UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D C 20460
DEC 10 2010
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
SUBJECT: Transmittal of Science Advisory Board Report
FROM: Vanessa T. Vu
Director, Science Advisory Board Staff Office (HOOF)
TO: Karen Sheffer
EPA Headquarters Library Repository (3404T)
This is to advise you that the Science Advisory Board Clean Air Scientific Advisory
Committee (CASAC) Ambient Air Monitoring and Methods Subcommittee (AAMMS),
issued a report numbered EPA-CASAC-11-002, CASAC Review of EPA's White Paper,
"Approach for the Development of a New Federal Reference Method (FRM) for Lead in Total
Suspended Particulates (Pb-TSP)", dated November 30, 2010.
Two copies of the report are attached and a third copy has been sent electronically to
the attention of Ms. Jeannie Turner at tumer.jeannie(S).eap.gov. The report is available in
electronic format on the Science Advisory Board's Web site at http://www.epa.gov/sab.
If you have any questions regarding this report, please contact the Designated Federal
Officer, Mr. Aaron Yeow directly at (202) 564-2050.
Attachments (2)
Internet Address (URL) • httpVAvww epa gov
Recycled/Recyclable • Printed with Vegetable Oil Based Inks on 100% Postconsumer, Process Chlorine Free Recycled Paper
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON D.C. 20460
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
November 30, 2010
EPA-CASAC-11-002
The Honorable Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, D.C. 20460
Subject: CASAC Review of EPA's White Paper, "Approach for the Development of a
New Federal Reference Method (FRM) for Lead in Total Suspended
Particulates (Pb-TSP)"
Dear Administrator Jackson:
In November 2008, EPA published the final rule for the National Ambient Air Quality
Standard (NAAQS) for lead, revising the primary standard by an order of magnitude, from
1.5 ug/m3 to 0.15 ug/m3 three-month average. The indicator for the Lead NAAQS, lead in total
suspended particulates (Pb-TSP), did not change. There are a variety of air monitoring methods
that can measure ambient lead concentrations. A Federal Reference Method (FRM) is a
monitoring method designated by EPA that meets specific requirements for adequacy to
demonstrate attainment of the NAAQS. As a result of the lowering of the Lead NAAQS and
improvements in analytical measurement technologies, EPA is considering developing a new
FRM for Pb-TSP. This is outlined in EPA's white paper, Approach for the Development of a
New Federal Reference Method (FRM) for Lead in Total Suspended Particulates (Pb-TSP).
EPA requested that the Ambient Air Monitoring and Methods Subcommittee (AAMMS, or the
Subcommittee) of the Clean Air Scientific Advisory Committee (CASAC) provide a peer review
of the white paper. As part of the peer review, the Subcommittee held a public teleconference on
September 15, 2010 and prepared comments in response to EPA's four charge questions, which
centered on extraction methods, analytical methods, validation and testing, and inter-laboratory
variability. The Chartered CASAC approved this letter during a teleconference on November 8,
2010. The CASAC and AAMMS memberships are listed in Enclosure A. CASAC's consensus
responses to the Agency's charge questions are presented in Enclosure B. Individual review
comments from the AAMMS are compiled in Enclosure C. Major highlights of the report are
presented below.
EPA's white paper presents a proposed new Pb-TSP FRM consisting of two extraction
methods, ultrasonic bath with a nitric acid/hydrochloric acid solution and hot block reflux with
nitric acid, followed by analysis using inductively coupled plasma-mass spectrometry (1CP-MS).
-------
CASAC finds that the two extraction methods as well as ICP-MS are appropriate. However,
CASAC recommends a strategy that provides flexibility to use additional, applicable extraction
methods (e.g. microwave digestion) and analytical measurement methods (e.g. Graphite Furnace
Atomic Absorption Spectroscopy), including those used as parts of the current FRM. As
previously recommended, EPA should consider developing performance-based FRMs.
EPA's white paper also presents an approach for intra-laboratory and inter-laboratory
validation and testing. CASAC finds that the criteria set forth for intra-laboratory validation
needs to be strengthened. CASAC views the comprehensiveness of the planned inter-laboratory
testing as at an appropriate level, but recommends that EPA consider separating the extraction
methods from the analytical methods so that any of the acceptable extraction methods can be
used with any of the acceptable analytical measurement methods. CASAC does not recommend
a specific level of acceptable inter-laboratory variability, but recommends identifying and taking
into consideration the data quality objectives of the overall measurement.
Although EPA's white paper is specific for Pb-TSP as the indicator, CASAC reiterates
prior recommendations that EPA consider moving away from using Pb-TSP as the indicator and
consider using lead in PM]0 (Pb-PM|0) as the indicator. Measuring Pb-TSP involves the use of
high-volume TSP sampling devices, whereas measuring Pb-PMio involves the use of low-
volume PMio sampling devices. Low-volume Pb-PMio sampling has several advantages over
high-volume Pb-TSP sampling, including the widespread use of such sampling and the ability to
measure multiple compounds in addition to lead. Field studies should be conducted with an
appropriate experimental design to assess low-volume Pb-PMio sampling, making comparison to
high-volume Pb-TSP sampling, to develop data for considering Pb-PMioas the indicator in
future Pb NAAQS reviews. If the Agency determines that Pb-PMio is not appropriate as an
indicator, CASAC encourages the development of a more robust particle sampler for TSP, such
as low-volume TSP sampling that would allow the use of alternative filter substrates to those
currently used in high-volume TSP samplers.
CASAC appreciates the opportunity to provide advice on this issue and looks forward to
receiving the Agency's response.
Sincerely,
/Signed/
Dr. Armistead (Ted) Russell, Chair
CASAC AAMM Subcommittee
/Signed/
Dr. Jonathan M. Samet, 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. 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 - Rosters
U.S. Environmental Protection Agency
Clean Air Scientific Advisory Committee (CASAC)
Ambient Air Monitoring and Methods Subcommittee (AAMMS)
CHAIR
Dr. Armistead (Ted) Russell, Professor, Department of Civil and Environmental Engineering,
Georgia Institute of Technology, Atlanta, GA
MEMBERS
Mr. George A. Allen, Senior Scientist, Northeast States for Coordinated Air Use Management
(NESCAUM), Boston, MA
Dr. Judith Chow, Research Professor, Desert Research Institute, Air Resources Laboratory,
University of Nevada, Reno, NV
Mr. Bart Croes, Chief, Research Division, California Air Resources Board, Sacramento, CA
Dr. Kenneth Demerjian, Professor and Director, Atmospheric Sciences Research Center, State
University of New York, Albany, NY
Dr. Delbert Eatough, Professor of Chemistry, Department of Chemistry and Biochemistry,
Brigham Young University, Provo, UT
Dr. Eric Edgerton*, President, Atmospheric Research & Analysis, Inc., Gary, NC
Mr. Henry (Dirk) Felton, Research Scientist, Division of Air Resources, Bureau of Air Quality
Surveillance, New York State Department of Environmental Conservation, Albany, NY
Dr. Philip Hopke, Bayard D. Clarkson Distinguished Professor, Department of Chemical and
Biomolecular Engineering. Clarkson University, Potsdam, NY
Dr. Rudolf Husar, Professor, Mechanical Engineering, Engineering and Applied Science,
Washington University, St. Louis, MO
Dr. Kazuhiko Ito*, Assistant Professor, Department of Environmental Medicine, School of
Medicine, New York University, Tuxedo, NY
Dr. Donna Kcnski*, Data Analysis Director, Lake Michigan Air Directors Consortium,
Rosemont, IL
Dr. Thomas Lumley*, Associate Professor, Biostatistics, School of Public Health and
Community Medicine, University of Washington, Seattle , WA
-------
Dr. Peter H. McMurry. Professor, Department of Mechanical Engineering, University of
Minnesota, Minneapolis, MN
Mr. Richard L. Poirot, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT
Dr. Kimberly A. Prather*, Professor, Department of Chemistry and Biochemistry, University
of California, San Diego, La Jolla, CA
Dr. Jay Turner, Associate Professor, Environmental & Chemical Engineering, Campus Box
1180 , Washington University , St Louis, MO
Dr. Warren H. White, Research Professor, Crocker Nuclear Laboratory, University of
California - Davis, Davis, CA
Dr. Yousheng Zeng, Air Quality Services Director, Providence Engineering & Environmental
Group LLC, Baton Rouge, LA
Dr. Barbara Zielinska*, Research Professor, Division of Atmospheric Sciences, Desert
Research Institute, Reno, NV
SCIENCE ADVISORY BOARD STAFF
Mr. Aaron Yeow, Designated Federal Officer, U.S. Environmental Protection Agency,
Washington, DC
* Did not participate in the review
-------
U.S. Environmental Protection Agency
Clean Air Scientific Advisory Committee (CASAC)
CHAIR
Dr. Jonathan M. Samet, Professor and Flora L. Thornton Chair, Department of Preventive
Medicine, University of Southern California, Los Angeles, CA
MEMBERS
Mr. George A. Allen, Senior Scientist, Northeast States for Coordinated Air Use Management
(NESCAUM), Boston, MA
Dr. Joseph D. Brain, Cecil K. and Philip Drinker Professor of Environmental Physiology,
Department of Environmental Health, Harvard School of Public Health, Harvard University,
Boston, MA
Dr. H. Christopher Frey, Professor, Department of Civil, Construction and Environmental
Engineering, College of Engineering, North Carolina State University, Raleigh, NC
Dr. Armistead (Ted) Russell, Professor, Department of Civil and Environmental Engineering,
Georgia Institute of Technology, Atlanta, GA
Dr. Helen Suh, Senior Lecturer on Environmental Chemistry and Exposure Assessment,
Department of Environmental Health, School of Public Health, Harvard University, Boston, MA
Dr. Kathleen Weathers, Senior Scientist, Gary Institute of Ecosystem Studies, Millbrook, NY
SCIENCE ADVISORY BOARD STAFF
Dr. Holly Stallworth, Designated Federal Officer, U.S. Environmental Protection Agency,
Washington, DC
-------
Enclosure B
Consensus Responses to Charge Questions on EPA's White Paper, Approach for the
Development of a New Federal Reference Method (FRM)for Lead in Total Suspended
Particulates (Pb-TSP)
1. What are the Panel's views on the two extraction method options of heated ultrasonic and
graphitic hot block with nitric/hydrochloric acid for the extraction of Pb from TSP?
EPA is proposing to replace the current extraction method described in Appendix G of 40
CFR Part 50 with one of the two methods listed above. The major aspects of the three methods
are compared in the following Table. From a chemical extraction point of view, the three
methods should be nearly equivalent for the extraction of Pb from particles collected on either a
glass or quartz fiber filter. As pointed out in the white paper, the hydrochloric acid added to the
Ultrasonic Bath technique will assist in the extraction of more refractory metals from the sample,
but is not needed for the recovery of Pb.
Method
Fraction Analyzed
Solution
Extraction
Current Method
3/4" x 8" strip
15 mL, 3MHNO3
30 min, boiling
Ultrasonic Bath
3/45' x 8" strip
ISmL, 1 MHNO3,
2MHC1
60 min, 80 °C
Hot Block Reflux
%" x 8" strip
20 mL, l:19v/v
HNO3
60 min, 95 °C
Although many versions of the traditional extraction approach using heated ultrasonic
water bath with nitric/hydrochloric acids have been used successfully, they do not represent the
current state-of-art extraction approaches (e.g., microwave digestion). Compendium Method IO-
3.1 (USEPA, 1999a) describes a validated microwave digestion method that EPA originally
developed in 1989. Qureshi, et al. (2006) describe a more contemporary approach to microwave
digestion. A survey of the recent literature should permit an easy update of EPA's Compendium
Method IO-3.1. Thus, it would be straightforward to incorporate microwave digestion in the
intercomparison tests and determine its equivalence to the other proposed methods. It has the
advantage of handling a greater sample throughput with the potential for reduced sample-to-
sample variability, given the features of current generation microwave digestion systems. Thus,
the Subcommittee recommends moving forward with the ultrasonic bath and hot block reflux
methods, but adding microwave digestion as an additional option. Given the variety of potential
methods available which appear to perform well, another approach would be to develop
performance-based standards (quantification of extraction efficiency) to define acceptable
extraction procedures for the FRM.
2. What are the Panel's views on ICP-MS as the analysis met hod for Pb-TSP?
The Subcommittee views that inductively coupled plasma-mass spectrometry (ICP-MS)
is an appropriate analytical method for the Pb-TSP FRM. ICP-MS has superior sensitivity and
specificity for Pb, which is a key criterion for the FRM. Furthermore, the method can generate
data for multiple elements and, depending on the sensitivity of the instrument, can generate data
-------
for the isotopic composition of Pb, which can provide insight into source attribution. Many
state/local agencies are already operating ICP-MS instruments, as demonstrated by the data
reported to EPA for PMio air toxics metals (including Pb) under the National Air Toxics Trends
Stations (NATTS) program.
There are advantages to designating a single analytical method as the FRM, especially if
it has sufficiently high performance that it would remain a suitable method, should revisions be
made to the Pb NAAQS. That said, the Subcommittee recommends a strategy that provides
flexibility, such that agencies can use other methods such as Graphite Furnace Atomic
Absorption Spectrometry (GFAAS), if they are deemed suitable. One approach would be to
designate a performance-based FRM, which would define key performance criteria for minimum
detection limit (MDL), precision, accuracy, linearity, and tolerance for interference. An FRM
based solely on performance, however, has the drawback that each monitoring agency would
have to demonstrate the ability to meet these performance criteria, which could be burdensome.
Alternatively, a single analytical method could be designated as the FRM and performance
criteria could be defined to serve as the basis for Federal Equivalent Method (FEM) designations.
In this case, the Subcommittee strongly recommends that the proposed evaluation of candidate
FRM methods also include analytical methods beyond ICP-MS (including, but perhaps not
limited to, GFAAS) with the intent that methods meeting the performance criteria would be
directly designated as an FEM. In effect, EPA would have completed the FEM process for any
monitoring agencies that would like to use these qualified methods. This approach should be
more cost effective and provide a high level of consistency because all candidate methods are
evaluated by the same labs using the same procedures.
The Subcommittee also recommends that EPA consider:
• Evaluating and possibly designating as an FRM or FEM commonly used ICP-MS
based analytical methods, such as EPA Compendium Method IO-3.5 (USEPA,
1999b) which is already used by many monitoring agencies for NATTS sample
analysis.
• Including low-volume PMio samples collected on Teflon filters with analysis by X-
Ray Fluorescence (XRF) as part of the candidate FRM testing. While this advisory
response focuses on the Pb-TSP method, EPA promulgated an FRM for the
determination of Pb in PMio (CFR, Appendix Q to Part 50) and in the future may
consider the feasibility of low-volume TSP sampling.
• Including the current FRM analytical method (GFAAS) in the evaluation of candidate
FRM methods to assess comparability and help bridge past and future data.
• Including the analysis of other elements in the candidate FRM tests. Since the
monitoring agencies would be able to analyze for multiple elements at a small
incremental cost (at least for methods such as ICP-MS), the documentation of method
performance for multiple elements would assist the agencies in prudent use of such
data for air toxics, studies, source apportionment, etc. However, it should be noted
that the high-volume TSP method (sampler and filter media) are less suitable for
other metals than a PMio Teflon filter sample due to the limitations inherent in the
TSP sampler as well as potential limitations of TSP filter media for trace metal
analysis.
-------
3. What are the Panel's views on the approach described for evaluating and testing the method
prior to proposal as a new FRMfor Pb-TSP?
The overall approach to validation and testing needs to include a comparison to the
existing extraction and analytical methods. A satisfactory comparison to the existing methods
will ensure that the data collected using a new FRM will be consistent with the older Pb database.
This is not likely to be an issue, but it is good practice and should prevent any unexpected
changes in the resultant dataset.
The test matrix, as outlined in the white paper, sets a low bar for validation of the
proposed methods for extraction and analysis of Pb from glass and quartz fiber filters. It also
does not include enough detail on issues such as how the spiked samples will be prepared, which
lab will determine the accuracy and precision of the spiked samples, and whether the sampled
filters will include potential interferences (such as heavy and light loadings of metals, and other
pollutants such as elemental carbon). To address these issues, the test protocol should include a
description of how the test strips are prepared and include an analysis of a subset of the prepared
strips by an acceptable reference laboratory. The selection process for the sampled filters should
be more clearly specified and should include some potential interferents for the extraction and
analytical processes. Archived filters are available and it is likely that state and local monitoring
agencies can assist with the selection of appropriate filters.
EPA has based the development of the Pb analytical FRM on the Office of Solid Waste's
SW-846 Method 6020A (USEPA, 2007). This method does not have adequate Quality
Assurance for analysis of samples used in a compliance oriented air monitoring network. The
following recommendations address some of the criteria that are specified in the test matrix in
the white paper, but need to be strengthened for a Pb FRM.
• The Low Level Calibration Verification (LLCV ) should be ± 10%.
• The Low Level Quality Control Sample (LLQC ) is not really that low since it is 30%
of the NAAQS. The recovery limits should be at minimum ± 15%.
• The recoveries of Standard Reference Material (SRM) spiked on unexposed filter
strips at all concentrations should be ± 10%.
• The 3 non-sequential results should not be averaged in the bias calculation. This will
not provide an indication of single sample bias, which is how actual samples are
analyzed. In general ± 10% bias is too loose, but 5% may be too stringent. A
precision of 10% seems reasonable, because it includes the uncertainty due to
potentially uneven loadings from sampled filters.
• The replicate analysis, as specified in EPA 6020A, only requires one replicate
analysis per batch. The batch size must be specified in the method since new
extractors can accommodate up to 96 samples. One replicate for every 20 samples
should be considered to be a minimum.
The Pb FRM development program should include Teflon filter media, as well as an
accommodation for the analysis of other pertinent elements for air toxics and source attribution
work. This will not conflict with the TSP Pb FRM field sampler since the extraction and
analytical portion of the FRM can be separated from the sample collection method. It may be
-------
appropriate to constrain evaluation of the high-volume TSP sampler filter extraction process to
Pb, since most air agencies would presumably use the PMio FRM with Teflon filters for other
metals analysis. Teflon filters from a PMio FRM are more suitable for trace metal analysis than
the quartz or glass fiber media needed for TSP samplers because of fewer contamination issues.
Teflon filters can also be used for XRF analysis, a cost effective approach for analysis of
multiple elements.
4 Inter-laboratory testing of the method will be done to assess between-laboratory variability
(CV) at the 95% confidence interval. What are the Panel's views on a reasonable level on
inter-laboratory variability?
The Subcommittee is generally comfortable with the comprehensiveness of the planned
inter-laboratory testing. A full array of laboratory and extraction method comparisons will be
performed using spiked and archived sample filter strips. The results will be analyzed to
distinguish the components of variance attributable to intra-laboratory precision and inter-
laboratory bias. The Subcommittee believes that the existing Federal Reference Method for
extraction and analysis should be added to the experimental design to support quantitative
comparisons between new and historical data.
The Subcommittee urges the Agency to consider decoupling the tests of extraction and
analysis, as is recommended by EPA's methods development and validation guidance (USEPA,
1992) cited in the white paper. Extraction and analysis could be distinguished by collecting
different laboratories' extracts for analysis by a single operator on a single instrument, and
sending extracts prepared by a single operator to different laboratories for analysis. The
procedure-resolved results would allow future consideration of alternative method combinations.
On the specific question of "a reasonable level of inter-laboratory variability", the
theoretical implications of target bias and precision values are discussed in the individual
comments by Dr. Warren White. The Subcommittee does not recommend a specific level, but
rather recommends identifying and taking into consideration the data quality objectives of the
overall measurement.
References:
Qureshi, S., Dutkiewicz, V.A., Khan, A.R., Swami, K , Yang, K.X., Husain, L., Schwab, J.J.,
Demerjian, K.L., 2006. Elemental composition of PM2.5 aerosols in Queens, New York:
Solubility and temporal trends. Atmospheric Environment, 40, S238-S251.
USEPA, 1992. Development and Validation of SW-846 Methods Phase 2: Formal Validation. In
Guidance for the Methods Development and Methods Validation for the RCRA Program,
Washington, DC.
10
-------
USEPA. 1999a. Compendium Method IO-3.1, Selection, Preparation and Extraction of
Filter Material. In Compendium of Methods for the Determination of Inorganic Compounds in
Ambient Air, EPA/625/R-96/010a, Cincinnati, OH.
USEPA. 1999b. Compendium Method 1O-3.5, Determination of Metals in Ambient Paniculate Matter
Using Inductively Coupled Plasma/Mass Spectrometry (1CP/MS). In Compendium of Methods for the
Determination of Inorganic Compounds in Ambient Air, EPA/625/R-96/010a, Cincinnati, OH.
USEPA, 2007. Method 6020A, Inductively Coupled Plasma-Mass Spectroscopy. In Test
Methods for Evaluating Solid Waste, Physical/Chemical Methods (SW-846), Washington, DC.
11
-------
Enclosure C
Compendium of Individual CASAC AAMMS Comments on EPA's White Paper, Approach
for the Development of a New Federal Reference Method (FRM)for Lead in Total Suspended
Particulates (Pb-TSP)
Comments from Mr. George A. Allen 13
Comments from Dr. Judith Chow 15
Comments from Mr. BartCroes 19
Comments from Dr. Kenneth Demerjian 21
Comments from Dr. Delbert J. Eatough 22
Comments from Mr. Dirk Felton 26
Comments from Dr. Philip Hopke 29
Comments from Dr. Jay Turner 30
Comments from Dr. Warren White 33
12
-------
Comments from Mr. George A. Allen
General background
This charge to the AAMMS does not include the sample collection method for the Pb
FRM ~ that remains the Hi-Volume sampler with quartz or glass-fiber filters at this time. The
issues discussed here are related to analysis of these filters for Pb and related issues, in the
context of revising the Pb FRM's analytical method to better suit the new and much (10 times)
lower PB NAAQS that is now in effect. As EPA notes, this large drop in the NAAQS requires
revision of the analytical method to insure that data of high quality can be reported at these much
lower concentration loadings on the FRM Hi-Vol filter. These lower loadings require assessment
of both the filter extraction process and the analysis method used for the Pb FRM. Analytical
methods have also changed dramatically since the Pb FRM was last revised; this process allows
the FRM to be updated to take advantage of these newer technologies.
It is important to note that for these components of the Pb FRM, there are likely to be
FEM methods approved that may be more practical for routine use. Still, it is important to have a
robust and well characterized FRM in place even though analysis of Pb from Hi-Vol filters by
any reasonable method with sufficient sensitivity is probably much more robust than the
uncertainties inherent in the existing Pb FRM Hi-Vol sampling method. This process updates the
"back end" of the method; I strongly encourage EPA to continue development of a "larger
particle Lower-Vol" Pb FRM sampler that can provide accurate and reproducible samples to
better harmonize the improvements in the Pb-FRM filter analysis being discussed here.
Charge Question 1. "What are the panel's views on the two extraction method options of heated
ultrasonic and graphitic hot block with nitric/hydrochloric acid for the extraction of Pb from
TSP? "
Either extraction method should meet the needs of a revised Pb FRM, and both of these
methods are currently approved by EPA as FEM methods for Pb extraction off a Hi-Vol filter.
Assuming their performance is similar across the expected lower range of Pb filter loadings, I
would recommend using the simpler method, which in this case appears to be the graphitic hot
block with nitric acid, Method EQL-0710-192. Jt is worth noting here that a graphitic hot block
may not be necessary; aluminum has been shown to work well and is a simpler approach. While
the other method (sonication and both acids, EQL-0510-191) may be needed for extraction of
other metals, it is not needed for Pb, and this is a Pb FRM, not an FRM intended for wider use.
Sonication adds complexity to the method especially if "trace" analysis protocols are used;
quartz fiber filters produces a "filter mush" that requires an extra step to remove the filter
material from the liquid, usually with a centrifuge. Glass fiber filters may also need this
treatment after sonication.
13
-------
Charge Question 2 "What are the panel's views on ICP-MS as the analysis method for Pb-
TSP? "
For the Pb-FRM, ICP-MS is clearly the preferred analysis technique. It is a highly
sensitive and specific method for Pb; there are no other obvious candidate methods.
Charge Question 3 "What are the panel's views on the approach described for evaluating and
testing the method prior to proposal as a new FRMfor Pb-TSP?
The described approach is rigorous. The description of the "archived" field (real
samples?) filters needs to be clarified ~ where are the samples from, what are the loadings, etc.
This use of real samples for assessing extraction, especially at lower Pb loadings, is an important
part of the evaluation since dry loading combined with a large amount of other material may
effect the extraction process. The information in Table 1 (dry NIST SRMs) needs clarification.
How will the dry SRM be applied to the filter? What does the "minimum weight" column mean?
There is an assumption here that quartz filters would be used since they have lower blank
levels for some metals of interest other than lead. But again, this is an FRM for Pb - not other
metals - and glass fiber filter Pb blanks are very low. An agency would presumably use glass
fiber filters for Pb since that media is much less fragile than quartz fiber filters. If other metals
were of interest, I 'd expect that they would be done using a low-volume PM10-FRM Teflon
filter sampler, usually using either XRF and/or ICP-MS. I would not expect agencies to routinely
do other metals off a Hi-Vol filter sample, and thus I would suggest that EPA simplify the Pb
FRM evaluation by constraining tests to glass fiber filters. For (PM10) "trace metals", the Hi-Vol
sampler is not the sampling method or filter media of choice. Trace level analysis requires "trace
level" protocols across the entire method, and a low-vol PM10 FRM Teflon sampling method is
better suited for this.
Charge Question 4 "Inter-laboratory testing of the method will be done to assess between
laboratory variability (CV) at the 95% confidence interval What are the panel's views on a
reasonable level of inter-laboratory variability?
Inter-laboratory method testing is an important component of the FRM assessment, since
it will better reflect the real-world performance of the method ~ which is usually somewhat
degraded from ideal testing conditions. For an acceptable level of inter-lab variability that
includes variability from filter extractions, a starting point for this value might be the recovery
range noted in the white paper: "The SRMs are expected to be recovered within 80 to 120% of
the certified value per method 6020A". If initial test results show a tighter recovery range, then
the inter-lab criteria could be tightened. I would expect much of the inter-lab variability to be
from the extraction process, since a well-controlled ICP-MS analysis for Pb should have
reasonably tight inter-lab variability at these levels.
14
-------
Comments from Dr. Judith Chow
This memo addresses the four questions on which the Subcommittee members were
asked to comment regarding "the White Paper on the Approach for the Development of a New
Federal Reference Method (FRM) for Pb in Total Suspended Particulates (TSP)." This
supplements prior comments to the first set of questions that was appended to the August 12,
2008 letter from Dr. Russell to Administrator Johnson.
Question 1 • What are the panel's views on the two extraction method options of heated
ultrasonic and graphitic hot block with nitric/hydrochloric acid for the extraction of Pb from
TSP?
Both the heated ultrasonic water bath (with 1.02 M nitric acid [HNO3]/2.23 M
hydrochloric acid [HCI]) and graphite hot block (with 3.5 % HNO3 [v/v] at 95 ± 5 °C) are
adequate methods to extract PM deposits on glass-fiber or quartz-fiber filters. It is assumed that
blank filters have been acceptance-tested to assure that Pb blank levels are negligible. However,
the extraction method should be performance-based (e.g., Ashley et al., 1999; Oehme and Lund,
1979) and not limited to the two most commonly used extraction methods defined in the white
paper. For example, Moreira et al. (2005) showed that the microwave and the heating block
digestion methods performed equally well for lead.
The procedure for the heated ultrasonic HNO3/HCI method is more cumbersome and
requires a centrifuge to complete the extraction. This method adds 15.0 ± 0.15 ml of the
HNOs/HCl solvent to the filter aliquot, followed by 1 hr of ultrasonic extraction at 80 ± 5 °C in a
loosely capped 50 ml extraction tube. After cooling, 25.0 ± 0.25 ml of distilled deionized water
(DDW) are added to bring the volume to 40 ± 0.4 ml in a tightly capped extraction tube. This
tube is then centrifuged for 20 minutes at 2500 RPM. Note that the water level in the ultrasonic
bath also needs to be set above the level of the extraction solution in the tubes but below the
level of the extraction tube caps to minimize contamination. Some of the solvent volume may be
lost during the ultrasonic heating stage, because the caps are left loose. The use of HCI in the
extract adds another potential source of contamination and produces a more complex analytical
matrix; HNO3 by itself is an effective Pb solvent (U.S.EPA, 2007).
The hot block digestion method is simple and straightforward. It involves transferring a
20 ml aliquot of diluted HNO3 (1:19 v/v) from concentrated HNO3 (67 - 70%) to each extraction
vessel, ensuring that the filter strip is covered with HNO3 extract, placing the vessel on the hot
block, covering it with ribbed watch glass or a cap with a central vent, and heating it at 95 ± 5 °C
for 60 min. The extract is to be brought to a final volume of 50 ml by dilution wrth reagent grade
DDW. Given the simpler matrix and reduced sample handling of the hot block digestion method,
one could reasonably expect to achieve more consistent results (i.e., better precision). There is a
possibility that deviations from procedures could produce insoluble lead chloride.
15
-------
Question 2: What are the panel's views on ICP-MS as the analysis method for Pb-TSP7
As stated in the white paper, Pb can be analyzed by ICP-MS, just as it can be analyzed by
atomic absorption spectrometry (AAS) with a graphite furnace and ICP-atomic emission
spectroscopy (ICP-AES) once it is in solution. It can be accurately quantified to several orders of
magnitude below the necessary working range. The ICP-MS should support a method detection
limit (MDL) of 0.0075 ng/m3. The instrument will be well within its capabilities, resulting in
accurate and precise results.
The analysis method for Pb should be performance-based. As long as the MDL,
precision (±15% at 90% confidence interval), and accuracy (±5%) are within the EPA's specified
levels, other analytical methods (e.g., AAS with graphite furnace, ICP-AES, X-ray fluorescence
[XRF], or proton-induced X-ray emission [PIXE]) should be available as options (e.g., Harper et
al., 2004; Margui et al., 2005). Unless the state or local agency is already set up for ICP-MS, it
will be costly to initially identify, procure, and set-up for acid digestion (~$5,000 - 10,000; e.g.,
hot block, flow hood) and ICP-MS (-$200,000). An advantage of ICP-MS is the ability to
quantify different isotopic abundances for Pb, which may be related to their sources (Moreira et
al., 2005; Oehme and Lund, 1979; Huggins, 2002; Herner et al., 2006).
Question 3 • What are the panel's views on the approach described for evaluating and testing the
method prior to proposal as a new FRMfor Pb-TSP?
The white paper states the proposed approach for determining method performance (i.e.,
analysis of standard reference materials [SRMs], verification of filters spiked with a known
concentration of NIST traceable Pb salts solution, examination of interference and filter matrix
effects, determination of MDL, and intra-laboratory method performance [e.g., assess bias within
±10% and precision within ±15%, evaluation of glass and quartz-fiber filter matrices, analysis of
spiked filter strips, SRMs, and real-world samples, tests for small variations in extraction
temperature and time, and evaluation of extract storage stability]).
Additional tests need to be planned and conducted to ensure 100% extraction efficiency.
Extraction of spiked-samples with MIST traceable Pb salt or SRMs does not necessarily verify
the extent of a complex matrix of ambient or source samples.
Question 4~ Inter-laboratory testing of the method will be done to assess between-laboratory
variability (CV) at the 95% confidence interval What are the panel's views on a reasonable
level of inter-laboratory variability7
Inter-laboratory testing as listed in the white paper (e.g., method performance assessment
of the four participating laboratories, analysis of spiked filter samples and actual filter samples)
is adequate. It would be worthwhile to examine previously published composition for Pb
analyses. It would also be helpful to specify the concentration range representing the measured
and calibrated concentration range. The total number of samples to be performed for each type of
16
-------
analysis should be more than 10 pairs by each participating laboratory to obtain a statistically
significant analysis.
In addition to the calculation of coefficient of variation (CV) at the 95% confidence
interval, equivalence and comparability should be established for the inter-laboratory comparison
as defined in Mathai et al. (1990). The U.S. EPA criteria for equivalence between Federal
Equivalent Methods (FEM) and FRM for PM2 5 mass concentration, such as: 1) linear regression
slope of 1 ± 0.05; 2) linear regression intercept of 0 ± 1 ug/m3; and 3) linear regression
correlation coefficient (r) of >0.97 can be modified and adapted for "equivalence." The
following criteria can be considered for "comparability" within the stated precision: 1) the slope
(by either ordinary least squares [OLS] or effective variance [EF] weighting) equals unity within
three standard errors, or average ratios (y/x) equal unity within one standard deviation; 2) the
intercept does not significantly differ from zero within three standard errors; and 3) the
correlation coefficient exceeds 0.9 (Berkson, 1950; Kendall, 1951; Madansky, 1959). This is a
less demanding definition than equivalence because it considers the reported precisions of the
two measurements being compared; these may be larger than the requirements for an FEM used
to determine compliance, but still sufficient to discern concentration differences (Watson and
Chow, 2002).
References
Ashley, K.; Song, R.G.; Esche, C.A.; Schlecht, P.C.; Baron, P.A.; Wise, TJ. (1999). Ultrasonic
extraction and portable anodic stripping voltammetric measurement of lead in paint, dust wipes,
soil, and air: An interlaboratory evaluation. J. Environ Monit, 1(5): 459-464.
Berkson, J. (1950). Are there two regressions? J. Am Stat Assoc., 45: 164-180.
Harper, M.; Hallmark, T.S.; Andrew, M.E.; Bird, A.J. (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.
Hemer, J.D.; Green, P.O.; Kleeman, M.J. (2006). Measuring the trace elemental composition of
size-resolved airborne particles. Environ. Sci Technol, 40(6): 1925-1933.
Huggins, F.E. (2002). Overview of analytical methods for inorganic constituents in coal.
International Journal of Coal Geology, 50(1-4): 169-214.
Kendall, M.G. (1951). Regressions, structure and functional relationship, Part I. Biometrika, 39:
96-108.
Madansky, A. (1959). The fitting of straight lines when both variables are subject to error. J Am
Slat Assoc , 54: 173-205.
Margui, E.; Queralt, I.; Carvalho, M.L.; Hidalgo, M. (2005). Comparison of EDXRF and ICP-
OES after microwave digestion for element determination in plant specimens from an abandoned
mining area. Anal Chim Ada , 549(1-2): 197-204.
17
-------
Mathai, C.V.; Watson, J.G.; Rogers, C.F.: Chow, J.C.: Tombach, I.H.; Zwicker, J.O.; Cahill,
T.A.: Feeney, P.J.; Eldred, R.A.; Pitchford, M.L.: Mueller, P.K. (1990). Intercomparison of
ambient aerosol samplers used in western visibility and air quality studies. Enivron. Sci Technol,
24(7): 1090-1099.
Moreira, F.R.; Borges. R.M.; 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.
Oehme, W.; Lund, W. (1979). Comparison of digestion procedures for the determination of
heavy metals (cadmium, copper, lead) in blood by anodic stripping voltammetry. Fresenius Z
Anal Chem., 298: 260-268.
U.S.EPA (2007). 40 CFR Ch. I; Appendix G to Part 50 - Reference Method for the
Determination of Lead in Suspended Paniculate Matter Collected from Ambient Air. Federal
Register, 50: 57-61.
Watson, J.G.; Chow, J.C. (2002). Comparison and evaluation of in-situ and filter carbon
measurements at the Fresno Supersite. J Geophys. Res , 107(D21): ICC 3-1-ICC 3-15. doi:
10.1029/2001JD000573.
-------
Comments from Mr. Bart Croes
Staff should be commended for taking a systematic approach towards implementation of
the revised National Ambient Air Quality Standard (NAAQS) for lead (Pb) in total suspended
participates (TSP). I appreciate the opportunity to comment during several stages of the process,
and agree with the general approach taken by U.S. EPA toward extraction and analysis as
described in the white paper. My comments reflect input from California Air Resources Board
(ARE) staff responsible for implementing U.S. EPA monitoring requirements and using the data
in source apportionment and health studies.
Charge Questions:
1 What are the panel's views on the two extraction method options of heated ultrasonic and
graphitic hot block with nitric/hydrochloric acid for the extraction of Pb from TSP7
Both methods are acceptable for extraction of Pb from TSP for NAAQS compliance
purposes. Since SLT agencies are already extracting Pb as part of source apportionment and air
toxics monitoring networks, and want to conserve resources, this flexibility is desirable. Pb is
easy to extract, but it would be desirable to have performance standards for extraction efficiency
in order to demonstrate equivalency for data from other existing networks.
2. What are the panel's views on ICP-MS as the analysis method for Pb-TSP?
ICP-MS is preferred as the FRM, but other analytical methods besides should be allowed.
The ARB air toxics network (Xontech 924, low volume sampler TSP, Teflon filter) uses ICP-MS
for toxic metals, including Pb, because of high accuracy and precision across many species (in a
single scan) over the entire ambient concentration range.
The Flame Atomic Absorption Spectroscopy (FAAS) method is used for the existing high
volume sampler TSP-Pb network by many SLT agencies, and also has good precision and
accuracy. The FAAS method should be retained as an FEM. ARB's high volume sampler TSP-
Pb network uses Graphite Furnace Atomic Absorption Spectroscopy (GFAAS) and has a
1.0 ng/m3 (0.0010 ug/m3) limit of detection (LOD), far below the new NAAQS of 0.15
ug/m . A cursory look at the last few months of data from two FRM sites (Calexico and
Tijuana) showed all results to be well above this LOD. The GFAAS method should be retained
as an FEM.
3. What are the panel's views on the approach described for evaluating and testing the method
prior to proposal as a new FRM for Pb-TSP7
With the inclusion of the A AS methods discussed above, the approach described for
evaluating and testing the proposed analytical method for a Pb-TSP FRM appears adequate,
especially the range of spiked and ambient samples and number of proposed facilities (four) for
the inter-laboratory comparison. The criteria that each laboratory will achieve bias (accuracy)
<10% and precision <15% for the spiked filter strips seems loose and could be tightened.
19
-------
Typically, the ARB laboratory sees <5% bias and <5% precision for metals analyses, but perhaps
<10% for both is a more realistic expectation.
4. Inter-laboratory testing of the method will be done to assess between-laboratory variability
(CV) at the 95% confidence interval. What are the panel's views on a reasonable level of inter-
laboratory variability?
An inter-laboratory bias (accuracy) <10% and precision <15% for the group of
measurements that make up a 3-month average is a reasonable expectation. The inter-laboratory
comparison should be weighted toward the ambient samples and spiked samples that are near the
level of the NAAQS.
20
-------
Comments from Dr. Kenneth Demerjian
Charge question 1. What are the panel's views on the two extraction method options of heated
ultrasonic and graphitic hot block with nitric/hydrochloric acid for the extraction of Pb from
TSP?
The ideal approach would be to develop performance based standards (quantification of
extraction efficiency) from which to quantify acceptable extraction procedures. The design of an
intercomparison study which would consider a wide-range of real-world samples (i.e. spanning
filter loading and particle composition) to evaluate these techniques will result in substantial
additional work for EPA. Although many versions of the traditional extraction approach using
heated ultrasonic water bath with nitric/hydrochloric acids have been used successfully, they do
not represent the current state-of-art (e.g. microwave digestion). If the agency or states view the
Pb measurement approach as an opportunity to consider the measurement of other PM metals the
newer digestion technologies may be of even greater interest.
Charge question 2 What are the panel's views on ICP-MS as the analysis method for Pb-TSP7
The ICP-MS is the state-of-art measurement method of choice.
Charge question 3 What are the panel's views on the approach described for evaluating and
testing the method prior to proposal as a new FRMfor Pb-TSP?
The subject white paper provides an acceptable description and adequate guidance
regarding the development of a new FRM for Pb-TSP and QA/QC requirements.
Charge question 4 Inter-laboratory testing of the method will be done to assess between
laboratory variability (CV) at the 95% confidence interval What are the panel's views on a
reasonable level of inter-laboratory variability'''
The proposed inter-laboratory testing procedures as outlined are adequate to judge the
performance and acceptability of individual laboratories. Differences in laboratory performance
may likely reside from sampling handling and preparation (i.e. extraction). An assessment of the
typical variability in extraction efficiencies may be the determining factor in what should be
viewed as to reasonable level of inter-laboratory variability.
21
-------
Comments from Dr. Delbert J. Eatough
Individual response to Charge Questions for the Clean Air Scientific Advisory
Committee's (CASAC) Ambient Air Monitoring & Methods Subcommittee Peer Review on the
Approach for the Development of a New Federal Reference Method (FRM) for Lead (Pb) in
Total Suspended Particulates (TSP).
Prepared in connection with the AAMMS committee consensus response to EPA in
connection with the September 15, 2010 AAMMS call.
1 What are the panel's views on the two extraction method options of heated ultrasonic and
graphitic hot block -with nitric/hydrochloric acid for the extraction of Pb from TSP7
EPA is proposing to replace the current extraction method described in Appendix G of 40
CFR Part 50 with one of the two methods listed above. The essence of the three methods, as I see
it, are compared in the following Table.
Method
Fraction Analyzed
Solution
Extraction
Current Method
3/4" x 8" strip
15mL,3MH-NO3
30 min, boiling
Ultrasonic Bath
3A" x 8" strip
15mL, 1 MHNO3,
2M HC1
60 min, 80 °C
Hot Block Reflux
%" x 8" strip
20 mL, 1:19 v/v
HN03
60 min, 95 °C
From a chemical extraction point of view, the three methods should be nearly equivalent
for the extraction of Pb from particles collected on either a glass or quartz fiber filter. As pointed
out in the EPA material provided for the consultation, the HC1 added to the Ultrasonic Bath
technique will assist in the extraction of more refractory metals from the sample, but is not
needed for the recovery of Pb. From, a personal point of view, I prefer the Ultrasonic Bath
technique because it minimizes the temperature to which the concentrated nitric acid solution is
heated. This method does, however add and additional step to remove the fibers produced. As
pointed out by others, all of the techniques are a little dated. Using a more current technique,
such as microwave digestion would be preferable. However, the more important question is the
blank, Pb recovery, and precision of each technique. These points cannot be established from the
material provided. The two newer techniques proposed by EPA for evaluation are covered in two
reports referenced in the white paper, e.g. EQL-0510-191,Determination of Lead in TSP by
Inductively Coupled Plasma Mass Spectrometry (ICP-MS) with Heated Ultrasonic Nitric and
Hydrochloric Acid Filter Extraction and EQL-0710-192, Heated Nitric Acid Hot Block
Digestion and ICP-MS Analysis for Lead (Pb) on TSP High-Volume Filters. The first is a report
from RTI to EPA on the indicated method and the second a method description from U.S. EPA
Region 9. Neither are complete scientific reports, giving details on the data behind the various
statements in the documents. The above listed important points of method blank. Pb recovery
and precision of the techniques cannot be established from these reports.
Likewise, there are no studies from any given laboratory which allow comparisons of the
three techniques. EPA is planning a detailed laboratory study as part of the protocol moving
22
-------
forward. This will be a valuable step and should do much to better define important elements
which should be part of the EPA inter-laboratory evaluation and testing program to follow.
Finally, I would like to suggest that all three methods, plus microwave extraction, and not
just the two outlined in the white paper should be evaluated on an equal footing. This will
provide a tie to previous analyses and give an indication as to whether a change from the older
analytical scheme is needed with respect to the extraction method.
2. What are the panel's views on ICP-MS as the analysis method for Pb-TSP?
The analytical techniques used for the determination of Pb in the various extract solutions
to be analyzed, as outlined in charge question 1, is required by EPA to have a MDL for Pb of 5%
of the N AAQS (0.15 u.g/m ), or 0.008 ug/m . As recently summarized by C 3 3 avender (EPA,
OAQPS) in a presentation at the March 2, 2009 National Air Quality Conference in Addison,
TX, expected MDLs for Pb in TSP and PMio methods are:
Estimated MDL: (ug/m3)
0.002 to 0.00005 •*
0.0007 b
0.00006 c
0.0003 c
Sample Analysis Methods
Graphite Furnace Atomic Absorption (GFAA)
Energy-dispersive X-Ray Fluorescence (EDXRF)
Inductively Coupled Plasma Mass Spectrometry (ICP-MS)
Inductively Coupled Plasma Atomic Emission spectrometry
(ICP-AES/OES)
"Estimated MDL retrieved from AQS
b40 CRF Part 50, Appendix Q
'Estimated MDLs reported in the Inorganic (1C) Compendium of Methods for the
Determination of Inorganic Compounds in ambient Air, EPA/625/R-96/01a, July 1999
Each of the methods listed above are capable of given data within the required MDL.
While 1CPMS has the anticipated lowest MDL in the Table, this extra level of lower detection is
not needed for the purposes of the NAAQS for Pb as currently outlined. While evaluating the
ICP-MS method as an alternate method for data analysis would be warranted, I see no reason to
dump the long standing GFAA method for which we have decades of experience. Moving ahead
with both methods in the evaluation proposed by EPA would seem warranted. The comparison
of the two methods will enhance our understanding of both precision and accuracy in the
determination of Pb under the current standard. Since both methods will use the various extract
solutions outlined in charge question 1, there is no practical impediment to this approach.
I would like to make a few final points in regards to the above Table. All of the
techniques, except for EDXRF analysis are ideally suited for analysis of solutions, such as those
produced by the procedures highlighted under charge question 1. However, the procedures
required to produce those solutions are laborious, expensive and prone to sample contamination
introduction. The solution extraction approach is required under the current FRM technique for
Pb in TSP. But making an effort to move any from this requirement would be highly desirable.
23
-------
EPA has put in place the possibilities for both low volume TSP samplers or the acceptance of a
PMio value for Pb measurements in the future in the 40 CFR 50 document. These were both
directions the AAMMS urged EPA to move in the past and 1 encourage EPA to actively
incorporate evaluation of these methods into some phases of the development of FRMs for Pb. If
such a path, is not taken, then comparison data which would allow the incorporation of either a
low volume TSP sampler or a PMio metric in the future will not be obtained and moving one of
these directions will not be possible. Either technique would allow the use of EDXRF (or
equivalent methods) for the direct analysis of collected samples without an extraction step. This
would be highly desirable.
3. What are the panel's views on the approach described for evaluating and testing the method
prior to proposal as a new FRMfor Pb-TSP?
As outlined in the charge questions and in the white paper, the approach which EPA will
take to validate any new methods for analysis of TSP PB is based on the guidance documents
and references provided with the charge questions. Particular weight is given to information
contained in Validation and Peer Review of U.S. Environmental Protection Agency Chemical
Methods of Analysis, prepared for The EPA Forum on Environmental Measurements (FEM);
FEM Document Number 2005-01, October 14, 2005. The approach to be used includes the
analysis of Standard Reference Materials (SRMs) and NIST-traceable Pb salt solutions;
assessment of method detection limits (MDLs), bias, precision; and inter-laboratory testing to
assess between laboratory variability. The approach is reasonable, well thought out and based on
solid guidance to EPA. 1 urge EPA to keep in focus the point that the lowest MDL is not the
prime factor which should be considered in making a decision. As long as any analytical
technique has a MDL within the requirements of the standard, it should be evaluated on an equal
footing with the methods with the lowest MDL. There are three important modifications I would
suggest to the approach outlined in the white paper:
(1). Include in the single laboratory evaluation of the four extraction methods outlined
in 1. above to better inform the process and set the protocols as EPA moves into
multilaboratory evaluations.
(2). Include a minimum of both ICP-MS and GFAA in both the single laboratory
study and multi-laboratory evaluations so that there is a solid basis for either
retaining the GFAA method in the FRM (along with ACP-MS if adopted) or
switching to only the ICP-MS method. I urge this evaluation because of the
legacy in both data and laboratory experience with the current analytical method.
(3). Include comparison measurements with both a low flow TSP sampler and a PMio
sampler to develop a data set which would allow EPA to consider moving away
from the high volume TSP sampling procedure in the future. I particularly urge
EPA to include the PMio measurements and direct analysis of the collected filter
material to allow the Pb program to become more consistent with sampling for the
PM standard in the future.
24
-------
4 Inter-laboratory testing of the method will be done to assess between-laboratory variability
(CV) at the 95% confidence interval. What are the panel's views on a reasonable level of inter-
laboratory variability?
I concur with the points in the consensus report.
25
-------
Comments from Mr. Dirk Felton
General Comment
The FRM for TSP Pb suffers from relatively high measurement uncertainties at the
higher concentrations near the NAAQS. This error is apparent from the field precision data
which includes variability due to sampler collection efficiency. The high volume TSP samplers
specified for use in the NAAQS program have variable collection efficiency due to wind
direction and wind speed. These sampler related collection errors are proportionally larger as the
concentration of Pb in the air increases. The graph below shows that when the ambient Pb
concentration nears the new NAAQS, the difference in concentration between the primary and
duplicate samplers increases. This increase in error is not related to extraction efficiencies which
are summarized in the next plot. Duplicate extractions are performed by extracting and
analyzing a second strip from the same filter. The 2009 Pb QA data provided by the NYSDEC
indicates that the current extraction and analysis methods are adequate but the field collection of
Pb samples needs improvement.
m
E
~E
I
01
*gt
ft
y
"E
0
NYSDEC TSP Pb Coilo Data: ICP-AES
01 ft
.14 "•
.11)
OrtQ .
Oftf. .
0.02 -
1 <-""»
. ^
\s^ *
^^^'
^^+ V= 0.94x4 0.00
^^*^ RJ = n QQ
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
Primary Sampler ug/rr>3
26
-------
Q.
X-
+ll
M
fc.
0*
**
ugPb/2nd fi
NYSDEC TSP Duplicate Extractions 2009
PC
or
15 -
1 fi -
^+
^^
^**^
^^
j^^ y = 1.02x-0.06
^^•*^ R2-1.00
_*-•''
0 5 10 15 20 25 30 35
ugPb/lst filter strip
ICPMS is used by most monitoring agencies in the NATTS program for the analysis of
low-volume PM-10 filters for toxics metals including Pb. The data for this program is used for
trends determinations and risk assessments so data accuracy is important throughout the
concentration range. The accuracy and sensitivity of the analysis must be much better than for
the Pb high volume analysis because of the smaller sampling volume for these samplers. The
results from the collocated non-source oriented NATTS site in New York demonstrate that the
low volume field samplers with ICPMS analysis produce very high quality data.
NYSDEC PM-10 Collo Pb Data: ICP-MS
I
.51
a.
i
0.02
0.015
0.01
.3 0.005
Q.
D
= l,00x- 0.00
R2 = 0.99
0.005
o.oi
Primary Sampler ug/m3
0.015
0.02
27
-------
Response to Charge Questions
1. What are the panel's views on the two extraction method options of heated ultrasonic and
graphitic hot block with nitric/hydrochloric acid for the extraction of Pb from TSP?
Either extraction method is acceptable for NAAQS Pb monitoring. Monitoring Agencies
are likely to prefer one method over another because it is either cost effective or because it is
better suited for the analyses of other elements.
Since Pb is relatively easy to extract, it would be preferable to have a range of acceptable
extraction parameters such as acid types and concentrations, temperatures, heating methods and
filtration included in the FRM. Performance standards for extraction efficiency could be
included to ensure that the method selected by specific laboratories is adequate for compliance
data. This approach would not only make the samples collected for Pb NAAQS monitoring
useful for other programs but will also make the samples collected for other programs such as air
toxics eligible for use in comparison to the Pb NAAQS.
2 What are the panel's views on ICP-MS as the analysis method for Pb-TSP?
The method is acceptable but is more accurate than is necessary for high volume TSP
filter analysis. The advantage of making this method a FRJvl is that it is also suitable for use with
low volume PM-10 filters. The FAAS method is adequate for high volume sample analysis and
is performed in-house by many State and Local monitoring agencies. The FAAS method should
be retained as an FEM but limited to use for high volume samples.
3 What are the panel's views on (he approach described for evaluating and testing the method
prior to proposal as a new FRM for Pb-TSP?
The approach looks adequate though some of the evaluation criteria are not very tight.
Recoveries, inter-method precisions and bias at high concentrations should be better than the
criteria set forth in the white paper. The criteria should be tightest at the level of the NAAQS.
The sampled filters should also be selected to look for real world interferences. The
white paper mentions that filters will be selected to cover a range of Pb but they should also
cover a range of other factors such as heavy crustal or EC loadings that may potentially interfere
with the analysis results.
4 Inter-laboratory testing of the method will be done to assess between-laboratory variability
(CV) at the 95% confidence interval What are the panel's views on a reasonable level ofinter-
laboralory variability?
The comparisons between labs should include criteria that indicate performance at low
concentrations, high concentrations and of course at the level of the NAAQS. The evaluation
should also emphasize the performance for real world samples.
28
-------
Comments from Dr. Philip Hopke
1. What are the panel's views on the two extraction method options of heated ultrasonic and
graphitic hot block with nitric/hydrochloric acid for the extraction of Pb from TSP?
The methods proposed appear to be adequate for extracting lead from TSP samples, but
they have failed to evaluate modern extraction techniques. There is no discussion of microwave
digestion where you can automate multiple samples. There are systems that can run up to 40
samples at a time with monitoring of the temperature of each extraction cell to ensure uniform
extraction. Thus, the Agency has missed an opportunity to permit efficient, uniform extraction
that would make the analyses faster, more uniform and simpler to run in large numbers.
2 What are the panel's views on ICP-MS as the analysis method for Pb-TSP?
TCP-MS is an appropriate method.
29
-------
Comments from Dr. Peter McMurry
1 What are the panel's views on the two extraction method options of heated ultrasonic and
graphitic hot block with nitric/hydrochloric acid for the extraction of Pb from TSP?
I am not an expert on methods used for extracting Lead (Pb), so I will not to comment on
them.
2 What are the panel's views on ICP-MS as the analysis method for Pb-TSP?
Again, I have no particular expertise on analytical methods used to determine Lead (Pb)
concentrations, so 1 will not comment on the choice of ICP-MS as the analysis method for the
FR.M. My only concern about this is that if state or local agencies use another method that
produces equivalent results, then perhaps it should also be allowed.
3 What are the panel's views on the approach described for evaluating and testing the method
prior to proposal as a new FRMfor Pb-TSP?
My primary concern is that the standard involves the analysis of samples collected with a
TSP sampler. There must be a reason that TSP samplers are used rather than, for example, PM10.
Are TSP samplers also used for other types of sampling? It would seem that mandating TSP as
well as PM10 sampling would impose unnecessary burdens on state and local agencies. Also,
TSP samples are less well characterized than PM10 samples.
Other AAMM members have had more experience than I at method evaluation. 1 defer to
their judgement on this topic.
4. Inter-laboratory testing of the method will be done to assess between-laboratory variability
(CV) at the 95% confidence interval. What are the panel's views on a reasonable level of inter-
laboratory variability7
It would appear that the interlaboratory testing plan is based on well-established
paradigms.
What procedures will be employed to ensure that a given laboratory continues to analyze
samples accurately over time, after the initial intercomparison is complete?
30
-------
Comments from Dr. Jay Turner
It is appropriate that the Agency revise the FRM for Pb-TSP to provide for suitable
sample analyses to determine compliance with the Pb NAAQS revisions promulgated in
November 2008. Advice has been solicited concerning the extraction and analytical methods.
method validation and testing, and target for inter-laboratory variability.
I commend the EPA for the thoughtful approach that is being taken to the revisions,
including the proposed methods evaluation study.
Charge Questions and Responses
1. What are the panel's views on the two extraction method options of heated ultrasonic and
graphitic hot block with nitric/hydrochloric acid for the extraction of Pb from TSP?
The two methods (EQL-0510-191 and EQL-710-192) are reasonable candidate methods.
They are consistent with current laboratory equipment and practices. Our group's recent
experience with graphite hot block sample digestions for air toxics PM10 metals (including but
not limited to Pb) has been very positive.
I am comfortable with more than one extraction method being included in the FRM. My
key concern is that clear equivalency criteria be developed so that additional, commonly-used
extraction protocols can designated as FEM methods. It would be beneficial if EPA included
certain additional extraction methods in the proposed evaluation study and made the FEM
designations, rather than placing this burden on state/local agencies. For example, Pb is a target
analyte for the National Air Toxics Trends Station (NATTS) network PM10 metals
measurements. Given that several state/local agencies are already performing these analyses, it
might be advantageous if the same extraction method could be used for the Pb-TSP analysis.
On a technical note, the extraction description in EQL-710-192 concludes with the step
"Shake the extract vigorously for 5 seconds (with the filter strip in the extraction vessel) and let
settle for at least an hour. The sample is now ready for analysis." What are the provisions for
storing the sample(s) for some length of time between extraction and analysis? Also, does the
one hour of settling time guarantee that no filtration of the extract is required? A filtration step is
included in EPA Compendium Method IO-3.1 (Selection, Preparation, and Extraction of Filter
Material), the metals extraction SOP posted under the air toxics section of the EPA AMTIC web
site (http://www.epa.gov/ttnamti 1 /files/ambient/airtox/metalsop.pdf). and the Technical
Assistance Document for the National Air Toxics Trends Station Program (Revision 2, April
2009; http://www.epa gov/ttn/amtic/files/ambient/ inorganic/mthd-3-5.pdf).
What are the plans for the initial demonstration of performance for each analytical
laboratory? While the candidate methods will be evaluated using the Text Matrix appended to
the White Paper, would there be advantages to having each analytical laboratory perform tests
31
-------
with SRM or some other standardized samples as part of an initial demonstration of
performance? This is not addressed in the candidate SOPs.
2 What are the panel's views on ICP-MS as the analysis method for Pb-TSP?
ICP-MS is appropriate as the reference method. It has suitable sensitivity and minimal
issues with interferences. In addition to contract laboratories, many state/local agencies are
already performing ICP-MS analysis for air toxics PM10 metals (see e.g. Table 8 in National Air
Toxics Trends Stations Quality assurance Annual Report, Calendar Year 2007;
http://www.epa.gov/ttnamtil/files/ambient/airtox/NATTS2007OAAnnualReport.pdf). Again,
there might be advantages to harmonization with the NATTS method which is EPA
Compendium Method IO-3.5 (Determination of Metals in Ambient Paniculate Matter Using
Inductively Couple Plasma/Mass Spectrometry). The EPA should consider including GFAAS in
the proposed evaluation study, possibly leading to designation as an FEM directly by the Agency.
Also, FAAS should be included in the evaluation study to provide connection to historical
analysis methods.
I encourage the designation of a single analytical method as the FRM, specifically a
method that has sufficiently high sensitivity to be suitable should the Pb NAAQS concentration
threshold be revised downward at a later time. In this case, any additional qualifying methods
for the current standard could be designated as FEM.
3 What are the panel's views on the approach described for evaluating and testing the method
prior to proposal as a new FRM for Pb-TSP?
The approach is sound although the test criteria ideally would be framed within the
context of a DQO, but there is no evidence that this has been done. Tests and criteria have been
defined in the Test Matrix appended to the White Paper. I have some concerns about the broad
range for the SRM recoveries (80-120%). While this recovery range is a commonly used
criterion, should it be tighter for NAAQS compliance measurements? Also, should recovery
tests be performed at even higher nominal mass loadings? The NAAQS is based on a three-
month rolling average. For 1 -in-6 day sampling this is nominally 15 samples and thus a few very
high concentration samples could lead to a violation and it would be important to have accurate
measurements at high concentrations since the NAAQS is an arithmetic average.
4 Inter-laboratory testing of the method will be done to assess between-laboratory variability
(CV) at the 95% confidence interval What are the panel's views on a reasonable level of inter-
laboratory variability7
The target performance for inter-laboratory variability should take into consideration the
overall measurement quality objectives and the performance achieved by the method(s) from the
intra-laboratory method performance evaluation.
32
-------
Comments from Dr. Warren White
As usual, 1 am impressed by the care and consideration the Agency gives to defining its
compliance measurements. I would also like to commend the authors of this white paper for
their inclusion of hyper-links to relevant Agency documents (on pages 1 and 2), which can
sometimes be hard for outsiders to locate.
I bring no specific personal expertise to this review, having no direct experience with
either high-volume (TSP) sampling or ICP-MS analysis. My comments accordingly focus on the
statistical aspects of the inter-laboratory testing, responding to charge question #4.
The experimental design for inter-laboratory testing is said to follow, generally, that of
Long et al. (1979). For tests on sampled filters, it anticipates selecting 28 = 7x2x2 archived
high-volume filters. This will provide 7 ambient levels on each of 2 filter media (glass and
quartz), for extraction by 2 different procedures (heated ultrasonic and heated block). Each high-
volume filter will be sectioned into eight strips; four pairs of adjacent strips will be distributed to
four different laboratories. Each laboratory will analyze each extract in triplicate and report the
individual results. The resulting data set will support the analysis of variance techniques with
which Long et al. resolved observed variations into components attributable to within-laboratory
and between-laboratory standard deviations.
In the Long et al. inter-laboratory evaluation, the role of the paired strips was to provide
matched tests of the alternative extraction procedures: "Each laboratory was requested to extract
one strip of each pair with the boiling HNO3 procedure and the other with the ultrasonic HNO3
procedure." This matching provided more statistical power for testing the equivalence of the two
extraction procedures. The present white paper instead envisions that "Each laboratory will
receive 7 pairs for the heated block method and 7 pairs for the heated ultrasonic method." This
means that the two extractions will be tested on unmatched samples, making their equivalence
more difficult to assess. The experimental design could be improved by splitting each pair
between the two extraction procedures, as was done by Long et al. This return to the earlier
design would also reduce the required number of archived filters (and lab analyses) by one-half,
to 14.
If more than four laboratories or two extraction methods might be compared, this would
require some reconfiguration of the present design. It is also worth noting that the inter-
laboratory discussion (pp. 27-28) in the EPA methods development and validation guidance
(USEPA, 1992), cited by the white paper, explicitly calls for decoupling the tests of the
extraction and determinative procedures:
When validating a sample preparation method, the participating laboratories should only
perform the sample preparation procedure The collected samples should then be sent to
one laboratory for analysis The analysis should be done by a single operator on a single
instrument in a single batch to minimize analytical variability inherent to the
determinative method Conversely, if a determinative method is to be validated, the
developer should have a single operator perform all of the sample preparation
33
-------
operations in order to minimize operator and laboratory variability inherent to the
sample preparative procedures. The sample extracts should then be split and sent to the
laboratories participating in the validation study for the analytical determination
On the charge question of "a reasonable level of inter-laboratory variability", we can note
that Long et al. judged that two observations by different laboratories using the previous FRM
would not differ by more than 28% of the concentration level more than 5% of the time due to
chance alone. Alternatively, if 10% bias and 15% precision are tolerated for each lab, then
differences of 50% = 2.77 x (10%2+15%2)1/2 could be expected 5% of the time. [The factor 2.77
is the two-sided 5% critical value (1.96) for the normal distribution, multiplied by the square root
of two to account for the independent uncertainties in two observations.] For comparisons
between n-sample averages, differences of 2.77 x (10%2+15%2/n)l/2 could be expected 5% of the
time. Averaging 16 (three months of one-day-in-six sampling) observations, for example, would
thus reduce the 5% confidence level difference from 50% to 30%.
Whether these are "reasonable" levels of variability requires a consideration of data
quality objectives (USEPA, 2006). Given the epistemic uncertainties in our exposure and risk
assessments for lead, an acceptable level of uncertainty for the indicator cannot be derived from
scientific principles alone.
USEPA, 1992. Development and Validation of SW-846 Methods Phase 2: Formal Validation. In
Guidance for the Methods Development and Methods Validation for the RCRA Program,
Washington, DC.
USEPA, 2006. Guidance on Systematic Planning Using the Data Quality Objectives Process.
EPA/240/B-06/001, Washington, DC. Available at: http://www.epa.gov/qualitv/qs-docs/g4-
final.pdf
34
-------
Comments from Dr. Yousheng Zheng
Charge Question 1 What are the panel's views on the two extraction method options of heated
ultrasonic and graphitic hot block with nitric/hydrochloric acid for the extraction of Pb from
TSP?
EPA has presented two options for lead extraction from TSP filters - Option 1: heated
ultrasonic nitric and hydrochloric acid filter extraction, and Option 2: heated nitric acid hot block
digestion. I don't have experience with these two specific extraction procedures. However, based
on my past experience with acid digestion of TSP samples, the mixture of nitric acid and
hydrochloric acid provides more aggressive and thorough extraction of metals than nitric acid
alone. More aggressive extraction may or may not be important for lead. It certainly makes
difference for extracting some other metals. With ICP-MS, multiple metal analyses will be easier
than the current FRM using FAAS. Multiple metal analyses will provide more useful information
for source identification and apportionment. With the ease and value of having multiple metals
analyses, it is anticipated that more agencies would analyze multiple metals beyond lead.
Therefore, a procedure that has stronger digestion and extraction power for not only lead, but
also other metals is beneficial.
In addition to hydrochloric acid, Option 1 uses a slightly higher nitric acid concentration
than Option 2 (64.4 mL of concentrated nitric acid in 1000 mL vs. 50 mL in 1000 mL). Option 1
also uses ultrasonic waves to provide agitation and to aid extraction.
Overall, Option 1 seems to be a stronger method than Option 2. Unless Option 1 is more
susceptible to interference due to presence of hydrochloric acid or chlorides, 1 would favor
Option 1. As part of EPA's further evaluation of the two candidate methods before final selection
of the FRM, 1 would suggest that EPA address vulnerability between the two methods in terms
of potential interference in the two methods, i.e., does the presence of hydrochloric acid make
the subsequent ICP-MS analysis more susceptible to interference in the context of a typical
ambient TSP samples as the matrix? As stated in Section 7.2 of SW-846 Method 6020A, "many
more molecular-ion interferences are observed when hydrochloric and sulfuric acids are used".
This question may have been investigated and answered, but not included in the materials
provided for this subcommittee review.
Charge Question 2: What are the panel's views on ICP-MS as the analysis method for Pb-TSP?
EPA has cited the advantages of using ICP-MS over the current FRM analytic method
(FAAS). I agree that ICP-MS should be a significantly better method than FAAS. My only
comments are:
1. Potential interferences (including isobaric elemental interferences, physical and
chemical interferences, and interferences that may be caused by introduction of
hydrochloric acid as discussed above;) should be addressed as part of EPA's
evaluation.
35
-------
2. If possible, EPA should consider some secondary objectives to leverage and
maximize the benefits of conducting the evaluation. These secondary evaluations may
include:
a. Analyzing other elements in addition to Pb. If the results are favorable, they
can be used to support monitoring agencies' effort to analyze multiple
elements at a small incremental cost and acquire valuable data for other
programs such as air toxics studies, source apportionment, etc.
b. Analyzing samples using the current FRM (i.e., FAAS) and XRF method in
parallel to the evaluation of the candidate ICP-MS methods so that
comparisons can be made. A comparison with FAAS may help understand the
past monitoring data and bridge the past and future data. A comparison with
XRF may be useful because a large body of data has been and can be
generated by XRF.
3. The candidate FRM methods are designed for analyzing TSP filters collected by high
volume samplers. With a large number of PM10 samplers in operations, it would be
desirable to evaluate suitability of applying the candidate ICP-MS methods to the
PM10 filters collected by more widely used PM10 samplers. If the new ICP-MS
based FRM is also suitable for analyzing the PM10 filters collected by common
PM10 samplers, it will make it easier for monitoring agencies to analyze these filters
for Pb and other metals, not necessarily for lead NAAQS attainment determination
but for other air programs.
Charge Question 3. What are the panel's views on the approach described for evaluating and
testing the method prior to proposal as a new FRM for Pb-TSP?
The proposed approach seems to be comprehensive. I don't have specific comments at
this time.
Charge Question 4. Inter-laboratory testing of the method will be done to assess between-
laboratory variability (CV) at the 95% confidence interval What are the panel's views on a
reasonable level of inter-laboratory variability?
The proposed inter-laboratory variability level seems to be reasonable. I don't have
specific comments at this time.
36
------- |