient concentration varies
accordingly.
XRF analysis of filters from this sampler would thus respond differently to fine Pb particles from
fume sources and coarse Pb particles from dust sources.
XRF as method of analysis
XRF is cost-effective, is sensitive enough for the levels under NAAQS consideration (see
below), and fits well with other aspects the Agency's monitoring strategy and infrastructure. It
has not previously been used for a NAAQS, however, and this first application raises issues of
calibration (see below), standardization (see below), and sample uniformity (noted above) that
wet-chemical methods do not present. I think Dirk Felton's suggestion to establish XRF as an
FEM with ICP-MS as the FRM is worth considering, with the caveat that methods requiring
extraction and digestion raise their own accuracy issues.
Adequacy of XRF bias, precision, and detection limit
The adequacy of XRF measurement capabilities depends on the MQOs (measurement quality
objectives) established for the analytical method, which in turn depend on the DQOs (data
quality objectives) established for compliance monitoring. In today's discussion it was noted
that DQOs required to protect public health will themselves depend on the level and form
eventually chosen for the NAAQS. With all these considerations yet to be finalized, there are
nevertheless certain points that are already clear.
Detection: The NAAQS level proposed in the Federal Register is in the range 0.1-0.3 ug/m3.
The existing CSN and IMPROVE networks demonstrate reliable (95% probability) XRF
detection of non-spurious Pb at filter loadings of 5-7 ng/cm2 (Attachment 2). For the low-
volume PMioc FRM sampler, this corresponds to a real detection limit of about 0.003 ug/m3,
more than an order of magnitude below the lowest contemplated NAAQS level.
Precision: The declared goal for collocated precision is a 15% CV at 90% confidence. Quality
assurance for IMPROVE includes regular XRF reanalyses of a fixed collection of about 70
representative ambient samples. Over 20 reanalyses have been performed of each sample at
approximately monthly intervals, yielding some 70 well-determined analytical CVs. The typical
(median) CV obtained for Pb has been 13% (Trzepla-Nabaglo and White, 2008;
B-40
-------
http://secure.awtna.org/presentations/AerosolAttnosphericOptics08/papers/81-Trzepla-
Nabaglo.pdf). These results do not reflect flow and other sampling uncertainties but do include
observations at all concentrations, with a (relatively low) mean loading of about 12 ng/cm2. As
Dirk Felton observed, precision for collocated samples will be sensitive to the minimum
concentration included in the calculations.
Bias: The declared goal is a system bias within 10% at 95% confidence. Demonstrating
attainment of these tolerances with XRF is likely to be a challenge. The need to verify that the
sample deposit is uniform has already been noted. The other main difficulty will be the absence
of a suitable NIST-traceable standard for calibration. NIST (2002) offers "air particulate on
filter media" as SRM 2783, with a certified Pb loading of about 32 ng/cm2, but gives a 95%-
confidence uncertainty of about 17% for this value. I am not aware of any peer-reviewed
examination of the claimed accuracies of commercially available calibration foils, or even
consistency among different foils.
Specific analysis details in the FRM
Some aspects of XRF analysis require more prescriptive detail than the draft FRM gives them.
The most important are two that relate to method accuracy.
Audit filters: Bias is to be assessed "through an audit using spiked filters." The preparation of
spiked standards for XRF analysis is significantly more complicated than simply depositing a
known quantity of standard solution on a glass-fiber hi-vol filter and letting it dry, as is now
done. Deposit uniformity is needed for quantitative XRF, as noted above. Achieving this
uniformity in a liquid deposit on a Teflon membrane is likely to require attention to surface
phenomena. The most relevant spiked filter would be created by actually sampling a pure Pb-
containing aerosol and determining the Pb loading from the weight gain. XRF results for such a
filter could be compared directly with those for ambient samples, but the production of such
filters would require development and validation.
Protocols: The principals of EDXRF are universal but there is no standard protocol for
implementing them, as the Agency discovered two years ago in its effort to "harmonize" XRF
reporting from different labs used by their PM2.5 speciation networks (Gutknecht et al., 2006;
http://www.epa.gov/ttn/amtic/files/ambient/pm25/spec/xrfdet.pdf). Different instrument systems
use different x-ray spectra generated by different configurations of source anode, secondary
target, and spectral filter, different geometries of irradiation and detection, and different spectral
decomposition software based on different interpretive strategies. Much of the spectral
processing in commercial instrument systems is proprietary and invisible to the user, making it
difficult to confirm which lines are used and how they are de-convoluted. Will the Agency
undertake to certify certain commercial systems for use?
For whatever analytical method is used, field blanks should be routinely exposed and analyzed to
detect possible contamination in the field and laboratory. CSN and IMPROVE both report
loadings below 3.5 ng/cm2 in 95% of their routine field blanks (Attachment 2), significantly
exceeding the FRM's proposed filter acceptance criteria (requiring 90% to be less than 4.8
ng/cm2).
B-41
-------
Method interferences
I know of no additional method interferences.
Attachment 1: material excerpted and annotated from Nuclear Instruments and Methods in
Physics Research B 160(2000) 126-138
Elemental composition and sources of air pollution in the city of Chandigarh, India, using
EDXRF and PIXE techniques
H.K. Bandhu, Sanjiv Puri, M.L. Garg, B. Singh, IS. Shahi, D. Mehta, E. Swietlicki, O.K.
Dhawan, P.C. Mangal, Nirmal Singh
•
O
50000
40000
30000
20000
Q 10000
O
O
9
-20
-10 0 10
Distance from centre (mm)
Fig. 5. Aerosol loading distribution for different elements as a
function of distance from the center measured using 2 mm
proton beam.
Samples were collected on 47 mm diameter, 0.8 1m pore size, cellulose nitrate filter papers
(Microdevices, Ambala, India). Filter paper was mounted in an aerosol filter holder (Millipore,
Cat No. xx50 04700) having an inlet dispersion chamber to produce optimum particle
distribution on the surface of the filter. The air through the filter paper was sucked at a flow rate
of 12 1 min"1 with the help of diaphragmatic vacuum pump (Millipore, Cat. No. xx55 22050) and
critical orifice (Millipore, Cat. No. xx50 000 00). The flow rate was monitored periodically for
B-42
-------
each sample with a rotameter and no cases of reduction of flow rate due to filter clogging were
experienced during the sampling. The collection surface was directed downward to prevent
particle collection by sedimentation and the filter holder was protected with a rain cover. All the
sampling sites chosen for sampling were located on the flat roof of building tops 40-60 feet high.
Attachment 2: material excerpted and annotated from Environ. Sci. Technol. 2008, 42, 5235-
5240
An empirical approach to estimating detection limits using collocated data
Nicole P. Hyslop and Warren H. White
TABLE 1. Summary of STE\I ami IMPROVE i, estimates, LI estimates
£c 95tb percent! le LQ 95% detection probability
element network {3Q-99th percent!les) IIKJ cm~2} (92-97% probability) IIKJ cnr2
Pb STN 3.5 (1.8-8.1) 6.5(6.3-7.0)
IMPROVE 3.1 (2.6-4,2) 5.1 (4.8-6.0)
T T
From field blanks From collocated sampling
To avoid type I errors a critictif h'mit U is set such that
measurements above that limit indicate the analyte is present
with a high level of confidence. For a given probability a of
type I error, L: is the minimum value satisfying the inequality
PrL>Ll£.= Ol^ft CD
To avoid type [] errors, a frmtf of detection Lr_, is set such
that atmospheric concentrations of the anal vie at or above
that threshold will be detected with a high level of confidence.
ID is dependent on Z< because the analyte must be measured
above Lc to be considered present. For a given probability
// of type II error, ID is the minimum value satisfying the
inequality
Pr(i (2)
B-43
-------
Dr. Yousheng Zeng
Charge Question 1: What are your comments on the use of the low-volume PM10c FRM
sampler as the Pb-PMio FRM sampler?
I support the EPA proposal to use the low-volume PMi0c FRM sampler as the Pb-PMi0 FRM
sampler. The PMi0c FRM sampler is better defined and better understood than the earlier PMio
sampler. This method will also provide consistency with PMio and PM2 5 monitoring network;
data comparability for evaluation of Pb-PMio and PMio inhalation pathway; and monitoring
operation efficiency (same samplers for both PMio and Pb-PMi0).
However, I share the same concern with some committee members. With this method, the
monitoring results will be naturally lower because PMio samples, not TSP samples, will be
collected for Pb analysis. If the revised Pb NAAQS is not set low enough to account for the
absence of Pb associated with particles larger than 10 u, the new Pb NAAQS may not provide
additional protection to human health.
Charge Question 2: What are your comments on the use ofXRF as the Pb-PMio FRM
analysis method?
I support the approach proposed by Mr. Dirk Felton to use ICPMS (or AA as he mentioned
during previous consultation meeting) as FRM for sample analysis and use XRF as FEM. A
similar approach has worked well for 862 where a manual method is the reference method and
instrumental methods are FEM and widely used in day-to-day monitoring operations.
Charge Question 3: What are your comments on the specific analysis details of the XRF
analysis method contained in the proposed Pb-PMio FRM analysis method description?
I don't have comments on this issue.
Charge Question 4: Do you think the precision, bias, andMDL of the XRF method for the
proposed Pb range will be adequate?
At this point, we really don't know what will be the final Pb NAAQS. It appears that the
proposed analysis method (either XRF or ICPMS) should be adequate to produce needed
monitoring data. However, it is highly recommended to use the Data Quality Objective (DQO)
model that EAP used for evaluation of PMc in 2004. During the public conference call on July
14th, 2008, EPA indicated that EPA was working on a DQO model for Pb. It would be most
desirable to use the DQO model to help finalize these parameters (precision, bias, and MDL).
Charge Question 5: Are there any method interferences that we have not considered?
I don't have comments on this issue.
B-44
-------
Dr. Barbara Zielinska
Charge questions regarding FRM for Lead in Pb-PMlO:
1. What are your comments on the use of the low-volume PMlOc FRM sampler as the Pb-
PM10 FRM sampler?
As stated in my comments from March 23, 2008, regarding previous consultation on this
subject, I support the use of the low-volume PMlOc FRM sampler as the Pb-PMlO FRM
sampler. This sampler has been well-tested, has well-defined cut-points and slopes and is
readily available.
2. What are your comments on the use of XRF as the Pb-PMlO FRM analysis method?
Although XRF method has many advantages (is nondestructive, sensitive, relatively simple
and inexpensive), it presents some problems related to the uniformity of material on the filter
collection surface. ICP-MS method is extremely sensitive for lead, has traceable standards
and the uniformity of material is not an issue. I would recommend ICP-MS as an FRM for
the analysis of lead and XRF as an FEM (or one of the FEMs).
3. What are your comments on the specific analysis details of the XRF method contained in
the proposed Pb-PMlO FRM analysis method description?
The XRF analysis method is well described in this document. Specific analysis details were
addressed during the advisory teleconference meeting and are reflected in the lead
discussants memos.
4. Do you think the XRF method precision, bias and MDL for the proposed Pb range will be
adequate?
The method MDL, precision and bias seem to be adequate. However, for very low ambient
concentrations of Pb, it may be challenging to achieve the required precision.
5. Are there any method interferences that we have not considered?
I am not aware of any additional interference.
B-45
-------
Enclosure C - Agency's Background and Charge Memorandum to the
CASAC AAMM Subcommittee
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
June 15,2008
MEMORANDUM
SUBJECT: CASAC Peer Review and Consultation on Monitoring Issues for Lead National
Ambient Air Quality Standard (NAAQS)
FROM: Lewis Weinstock
Acting Group Leader
Ambient Air Monitoring Group
Office of Air Quality Planning and Standards (C304-06)
TO: Fred Butterfield
Designated Federal Officer
Clean Air Scientific Advisory Committee
EPA Science Advisory Board Staff Office (1400F)
Attached are materials for review by the Clean Air Scientific Advisory Committee's
(CASAC) Ambient Air Monitoring and Methods (AAMM) Subcommittee. These materials will
be the subjects of a peer review and consultation by the AAMM Subcommittee, scheduled for a
teleconference to be held on July 14, 2008. I am requesting that you forward these materials to
the AAMM Subcommittee to prepare for the peer review and consultation.
This project, entitled Lead (Pb) National Ambient Air Quality Standards (NAAQS)
Review: Monitoring Issues, has been requested by EPA's Office of Air Quality Planning and
Standards (OAQPS), within EPA's Office of Air and Radiation, in anticipation of potential
revisions to the Pb NAAQS. The peer review will cover the proposed Federal Reference Method
(FRM) for the measurement of Pb in particulate mater less than 10 micrometers in diameter (Pb-
PMio). The consultation will cover the need and approach for development of a low-volume Pb
in total suspended particulate (Pb-TSP) method as an FRM or Federal Equivalent Method
(FEM). Charge questions associated with both the peer review and the consultation are provided
below.
The upcoming consultation will support the EPA by providing scientific advice as the
EPA Administrator considers potential revisions to the Pb NAAQS; a notice of final rulemaking
C-l
-------
is to be signed by September 15, 2008. We are requesting an expedited schedule to assist EPA in
meeting the September 15, 2008 deadline for finalizing the Pb NAAQS review.
We appreciate the efforts of you and the Subcommittee to prepare for the upcoming
meeting and look forward to discussing this project in detail on July 14, 2008. Questions
regarding the enclosed materials should be directed to Mr. Kevin Cavender, EPA-OAQPS
(phone: 919-541-2364; e-mail: cavender.kevin@epa.gov).
Document Associated with Subcommittee's Peer Review:
• Attachment 1 - Draft Federal Reference Method (FRM) Lead in PM10 (Pb-PMl0)
Background and Summary: In order for monitoring data to be used in determination of
attainment with the NAAQS, the data must be collected with a FRM or FEM. A number of
options under consideration for the Pb NAAQS indicator would require the EPA to develop a
FRM and FEM criteria for the measurement of Pb in PMi0. The EPA has proposed language
for a FRM for Pb-PMi0 based on the existing FRM sampler for low volume PMi0c in
Appendix O to Part 50 of the Code of Federal Regulations (CFR) coupled with analysis by x-
ray fluorescence (XRF). The attached document includes the proposed regulatory text for the
FRM for Pb in PMi0.
Charge Questions:
What are your comments on the use of the low-volume PMwc FRM sampler as the Pb-PMw
FRM sampler?
What are your comments on the use of XRF as the Pb-PMw FRM analysis method?
What are your comments on the specific analysis details of the XRF analysis method
contained in the proposed Pb-PMw FRM analysis method description?
Do you think the precision, bias andMDL of the XRF method for the proposed Pb range will
be adequate?
Are there any method interferences that we have not considered?
Document Associated with Subcommittee's Consultation:
• Attachment 2 - Options for the Development of a Low Volume Lead in Total Suspended
Particulate (Pb-TSP) Sampler
Background and Summary: Problems with the current high-volume Pb-TSP sampler have
been highlighted as part of the on-going Pb NAAQS review. As part of the NAAQS review,
EPA proposed network design requirements that could result in the need for a significant
expansion and/or reallocation of Pb monitors. Due to the concerns over the existing high-
volume Pb-TSP sampler, EPA requested comments on the need for a FRM or FEM low-
C-2
-------
volume Pb-TSP sampler. The attached document discusses options for the development of a
low-volume Pb-TSP sampler for use in the Pb network.
Charge Questions:
Would a low-volume Pb-TSP sampler be an improvement over the existing high-volume Pb-
TSP sampler? What advantages and disadvantages do you see associated with a low-volume
Pb-TSP sampler?
What inlet designs would be best suited for a low volume Pb-TSP sampler? What designs
are not appropriate for a low-volume Pb-TSP sampler?
What is your preferred approach for the development of a low-volume Pb-TSP sampler, and
why?
If the EPA were to develop a low-volume Pb-TSP FRM, how important is it that the sampling
capture efficiency be characterized for varying particle sizes?
If the EPA were to develop a low-volume Pb-TSP FRM, should the new FRM replace the
existing high-volume Pb-TSP FRM, or should the EPA maintain the existing FRM?
Is it appropriate to accept alternative sampler and inlet designs as FEM?
Are the proposed FEM testing criteria for Pb methods adequate to ensure equivalence of
alternative sampler and inlet designs? If not, what additional testing requirements should be
considered?
Attachments
cc: Fred Dimmick, OAQPS/NERL
Robert Vanderpool, ORD/NERL
Karen Martin, OAQPS/HEID
Deirdre Murphy, OAQPS/HEID
Kevin Cavender, OAQPS/AQAD
Tim Hanley, OAQPS/ AQAD
Joann Rice, OAQPS/ AQAD
Phil Lorang, OAQPS/ AQAD
James Hemby, OAQPS/ AQAD
C-2
------- |