&EPA
United States
Environmental Protection
Agency
PB2009-102750
Sampling and Analysis of
Asbestos Fibers on Filter
Media to Support Exposure
Assessments: Bench-Scale
Testing

-------

-------
EPA 600/R-08/046
November 2008
www.epa.gov
Sampling and Analysis of
Asbestos Fibers on Filter
Media to Support Exposure
Assessments: Bench-Scale
Testing
Daniel A. Vallero, Ph.D.
U.S. Environmental Protection Agency
National Exposure Research Laboratory
Research Triangle Park, NC 27711
John R. Kominsky
Environmental Quality Management, Inc.
Cincinnati, OH 45240
Michael E. Beard and Owen Crankshaw
RTI International
Research Triangle Park, NC 27709
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460

-------
April 6, 2008
(Revised November 5, 2008)

-------
NTIS DISCLAIMER
This document has been reproduced from the best
copy furnished by the sponsoring agency.

-------
ABSTRACT
Sampling efficiency is essential in exposure assessments of contaminants in air, as well as
other matrices. In the measurement of airborne contaminants, it is critical to collect a sample of air
containing representative contaminants in the air of concern, that is, contaminant concentration and
size distribution in the sampled air must be the same as that of the air of concern. Typically, mixed
cellulose ester (MCE, 0.45 or 0.8 |im pore size) and to a much lesser extent, capillary-pore
polycarbonate (PC, 0.4 |im pore size) membrane filters are used to collect airborne asbestos for
count measurement and fiber size analysis. A literature review did not identify any study that
compared the fiber retention efficiencies of 0.45 |im and 0.8 |im pore size MCE or 0.4 |im pore
size PC membrane filters for asbestos aerosols. In this research study chrysotile asbestos (fibers
both shorter and longer than 5 jam) were generated in an aerosol chamber and sampled by 25-mm
diameter MCE filter media to compare the efficiency of a 0.45 |im pore size filters versus 0,8 p.
pore size filter media. In addition, the effect of plasma etching times on fiber densities was
evaluated. Polycarbonate filters were not tested in this study.
This study demonstrated a significant difference in fiber retention efficiency between 0.45
|im and 0.8 |im pore size MCE filters for chrysotile asbestos aerosols (structures >0.5 |im length; s
= 0.5 |im). That is, the fiber retention efficiency of a 0.45 |im pore size MCE filter is statistically
significantly higher than that of the 0.8 |im pore size MCE filter. However, for chrysotile asbestos
structures > 5|im in length, there is no statistically significant difference between the fiber
retention efficiencies of the 0.45 jam and 0.8 |im pore size MCE filters. The mean density of
chrysotile asbestos fibers (>0.5 |im in length) increased with etching time (2, 4, 8, and 16 minutes).
Regression analysis of etching time and density showed that doubling the etching time adds an
average of 13% to the total chrysotile asbestos density within the density range tested. Plasma
etching time had no effect on the reported fiber densities of fibers longer than 5 |im.
Many asbestos exposure risk models attribute most of the health effects to fibers longer
than 5 |im in length. In these models, both the 0.45 |im and 0.8 |im pore size MCE filter can
produce suitable estimates of the airborne asbestos densities. However, some models suggest a
more significant role for asbestos fibers <5 |im in length. Exposure monitoring for these models
should consider only the 0.45 |im pore size MCE filters as recommended by the U.S. EPA
AHERA protocol and other methods.
11

-------
This report is based upon study findings and information submitted in fulfillment of
Contract No. 68-C-00-186, Task Order No. 0020 by Environmental Quality Management, Inc.
under the sponsorship of the United States Environmental Protection Agency, covering a period
from April 28, 2006 to December 24, 2006, and work was completed as of December 31, 2006.
The information in this document has been funded wholly by the United States Environmental
Protection Agency under Contract No. 68-C-00-186, Task Order No. 0020 to Environmental
Quality Management, Inc. It has been subjected to the Agency's peer and administrative review
and has been approved for publication as an EPA document. Mention of trade names or
commercial products does not constitute endorsement or recommendation by EPA for use.
in

-------
CONTENTS
Section	Page
Abstract	ii
Contents 	iv
Tables	v
Figures	vii
1	Introduction	1
1.1 Proj ect Obj ectives	4
2	Study Design and Methodology	5
2.1	Preparation of Samples for Analysis	5
2.1.1 Dust Generation and Collection System	5
2.1.1.1	FluidizedBed Generator	6
2.1.1.2	Sonic Velocity Disperser and Settling Tower	6
2.1.1.3	Sample Collection Chamber	6
2.1.1.4	Dust Feeder	7
2.1.1.5	Sonic Velocity Disperser and Settling Tower	8
2.1.1.6	Sample Collection Chamber	8
2.2	Sample Analysis Strategy	9
2.2.1	Fiber Retention Efficiency of 0.45 |im and 0.8 |im Pore Size MCE Filters	9
2.2.2	Effect of Plasma Etching Time of Total Chrysotile Asbestos Density	9
2.3	Analyti cal Methodol ogy	9
2.3.1	TEM Specimen Preparation	9
2.3.2	TEM Measurement Strategy	10
2.3.3	Determination of Stopping Point	11
2.4	Quality Control/Quality Assurance	11
2.4.1	MCE Filters (0.45 and 0.8 |im Pore Size)	11
2.4.2	Lot Blanks	11
2.4.3	Laboratory Blanks	12
iv

-------
2.4.4 Interlaboratory QA/QC	12
2.4.4.1	Duplicate Analysis	12
2.4.4.2	Verified Counts	13
3	Results and Discussion	15
3.1	Fiber Retention Efficiency (Fibers >0.5 |im) 0.45 & 0.8 |im Pore Size MCE Filters 15
3.2	Effect of Plasma Etching on Total Chrysotile Asbestos Density (Fibers >0.5 |im) 18
3.3	Fiber Retention Efficiency (Fibers >5 |im) 0.45 and 0.8 |im Pore Size MCE Filters 22
3.4	Effect of Plasma Etching on Total Chrysotile Asbestos Density (Fibers >5 |im) 23
4	Conclusions and Recommendations	25
4.1	Fiber Retention Efficiency of 0.45 and 0.8 |im pore size MCE Filters	25
4.2	Effect of Etching Time for 0.45 |im pore size MCE Filters	26
4.3	Additional Recommendations	26
5	Acknowledgments	27
6	References	28
Appendix Chrysotile Asbestos Data Listing
TABLES
Number	Page
1	Batch setup for 25-mm diameter MCE filter loading experiment	5
2	Plasma etching time for 0.45 |im pore size MCE filters	9
3	Interlab oratory duplicate analysis of MCE filters for chrysotile asbestos by TEM	13
4	Interlab oratory verified count analysis of MCE filters for chrysotile asbestos by TEM	14
v

-------
5	Mean chrysotile asbestos density (s/mm2) by batch and fiber length	16
6	Mean total chrysotile asbestos (fibers >0.5 |im) density (s/mm2) by batch and filter type	17
7	Mean total chrysotile asbestos (fibers >0.5 |im) density (s/mm2) for variable etching times .18
8	Mean chrysotile asbestos density for fibers >5 |im (s/mm2) by batch and filter type	22
9	Mean chrysotile asbestos density for fibers >5 |im (s/mm2) for variable etching times	23
vi

-------
FIGURES
Number	Page
1	Transmission electron microscope photograph of 0.4 |im pore size capillary pore
polycarbonate membrane filter (16,000X magnification)	2
2	Scanning electron microscope photograph of 0.8 |im pore size cellulose ester
membrane filter (8,000X magnification)	2
3	A. Fiber densities (fibers > 0.5 |im in length) observed on 0.45 |im pore size MCE Filter by
etch time in minutes: A. Mean fiber densities; B. Scatter plot showing range of density
values of individually sampled filters (N=48)	20
4	Fiber densities (fibers >0.5 to 5 |im in length) observed on 0.45 |im pore size
MCE filter	21
5	Fiber densities (fibers >5 to 10 |im in length) observed on 0.45 |im pore size
MCE filter	21
6	Fiber densities (fibers >10 |im in length) observed on 0.45 |im pore size
MCE filter	22
vii

-------

-------
SECTION 1
INTRODUCTION
Over the years, a number of optical and electron microscopy methods have been
developed to detect and quantify asbestos in air, as well as in other matrices. Each method has
its own strengths and weaknesses, and they must be carefully evaluated to determine how best to
detect and quantify asbestos under a given circumstance.1"10
Typically, mixed cellulose ester (0.45 |im or 0.8 |im pore size) and to a lesser extent,
capillary-pore polycarbonate (0.4 |im pore size) membrane filters are used to collect airborne
asbestos for count measurement and fiber size analysis. It is important to recognize that pore
size specification for a membrane filter is an absolute specification only for capillary-pore type
filters such as the polycarbonate (PC). The pore size rating for tortuous path filters, such as the
mixed-cellulose ester (MCE) filters, is an effective pore size and not a specification that particles
exceeding that size are retained by the filter.11
The two types of filters differ in their chemical and physical composition. Polycarbonate
filters have a smooth filtering surface; the pores are cylindrical, almost uniform in diameter, and
essentially perpendicular to the surface (Figure 1). A mixed-cellulose ester filter is a thicker
filter with a sponge-like appearance and relies on a tangled maze of cellulose ester strands to trap
fibers (Figure 2). For microscopic analysis of asbestos deposited on the filter, it is critical that
the fibers be in a single plane to assure they are in focus during the analysis. This requirement is
simple to achieve for PC filters because of the smooth filtering surface. Whereas, the MCE filter
requires two additional steps in the direct preparation procedure. The MCE filter must be
collapsed with an organic solvent and then the top layer of the collapsed filter material must be
etched away with a low temperature plasma asher.
The U.S. EPA1 and the National Institute for Occupational Safety and Health (NIOSH)5
recommend using 0.45 |im pore size MCE filters when performing transmission electron
microscopy (TEM) analysis on the samples because the particles deposit closer to the surface
than in larger pore size (e.g., 0.8 |im pore size) MCE filters. However, the higher pressure drop
through the 0.45 |im pore size MCE filters normally preclude their use with battery-powered
personal sampling pumps.5 In order to obtain a uniform distribution of collected particulates
across the surface of the collecting filter, EPA1 requires a 5.0 |im pore size MCE backing filter
1

-------
be placed behind the collecting filter followed by a cellulose support. This tandem filter
assembly further increases the pressure drop, which at given velocity is directly proportional to
the thickness of the filter. ISO Method 10312:1995 also recommends the tandem filter assembly
for 0.45 jam pore size MCE as well as for the 0.4 jam pore size PC filters.'
Figure 1. Transmission electron microscope photograph of a 0.4 jtun pore size
capillary pore polycarbonate membrane filter (16,000X magnification)
(Source RT1 International, S. Doom with permission.)
Figure 2. Scanning electron microscope photograph of a 0.8 fini pore size
mixed cellulose ester membrane filter (8,000X magnification).
(Source MVA Scientific, J. Millette with permission.)
2

-------
Studies reporting the fiber retention efficiencies of MCE and PC membrane filters for
asbestos aerosols are meager. One study investigated the collection efficiencies of 8 |im pore
size MCE filters and 0.2, 0.4, and 0.8 |im pore size PC filters for aerosols of chrysotile
asbestos.13 For MCE filters with 8-|im pores, the collection efficiency at a face velocity of
3.5 cm/s fell from 100% for fibers >5 |im in length to 75% for fibers of 2 |im in length, and to
25% for fibers approximately 0.5 |im in length. For PC filters with pore diameters of 0.2, 0.4,
and 0.8 |im, collection efficiencies began to drop for fiber lengths <3 |im and fiber diameters
<0.2 |im. For 0.2 |im pores, the efficiencies for fibers >0.5 |im did not drop below
approximately 80%, whereas for 0.8 |im pores, the efficiencies dropped to near zero for fiber
lengths below 0.5 |im and diameters below 0.05 |im. This study showed that fiber retention
efficiencies decrease substantially with fiber length for both MCE and PC pore filters of larger
pore size. The orientation of the airborne fibers as they approach the filter pore entrances may
have an important effect on their ability to penetrate the filter.
A literature review did not identify any study that compared the fiber retention
efficiencies of 0.45 |im and 0.8 |im pore size MCE or 0.4 |im pore size PC membrane filters for
asbestos aerosols.12 Information culled from an informal survey12 of asbestos analytical
laboratories, members of the American Society for Testing and Materials (ASTM) and
Environmental Information Association (EIA) revealed that MCE filters were primarily used for
airborne asbestos sampling. Accordingly, it was concluded that testing of the PC filters would
not be conducted in this study allowing the project to concentrate its efforts and funding on 0.45
|im and 0.8 |im pore size MCE filters that are widely used in asbestos exposure studies today.
Therefore, U.S. EPA's National Exposure Research Laboratory (NERL) conducted a
study in which chrysotile asbestos (fibers both shorter and longer than 5 |im) were generated in
an aerosol chamber and sampled by 25-mm diameter MCE filter media to compare the efficiency
of 0.45 |im pore size versus 0.8 |i pore size filter media. In addition, the effect of plasma etching
times on fiber densities was evaluated.
3

-------
1.1 Project Objectives
The goal of this research study was to determine the effect of mixed cellulose ester
membrane filter pore size on fiber retention efficiency of chrysotile asbestos fiber aerosols. The
following are the specific objectives of this study:
•	Compare the fiber retention efficiency (structures >0.5 |im in length) of chrysotile
asbestos aerosols of 0.45 |im and 0.8 |im pore size mixed cellulose ester filters.
•	Compare the fiber retention efficiency (structures >5 |im in length) of chrysotile
asbestos aerosols of 0.45 |im and 0.8 |im pore size mixed cellulose ester filters.
•	To evaluate the effect of plasma etching time (2, 4, 6, 8, and 16 minutes) on 0.45 |im
pore size mixed cellulose ester filters on total chrysotile asbestos density (structures
>0.5 |im in length).
•	To evaluate the effect of plasma etching time (2, 4, 6, 8, and 16 minutes) on 0.45 |im
pore size mixed cellulose ester filters on total chrysotile asbestos density (structures
>5 |im in length).
4

-------
SECTION 2
STUDY DESIGN AND METHODOLOGY
2.1 Preparation of Samples for Analysis
SRI International (SRI) loaded the 25-mm diameter (0.45 |im and 0.8 |im pore size) MCE
filters with chrysotile asbestos in an aerosol chamber. The filters were prepared at two fiber
loading levels: "low" nominal loading (2-5 fibers per grid opening) and "high" nominal loading
(>5 fibers per grid opening). The filters were prepared in four batches of 18 filters each as
shown in Table 1.
Table 1. Batch setup for 25-mm MCE filter loading experiment

Filter Pore Size/Loading and Number of Samples
Batch
0.45 jim
0.8 jim

Low
High
Low
High
1
12
-
6
-
2
-
6
-
12
3
6
-
12
-
4
-
12
-
6
Total
18
18
18
18
2.1.1 SRI Dust Generation and Collection System
Test atmospheres of dusts and fibers are dynamically generated in a dust generation and
collection system engineered and built by SRI. The main components are:
•	A fluidized bed generator, which delivers a continuous stream of aerosol material;
•	A sonic velocity disperser, which disperses, de-agglomerates, and dilutes the
aerosol;
•	A settling tower, where large particles are removed; and
•	Sample collection chambers, where 320 samples can be collected simultaneously.
5

-------
All of the air streams pass through ionizers to prevent static charge effects. The
components are described in detail below.
2.1.1.1.	Fluidized Bed Generator
A variety of SRI custom-designed and constructed feeders can be used to introduce
particulates and fibers into the collection system. Asbestos fiber atmospheres have been
generated using a two-component fluidized bed consisting of bronze powder and sized chrysotile
asbestos fibers. By proper adjustment of air flow through the bottom and across the top of the
bed, a pressure differential is established sufficient to fluidize the bronze powder bed and the
chrysotile asbestos fibers are stripped at a low rate and fed to the sonic velocity disperser. There
is a concentration gradient using this system because the chrysotile asbestos is depleted from the
fluidized bed. However, because a homogeneous atmosphere is produced, all 320 sampling ports
will still collect an equivalent amount of chrysotile asbestos. By varying the sampling time, the
chrysotile asbestos loading on the cassettes can be adjusted.
2.1.1.2.	Sonic Velocity Disperser and Settling Tower
The air stream from the dust feeder carries the aerosol to the sonic velocity disperser.
Dilution air is also delivered to the sonic velocity disperser, where it de-agglomerates the aerosol
under the action of an on-line static eliminator and high air velocity. The aerosol then enters the
settling tower, the linear velocity is reduced, and the larger particles settle out to the base of the
settling tower. The diluted aerosol is then divided uniformly among the four collection
chambers.
2.1.1.3.	Sample Collection Chamber
The base section of the sample collection chamber consists of layers of gaskets and
machined aluminum sheets. Eighty sampling ports are situated in an 8 x 10 matrix arrangement.
Downstream from each port is a critical flow orifice. The mounting sheet in which the 80 critical
flow orifices are embedded forms the upper section of a vacuum chamber, so that a vacuum to
6

-------
this chamber creates the necessary pressure differential to operate the orifices. Aerosol enters
the collection chambers through 20 symmetrically located passages. The 320 orifices (80 for
each of four sample collection chambers) all have the same diameter and were calibrated at the
time the system was constructed to ensure that all the ports sample at the same flow rate. The
orifices form a matched set, with a maximum flow rate of 2 L/min through each air monitor in
the system.
The collection chambers can be opened from the top by removing a cover. Air monitor
cassettes are connected to the sampling port by a Luer fitting. A variety of cassette and filter
types and sizes, including 25- and 37-mm-diameter cassettes, mixed cellulose ester filters, and
polycarbonate filters, can be accommodated in the collection chambers.
2.1.1.4 Dust Feeder
The chrysotile asbestos-containing powder was metered into the collection system by a
grooved disk, which rotates at a known rate. The powder is pneumatically unloaded from a
groove in the disk and then conveyed to a sonic velocity disperser.
Powder is loaded into the top of the powder hopper through the powder feed port. The
powder then drops down into the hopper connector, where it is pushed into the groove of the disk
by rubber wipers attached to the bottom of the agitator shaft. A spring-loaded guard ring
surrounds the hopper connector and scrapes the disk to prevent the disk from carrying away
excess powder. The rotation of the disk continuously carries the powder in the groove to the
unloading nozzle, where it is removed pneumatically by compressed air. The powder feed rate is
determined entirely by the rotation speed of the disk and the size of the groove. The loading of
powder on sample filters is further adjusted by varying the collection time.
Chrysotile asbestos fiber atmospheres are generated using a two-component fluidized bed
consisting of bronze powder and sized chrysotile asbestos fibers. By proper adjustment of air
flow through the bottom and across the top of the bed, the bronze powder bed is fluidized and the
chrysotile asbestos fibers are stripped at a low rate and fed to the sonic velocity disperser. There
is a concentration gradient using this system because the chrysotile asbestos is depleted from the
fluidized bed. However, because a homogeneous atmosphere is produced, each sampling port
still collects an equivalent amount of chrysotile asbestos. By varying the sampling time, the
7

-------
chrysotile asbestos loading on the cassettes is adjusted. By using a combination of the fluidized
bed and the powder feeder, a variety of fibers and particulates is loaded onto a filter.
2.1.1.5	Sonic Velocity Disperser and Settling Tower
The air stream from the dust feeder carries the aerosol to the sonic velocity disperser.
Dilution air is also delivered to the sonic velocity disperser, where it de-agglomerates the aerosol
under the action of an on-line static eliminator and high air velocity. The aerosol then enters the
settling tower, the linear velocity is reduced, and the larger particles settle out to the base of the
settling tower. The diluted aerosol is then divided uniformly among the four collection
chambers.
2.1.1.6	Sample Collection Chamber
The base section of the sample collection chamber consists of layers of gaskets and
machined-aluminum sheets. Eighty sampling ports are situated in an 8 x 10 matrix arrangement.
Downstream from each port is a critical flow orifice. The mounting sheet in which the 80 critical
flow orifices are embedded forms the upper section of a vacuum chamber, so that a vacuum to
this chamber creates the necessary pressure differential to operate the orifices. Aerosol enters
the collection chambers through 20 symmetrically located passages. The 320 orifices (80 for
each of four sample collection chambers) all have the same diameter and were calibrated at the
time the system was constructed to ensure that all the ports sample at the same flow rate. The
orifices form a matched set, with a maximum flow rate of 2 L/min through each air monitor in
the system. SRI collected 100 filters in each batch, and utilized 80 of the primary filter cassettes
(e.g., in a 0.45 |_im pore size batch the 0.45 |_im pore size filters are the primary filter cassettes)
and 20 of the secondary filter cassettes (e.g., in a 0.45 |_im pore size batch the 0.8 |_im pore size
filters are the secondary filter cassettes), so that the variable loadings of different batches could
be adequately measured and controlled. The 80 primary filter cassettes and 20 secondary filter
cassettes were divided evenly between the four quadrants of the chamber. It should be noted that
the fiber loading process is trial and error. That is, the chrysotile asbestos structures per area of
filter will be different for two filters in the same loading category.
8

-------
The collection chambers are opened from the top by removing a cover. Air monitoring
filter cassettes are connected to the sampling port using a luer fitting. Quality control activities
include checking each orifice flow rate with a digital flow meter before and after sample
generation and analyzing for background levels to prevent carryover contamination.
2.2 Sample Analysis Strategy
2.2.1	Fiber Retention Efficiency of 0.45 jim and 0.8 jim Pore Size MCE Filters
Seventy-two filter samples were prepared and analyzed to test pore size differences and
fiber loading differences between the two MCE filter types (see Table 1). Eighteen filters were
analyzed for each of four batches. Twelve of the primary filters and six of the secondary filters
were analyzed for each batch.
2.2.2	Effect of Plasma Etching Time on Chrysotile Asbestos Density
Annex A "Determination of Operating Conditions for Plasma Asher" of ISO Method
10312:1995 requires etching of collapsed filters for 8 minutes using operating parameters
determined for completely ashing uncollapsed filters in 15 minutes. Including the specified 8
minute etching time, three additional etching times were used to etch the 0.45 |_im MCE filters
(Table 2). Hence, a total of 12 filters were etched for each of four different times (2, 4, 8, and
16-minutes). The filters were loaded at a "high" nominal loading.
Table 2. Plasma etching time for 0.45 jum pore size MCE filters
Filter Loading
Plasma Etching Time (Minutes) and
Number of Samples

2
4
8
16
High
12
12
12
12
2.3 Analytical Methodology
2.3.1 TEM Specimen Preparation
9

-------
TEM specimens were prepared from the air filters using the dimethylformamide (DMF)
collapsing procedure of ISO 10312:1995, as specified for cellulose ester filters. DMF was used
as the solvent for dissolution of the filter in the Jaffe washer. Prior to etching the filters, a March
Plasmod asher was calibrated in accordance with ISO 10312:1995 procedures whereby an
uncollapsed filter was oxidized under controlled settings in approximately 15 minutes. After
asher calibration, the filters were prepared using ISO 10312:1995 procedures and etched for
either 2, 4, 8, or 16-minutes. For each filter, an equal number of grid openings were examined
on at least two prepared TEM specimen prepared from a one-quarter sector of the filter using 200
mesh-indexed copper grids. The remaining part of the filter was archived in the original cassette
in clean and secure storage.
2.3.2 TEM Measurement Strategy
1.	The minimum aspect ratio for the analyses was 3:1, as permitted by ISO 10312:1995.
As required in the ISO Method, any identified compact clusters and compact matrices
were counted as total chrysotile asbestos fibers, even if the 3:1 aspect ratio was not
met.
2.	All fibers larger than or equal to (>) 0.5 |_im in length were quantified with the
following breakdown according to ranges by length: a) >0.5 to 5.0 |_im; b) >5.0 to
10.0 |_im; and c) >10.0 |_im.
3.	The fiber counting data was distributed approximately equally among a minimum of
two specimen grids prepared from different parts of the filter sector.
4.	The TEM specimen examinations were performed at approximately 20,000
magnification.
5.	Phase contrast microscopy-equivalent chrysotile asbestos structures (PCME) were
also determined. PCME chrysotile asbestos structures, as defined by ISO
10312:1995, are
>5 |_im in length and from 0.2 to 3.0 |_im in diameter with an aspect ration >3:1.
10

-------
2.3.3 Determination of Stopping Point
The analytical sensitivity was > 6 chrysotile asbestos structures per square millimeter
(s/mm ). In principle, any analytical sensitivity can be achieved by increasing the number of
grid openings or fields examined. Likewise, statistical uncertainty around the number of fibers
observed can be reduced by counting more fibers. Stopping rules are needed to identify when
microscopic examination should stop, both at the low end (zero or very few fibers observed) and
at the high end (many fibers observed). The analysis was terminated upon completion of
counting >25 chrysotile asbestos structures in a minimum of 10 grid openings or 100 chrysotile
asbestos structures in 4 grid openings. In any case completion of the grid opening being
analyzed when the stopping rules have been met was completed.
2.4 Quality Control/Quality Assurance
2.4.1	MCE Filters (0.45 jim and 0.8 jim pore size)
The filter samples generated by SRI were monitored for absolute concentration and for
intra-batch uniformity by an independent quality control (QC) laboratory, RTI International
(RTI). RTI prepared and analyzed samples and provided feedback to SRI regarding filter density
so that the batches meet the target densities. They also used the data to validate the uniformity of
density of filters within each batch. For each batch of filters produced, a relative standard
deviation (RSD) of fibers per grid opening was developed with 40 grid openings analyzed.
Based upon historical RSD levels for SRI filters, each batch was expected to have an RSD at or
below 0.50, which was the case for this study.
2.4.2	Lot Blanks
Before filter samples were loaded with chrysotile asbestos two unused filters from each
filter lot of 0.45 and 0.8 jam filters were analyzed by the QC laboratory to determine the mean
chrysotile asbestos structure count. The lot blanks were analyzed for chrysotile asbestos
structures by using ISO 10312:1995. In all cases the mean count for all types of chrysotile
asbestos structures was
<18 structures/mm2.
11

-------
2.4.3 Laboratory Blank
Laboratory blanks are unused filters that are prepared and analyzed in the same manner
as the field samples to verify that reagents and equipment are free of the subject analyte, and that
contamination has not occurred during the analysis process. The laboratory analyzed two 0.45
|im and two 0.8 |im pore size MCE filters. Blanks were prepared and analyzed along with the
other samples. Chrysotile asbestos was not present on any of the samples at an analytical
sensitivity of 8.9 s/mm2.
2.4.4 Interlaboratory QA/QC
After analysis by the primary laboratory (Clayton Group Services, Inc.), selected filters
and grid preparations were sent to the QC laboratory for analysis as an independent QA/QC
check. The QA/QC sample analyses included duplicates and verified counts by TEM.
2.4.4.1 Duplicate Analyses
The duplicate analyses was conducted by repreparing and analyzing the same filter
using the same ISO 10312:1995 counting rules. Results of the QC duplicate analysis are
presented in Table 3. In Table 3, the third column lists the number of structures analyzed, and
the fourth column lists the density of chrysotile asbestos structures per unit area. Note: The
primary laboratory used a grid opening size of 0.011 mm2, and the QC laboratory used a grid
opening size of 0.0086 mm2. Column 5 presents the results of the duplicate sample variability.
All four interlaboratory duplicate samples met the acceptance criteria.
12

-------
Table 3. Interlaboratory duplicates analysis of MCE filters for chrysotile asbestos by TEM
Sample No.
Laboratory
Analyses
Actual
Variability11
Accepted
Variability
# Structures
Structures/mm2
A0611022-
001A
Primary
26
230


QC
23
270
1.8
2.24
A0611022-
002A
Primary
28
250


QC
19
220
1.3
2.24
A0611022-
003A
Primary
34
300


QC
27
310
0.39
2.24
A0611022-
004A
Primary
31
280


QC
22
280
0.86
2.24
aAnalvtical Variability = (Analysis A) - (Analysis B)
V(Analysis A + Analysis B)
This variability is the absolute value of the difference of the two analyses, divided by the square root of the
sum, which is an estimate of the standard deviation of the difference based on a Poisson counting model. The
value 2.24 was selected as targeting false positive rates of 2.5% (1/40) for the Poisson model.
2.4.4.2 Verified Counts
Verification counting involved re-examination of the same grid openings analyzed by the
primary laboratory. The verification counting was performed on two of the analyses for each of
the filter pore sizes. Verified counting was conducted using the procedure defined in NISTIR
5351, "Airborne Asbestos Method: Standard Test Methodfor Verified Analysis of Asbestos by
Transmission Electron Microscopy - Version 2.0."
Results of interlaboratory QC verified counting by TEM are presented in Table 4. In
Table 4, the third column gives the total number of chrysotile asbestos structures counted in the
specified grid openings which were determined to be true positives (TP). Column 4 gives the
number of false positives (FP) and Column 5 gives the number of false negatives (FN). The
results of all four analyses are combined at the bottom, and ratios of true positives, false
positives, and false negatives are developed in the final two rows for both the primary and QC
laboratory. Column 6 shows the "pass" (Yes) or "fail" (No) status of the comparison. The
acceptable variability is >80% true positives, <20% false negatives, and <20% false positives.
All interlaboratory verified count analysis met the acceptance criteria.
13

-------
Table 4. Interlaboratory verified count analysis of MCE filters for chrysotile asbestos by
TEM
Sample No.
Laboratory
Number of Structures
Pass?
True Positive
False Positive
False Negative
A0611024-
Primary
24
0
1

003A
QC
25
1
0

A0611024-
Primary
35
3
0

002A
QC
32
0
3

A0611024-
Primary
5
0
1

004A
QC
6
1
0

A0611024-
Primary
8
0
0

005A
QC
7
0
1

Totals
Primary
72
3
2

QC
70
2
4

Percentages
Primary
97%
4%
3%
Yes
QC
95%
3%
5%
Yes
14

-------
SECTION 3
RESULTS AND DISCUSSION
3.1 Fiber Retention Efficiency (Fibers >0.5 jim) of 0.45 and 0.8 jim Pore Size MCE
Filters
Chrysotile asbestos fibers are the most difficult to see and count after capture, so
they present a worst case for retaining and counting fibers after they are retained on
filters. Amphiboles are easier to see and count after capture. Therefore, this study was not
so concerned about the capture of fibers, but in seeing them with electron microscopy
after capture. Deeply embedded fibers present a particular challenge. Filtration theory13
states that the most penetrating particle size decreases with decreasing size of filter
medium. Thinner fibers, therefore, penetrate a filter matrix more deeply than thicker
fibers, making microscopy more difficult.
Due to the fact that it is the most common fiber type in most asbestos exposure
scenarios to date and owing to its finely fibrous nature it is also the ideal form of asbestos
to study post-preparation fiber retention in filters.14 equation for predicting most
penetrating particle diameter (dp.min) is:
Where, K is the hydrodynamic factor, a = solidity of filter (1 - porosity), X is the mean
free path of the gas molecules, &is the Boltzmann constant, T is the absolute temperature,
rj is the air viscosity, df is the filter fiber diameter, and U is average air velocity inside the
filter medium. Therefore, the most penetrating particle diameter decreases with
decreasing pore size in the filter medium. This relationship holds for both fibrous and
membrane filters16, and in porous-membrane filters 14
Of all asbestos fiber types, chrysotile is the most likely to penetrate the tortuous
matrix of MCE filter material, thus optimizing the ability of the study to employ electron
microscopy to characterize differences in asbestos post-preparation fiber retention, due to
MCE pore size (larger pore sizes equate to greater potential penetration of fibers into the
15

-------
matrix) and due to differential plasma etching time. Amphibole asbestos fibers, with their
larger average diameter and length,17 are less likely to penetrate the MCE matrix, and
therefore more easily visible than most chrysotile fibers by microscopy. In addition,
since chrysotile asbestos is by far the most commonly seen asbestos type on air filters
(such as from remediation sites), it best reflects real-world situations. Thus, these results
for chrysotile asbestos provide an indication of filter effectiveness for numerous fibers,
including amphibole asbestos.
A total of 72 filters (18 of 0.45 [j,m pore size and 18 of 0.8 [j,m pore size) were loaded
with chrysotile asbestos at two filter loadings (low = 2-5 fibers/grid opening; and high = >5
fibers/grid opening). The experiment was conducted in 4 batches of 18 filters each (Table 1).
All chrysotile asbestos structures >0.5 |im in length were quantified and categorized
according to three ranges by length: >0.5 to 5 |im; >5 to 10 |im; and >10 |im. The chrysotile
asbestos fiber distribution for the low and high filter loadings is presented in Table 5.
Table 5. Mean chrysotile asbestos density (s/mm2) by batch and fiber length
Mean fiber density (s/mm2) by length of fibers (standard deviation)
Batch	Filter Low Loading	Filter High Loading
>0.5-5 |im >5-10 |im >10 |im >0.5-5 |im >5-10 |im >10 |im
0.45 |im pore size
1	237 (113) 63 (22) 21 (112)
2	-	-	-	585 (92) 284(86) 89(30)
3	317(41) 71(21) 21(14)
4	-	-	-	1200(296) 235 (91) 66(48)
0.8 |im pore size
1	194(44) 60(21) 19(13)
2	-	-	-	429 (80) 225 (76) 88 (33)
3	288 (69) 78 (30) 22(10)
4	-	-	-	960 (299) 261 (70) 71 (27)
The mean filter density (total chrysotile asbestos structures per mm2) for the two filter
types in each batch is presented in Table 6. In each batch, the mean density on the 0.45 |im
filters is higher than on the 0.8 |im filters. In Batch 2, the difference is statistically significant
using both the two-sample t-test (p = 0.008) and the nonparametric Wilcoxon Rank-Sum test
(p = 0.01). In the other 3 batches, the difference is not statistically significant.
16

-------
Table 6. Mean total chrysotile asbestos density (s/mm2) by batch and filter type
Batch Mean fiber density (s/mm2) by filter pore size and nominal loading (standard
deviation)
0.45 jim	0.8 jim
Low Loading High Loading Low Loading High Loading
1	321 (114)	-	274(57)
2	-	958 (170)	-	743 (125)
3	413 (55)	-	388 (89)
4	-	1512(369)	-	1304 (336)
It is apparent from Table 6 that the two "Low Loading" batches differ substantially, as do
the two "High Loading" batches. For example, the 0.8 |im density in Batch 3 is higher than the
0.45 |im density in Batch 1, even though both batches were loaded at the same nominal level.
Likewise, the 0.8 |im density in Batch 4 is higher than the 0.45 |im density in Batch 2. In fact,
the between-batch differences (at the same nominal loading) are greater than the differences
between the two filter types. Thus, it is not appropriate to combine the 4 batches into a single
dataset for purposes of an overall comparison between the two filter types (Primary Objective 1).
To make the overall comparison, the sum of the Wilcoxon statistics for the 4 separate
batches was used. In each batch, the Wilcoxon statistic is the rank-sum for the 0.45 |im densities
in the 18 samples comprising the batch. Under the null hypothesis that the two filter types have
the same fiber retention efficiency, this statistic has (approximately) a normal distribution with
mean 57 (Batches 2 and 3) or 114 (Batches 1 and 4), and variance 114 (all batches). Thus, under
the null hypothesis, the sum of the 4 Wilcoxon statistics is approximately normal with mean 342
and variance 456. The observed value of the sum is 395.5, resulting in a test statistic z = (395.5-
342)/21.4 = 2.50, with a p-value of 0.01. Thus, the null hypothesis is rejected, and we conclude
that the fiber retention efficiency of the 0.45 |im pore size filter for fibers >0.5 |im in length is
significantly higher than that of the 0.8 |im pore size filter. However, for fibers >5 |im in length
there is no difference in the two filter pore sizes (see Figures 5 and 6, Section 3.2).
3.2 Effect of Plasma Etching Time on Total Chrysotile Asbestos Density (Fibers >0.5
jim)
Four different etching times were used to etch 0.45 |im filters. A total of 12 filters were
etched for each of the 4 times (2, 4, 8 and 16 minutes). The filters were loaded in at the "High"
nominal loading. Table 7 shows the mean total chrysotile asbestos density (s/mm2) for fibers
>0.5 jam in length for each etching time.
17

-------
Table 7. Mean total chrysotile asbestos (fibers >0.5 jim) density (s/mm2) for variable
etching times
Filter Loading Plasma etching time for 0.45 jim MCE filters (minutes). Standard
deviation in parentheses.
2	4	8	16
High	1123 (295)	1251 (442)	1512(369)	1635 (236)
The mean fiber retention density increases with etching time. To examine the relationship
between etching time and density, two regression models were fit to the data. The first model
assumes a linear relationship between etching time (t) and density (TA):
TA = a + b*t	(1)
The fitted equation was
TA (s/mm2) =1113+ 35.7*t	(R2 = 0.24)
The relatively low value of R2 is due to the considerable variability in densities observed
at each etching time. However, the coefficient of t is highly significant (SE = 9.27). This
regression indicates that each additional minute of etching time adds an average of
35.7 s/mm2 to the total chrysotile asbestos density within the range tested.
The second regression model assumes a logarithmic relationship between density and
etching time, of the form
TA = a + b*log(t)	(2)
Here, "log" denotes the natural logarithm (In). The fitted equation was
TA (s/cm2) = 931 + 259*ln(t)	(R2 = 0.27)
R2 is slightly higher than for the linear model. Again, the regression is highly significant
(SE of coefficient = 63). This model estimates that a doubling of the etching time adds an
average of 13% to the total chrysotile asbestos density within the range tested.
On physical grounds, it would appear that a point of diminishing returns for increased
18

-------
etching time would be reached, i.e., there is a level of etching time beyond which no increase in
density is expected (Figures 3 and 4). The data from this experiment do not appear to shed light
on what this level might be. For example, the increase in density from 8 to 16 minutes is
comparable to that from 2 to 4 minutes. However, the increase in density with etching time does
not appear to be the case for fibers >5 |im in length (Figures 5 and 6). These data suggest that
the etching time of 8 minutes that is specified in ISO 10312:1995 is adequate for fibers >5 |im in
length. If fibers <5 |im in length are of interest, additional research may be needed to determine
the optimum etching time.
Etch Time (min)
A
19

-------
2500
2000
1500
1000
~
~
~
~
500
~
~
B
8	10
Etch Time (min)
12
14
16
18
Figure 3. A. Fiber densities (fibers > 0.5 jim in length) observed on 0.45 jim pore size MCE
Filter by etch time in minutes: A. Mean fiber densities; B. Scatter plot showing range of
density values of individually sampled filters (N=48).
Etch time (min)
Figure 4. Mean chrysotile asbestos fiber densities (fibers > 0.5 to 5 jim in length) observed
on 0.45 jim pore size MCE filter at varying etching times.
20

-------
E
E
£
0)
0)
Si
E
3
1100
900
700
500
300
100
10
Etch time (min)
15
20
Figure 5. Mean chrysotile asbestos fiber densities (fibers 5 to 10 jim in length) observed on
0.45 jim pore size MCE filter at varying etching times.
300
10
Etch time (min)
15
20
Figure 6. Mean chrysotile asbestos fiber densities (fibers >10 jim in length) observed on
0.45 jim pore size MCE filter at varying etching times.
3.3 Comparison of Fiber Retention Efficiency of 0.45 jim and 0.8 jim Pore Size Filters
for Fibers > 5 jim in Length
The mean filter loading (total chrysotile asbestos structures, > 5 [j,m, per mm2) for the two
filter types in each batch is shown in Table 8.
21

-------
Table 8. Mean chrysotile asbestos density (s/mm2) for fibers >5 jim in length by batch and
filter type
Batch	Mean fiber density (s/mm2) by filter pore size and nominal loading
(standard deviation)
0.45 jim	0.8 jim
Low Loading High Loading Low Loading High Loading
1	84 (30)	-	80 (28)
2	-	373 (102)	-	313 (104)
3	92 (33)	-	100(32)
4	-	301 (118)	-	333 (85)
In Batches 1 (Low loading) and 2 (High loading), the mean density on the 0.45 jam filters
is higher than on the 0.8 jam filters. In Batches 3 (Low loading) and 4 (High loading), the
reverse is true; i.e., the 0.8 |im filters are higher. None of the differences are statistically
significant using both the two-sample t-test. When Batches 1 and 3 (Low loading), and Batches
2 and 4 (High loading), are combined, the differences between the filter types are even smaller.
We conclude that, for fibers > 5 |iin, there is no difference between the fiber retention
efficiencies of the 0.45 |im and 0.8 |im filters.
3.4 Effect of Plasma Etching Time on Chrysotile Asbestos Fibers >5 jim
Table 9 shows the mean total chrysotile asbestos density for fibers > 5 [j,m (s/mm2) for
each etching time. The mean densities for the 2, 8 and 16 minute etching times are virtually
identical. The mean density for the 4 minute etching time is a little lower.
Table 9. Mean chrysotile asbestos density for fibers > 5 jim (s/mm2)
for variable etching times (standard deviation)
Filter Loading	Plasma etching time for 0.45 jim MCE filters (minutes)
2	4	8	16
High	301 (75)	232(99)	301 (118)	303 (77)
To examine the relationship between etching time and density, two regression models
were fit to the data. The first model assumes a linear relationship between etching time (t) and
density (TA):
22

-------
TA = a + b*t
(3)
The fitted equation was
TA (s/mm2) = 267 + 2.2*t
(R2 = 0.016)
The regression is not statistically significant.
The second regression model assumes a logarithmic relationship between density and
etching time, of the form
Again, the regression is not statistically significant. The fact that neither regression is
statistically significant indicates that, for 0.45 |im filters, there is no statistically significant
relationship between etching time and density of fibers > 5 |im.
This is consistent with a study conducted by Chatifield.11 The study showed that fiber
densities for fibers longer 5 |im are similar for 0.2 |im pore size PC filters and various etching
schedules for 0.22 |im pore size MCE filters. In particular, plasma etching had no effect on the
reported fiber densities of fibers longer than 5 |im. At the 1% significance level, there was no
statistically significant differences between the mean fiber densities for any of the etching
preparations evaluated.
TA = a + b*log(t)
(4)
Here, "log" denotes the natural logarithm. The fitted equation was
TA (s/mm2) = 266 + 10.5*log(t)
(R2 = 0.007)
23

-------
SECTION 4
CONCLUSIONS AND RECOMMENDATIONS
Fiber Retention Efficiency of 0.45 and 0.8 jim Pore Size MCE Filters
The type of asbestos chosen for this study was chrysotile asbestos, due to its finely
fibrous nature. Of all asbestos fiber types, chrysotile is the most likely to penetrate the
tortuous matrix of MCE filter material, thus optimizing the ability of the study to see
differences in asbestos fiber retention efficiency, due to MCE pore size (larger pore sizes
equate to greater potential penetration of fibers into the matrix) and due to differential
plasma etching time. Amphibole asbestos fibers, with their larger average diameter and
length, are less likely to penetrate the MCE matrix. In addition, since chrysotile asbestos
is by far the most commonly seen asbestos type on air filters (such as from remediation
sites), it best reflects real-world situations. Thus, these results for chrysotile asbestos
provide an indication of filter effectiveness for numerous fibers, including amphibole
asbestos.
Conclusion—The null hypothesis was that the two mixed-cellulose ester (MCE) filter
types (0.45 |im and 0.8 |im pore size) have the same fiber retention efficiency for
chrysotile asbestos aerosol (structures >0.5 |im length). The null hypothesis was rejected,
and it is concluded the fiber retention efficiency of the 0.45 |im pore size MCE filter is
statistically significantly higher than that of the 0.8 |im pore size MCE filter (p=0.01) for
fibers >0.5 |im in length. However, for chrysotile asbestos structures >5 |im in length,
there is no statistically significant difference between the fiber retention efficiencies of
0.45 |im and 0.8 |im pore size MCE filters (p>0.05).
Recommendation—This research study demonstrates that the fiber retention efficiency
of a 0.45 |im pore size MCE filter for aerosols of chrysotile asbestos fibers (structures
>0.5 |im) is greater than that for a 0.8 |im pore size MCE filter. However, there is no
difference in fiber retention efficiency between these pore sizes for structures longer than
5.0 |im. If the exposure study is focused on fibers less than 5.0 |im, the investigator

-------
should use filters with 0.45 jam pore size. If the exposure study is only interested in
structures longer than 5.0 |im, then either filter pore size may be used.
4.2	Effect of Etching Time for 0.45 jim MCE Filters
Conclusion—There is a significant difference in the effect of etching times for fibers
<5.0 |im and fibers >5.0 |im in length. The mean density of chrysotile asbestos fibers
>0.5 |im in length increases with etching time (2, 4, 8, and 16-minutes) of 0.45 |im pore
size MCE filters. Regression analysis of etching time and densities showed that doubling
the etching time adds an average of 13% to the total chrysotile asbestos density within
the range tested. This increase is a diminishing percentage of the total fiber count as the
etching time increases; e.g., 20% at 2 minutes, and 12% at 8 minutes. There is likely an
etching time beyond which no increase in density is expected and in fact would decrease;
the data from this experiment did not identify this etching time. However, etching the
filter for longer periods may remove too much filter so that a specimen for TEM analysis
cannot be prepared.
For fibers > 5.0 |im in length, there is no significant difference in numbers of structures
counted at the etching times used in these tests.
Recommendation—Since most asbestos exposure risk models include fibers >5.0 |im in
length, the 8 minute etching time specified in ISO 10312:1995 is adequate. However, if
an exposure study is focused on fibers < 5.0 |im in length, the etching time of 8-minutes
should be reviewed. A study should be conducted to determine the etching time beyond
which no significant increase in asbestos density of fibers <5.0 |im in length is expected.
4.3	Additional Recommendations
• NIOSH Method 7402 notes that a 0.45 |im pore size filter may be difficult to use
with some personal sampling pumps due to the pressure drop across this filter. The
tandem MCE filter assembly (0.45 |im pore size filter and 5 |im pore size diffusing
filter) recommended by AHERA (40 CFR §761), ISO Method 10312:1995, and
25

-------
ASTM Method D 6281-04 may preclude the use of some battery-powered personal
sampling pumps due to the resultant high pressure drop. Analysis of filters by TEM
require the use of the 5 |im pore size diffusing filter to assure uniform deposition on
the primary collection filter. A study should be conducted to evaluate the difference
between chrysotile asbestos aerosols collected on 0.45 |im and 0.8 |im pore size
MCE filters with and without the 5 |im pore size MCE diffusing filter. Also,
specifications for personal pumps should be investigated to determine optimum
requirements for sampling using the 0.45 |im pore size collection filter and 5 |im
pore size diffusing filter combination.
• This study has focused on MCE filters since this filter type is the primary choice for
air monitoring. Exposure to asbestos through inhalation is considered the most
likely route for asbestos exposure. Polycarbonate (PC) filters are used in monitoring
asbestos in water and possibly by some studies of inhalation. Since no data has been
found comparing the relative effectiveness of MCE and PC filters, research should
be considered to compare the retention of asbestos fibers on 0.45 |im pore size MCE
filters to 0.4 |im pore size polycarbonate filters.
SECTION 5
ACKNOWLEDGMENTS
This report is the result of recommendations and insights from numerous scientists interested in
advancing the state-of-the-science of airborne fiber measurements. In particular, the experts in
the EPA Regional Offices, most notably Julie Wroble, Mark Maddaloni, Aubrey Miller, Phil
King and Mary Goldade, provided reviews and recommendations from the onset. The comments
and recommendations gained from the peer review of a companion article to be published in the
Journal of Occupational and Environmental Hygiene also enhanced the presentation of the
information gained from this research.
26

-------
SECTION 6
REFERENCES
1.	US Code of Federal Regulations, 40 CFR Part 763, Appendix A to Subpart E. "Interim
Transmission Electron Microscopy Analytical Methods - Mandatory and Nonmandatory -
and the Mandatory Section to Determine Completion of Response Action."
2.	American Society for Testing and Materials, "Standard Test Method for Airborne
Asbestos Concentration in Ambient and Indoor Air as Determined by Transmission
Electron Microscopy Direct Transfer (TEM)", ASTM D 6281.
3.	International Organization for Standardization, "Ambient Air - Determination of asbestos
fibres - Direct-transfer transmission electron microscopy method," ISO 10312, 1995.
4.	International Organization for Standardization, "Ambient Air - Determination of asbestos
fibres - indirect-transfer transmission electron microscopy method," ISO 13794, 1999.
5.	National Institute for Occupational Safety and Health. "Asbestos by TEM," Method
7402, Issued August 15, 1994. NIOSH Manual of Methods,
http://www.cdc.gov/niosh/nmam/.
6.	United States Environmental Protection Agency. "Superfund Method for the
Determination of Asbestos in Ambient Air, Part 1: Method." EPA 540/2-90-005a. May
1990.
7.	United States Environmental Protection Agency. "Methodology for the Measurement of
Airborne Asbestos by Electron Microscopy," Yamate, G. Agarwal, S. C. and Gibbons, R.
D. Draft Report, EPA Contract 68-02-3266 for Environmental Monitoring Systems
Laboratory, Office of Research and Development, Research Triangle Park, NC 27711.
27

-------
8.	National Institute for Occupational Safety and Health. "Asbestos and Other Fibers by
PCM," Method 7400, Issue 2, August 15, 1994. NIOSH Manual of Methods,
http://www.cdc.gov/niosh/nmam/.
9.	Occupational Safety and Health Administration, U. S. Department of Labor. "Asbestos
in Air, Method Number ID-160." July 1997,
http://www.osha.gov/dts/sltc/methods/toc.html
10.	International Organization for Standardization, "Air Quality - Determination of the
number concentration of airborne inorganic fibers by phase contrast microscopy -
Membrane filter method." ISO 8672, 1993.
11.	Chatfield, E.J. Measurements of Chrysotile Fiber Retention Efficiencies for
Polycarbonate and Mixed Cellulose Ester Filters, Eric J. Chatfield, Advances in
Environmental Measurement Methods for Asbestos, American Society for Testing and
Materials, Special Technical Publication 1342, 2000.
12.	Beard, M.E. and J.R. Kominsky. Sampling and Analysis of Asbestos Fibers on Filter
Media to Support Exposure Assessments: Scoping Effort. Report prepared by
Environmental Quality Management, Inc., Cincinnati, OH and RTI International,
Research Triangle Park, NC for U.S. EPA, National Exposure Research Laboratory, Task
Order No. 0020, Contract No. 68-C-00-186 (December 2006).
13.	Spurny, K. On the Filtration of Fibrous Aerosols. J. Aerosol. Sci. 16(3): 450-455
(1986).
14.	Baron, P. A. and K. Willeke, (2005). Aerosol Measurement: Principles, Techniques, and
Applications. 2nd Edition. Wiley-Interscience, Hoboken, NJ. 213-215.
15.	Lee, K.W. and B.Y.H. Liu.(1980). On the minimum efficiency and the most penetrating
particle size for fibrous filters. Journal of the Air Pollution Control Assocation. 30: 377-
381.
28

-------
Wylie AG, R.L. Virta, and E. Russek (1985). Characterizing and discriminating airborne
amphibole cleavage fragments and amosite fibers: implications for the NIOSH method:
American Industrial Hygiene Journal. 46:197-201.
Rubow, K.L. (1981). Submicrometer Aerosol Filtration Characteristics of Membrane
Filters. Ph.D. thesis, University of Minnesota, Minneapolis, MN.
29

-------
Appendix - Chrysotile Asbestos Data Listing
Batch
Loading
Sample ID
Grid
Openings
Counted
Grid
Opening
Size
(mm)
Filter Type
(MCE =
mixed
cellulose
ester)
Size
(mm)
Effective
Filter
Area
(mm2)
Total
Chrysotile
Asbestos
Structures
(s/mm2)
1
Low
A0611022-
002A
10
0.01122
MCE Filter,
0.45 jjm
25
385
249.6
1
Low
A0611022-
003A
10
0.01122
MCE Filter,
0.45|jm
25
385
303.0
1
Low
A0611022-
006A
10
0.01122
MCE Filter,
0.45|jm
25
385
249.6
1
Low
A0611022-
008A
10
0.01122
MCE Filter,
0.45|jm
25
385
606.1
1
Low
A0611022-
01 OA
10
0.01122
MCE Filter,
,45|jm
25
385
267.4
1
Low
A0611022-
012A
10
0.01122
MCE Filter,
0.45|jm
25
385
311.9
1
Low
A0611022-
013A
10
0.01122
MCE Filter,
0.45|jm
25
385
294.1
1
Low
A0611022-
014A
10
0.01122
MCE Filter,
0.45|jm
25
385
267.4
1
Low
A0611022-
015A
12
0.01122
MCE Filter,
0.45|jm
25
385
185.7
1
Low
A0611022-
016A
10
0.01122
MCE Filter,
0.45|jm
25
385
481.3
1
Low
A0611022-
017A
10
0.01122
MCE Filter,
0.45|jm
25
385
338.7
1
Low
A0611022-
018A
10
0.01122
MCE Filter,
0.45|jm
25
385
294.1
1
Low
A0611022-
001A
10
0.01122
MCE Filter,
0.8|jm
25
385
231.7
1
Low
A0611022-
004A
10
0.01122
MCE Filter,
0.8|jm
25
385
276.3

-------
Batch
Loading
Sample ID
Grid
Openings
Counted
Grid
Opening
Size
(mm)
Filter Type
(MCE =
mixed
cellulose
ester)
Size
(mm)
Effective
Filter
Area
(mm2)
Total
Chrysotile
Asbestos
Structures
(s/mm2)
1
Low
A0611022-
005A
10
0.01122
MCE Filter,
0.8|jm
25
385
267.4
1
Low
A0611022-
007A
11
0.01122
MCE Filter,
0.8|jm
25
385
210.7
1
Low
A0611022-
009A
10
0.01122
MCE Filter,
0.8|jm
25
385
285.2
1
Low
A0611022-
011A
10
0.01122
MCE Filter,
0.8|jm
25
385
374.3
2
High
A0611023-
001A
8
0.01122
MCE Filter,
0.45|jm
25
385
1114.1
2
High
A0611023-
002A
8
0.01122
MCE Filter,
0.45|jm
25
385
1192.1
2
High
A0611023-
004A
10
0.01122
MCE Filter,
0.45|jm
25
385
748.7
2
High
A0611023-
005A
10
0.01122
MCE Filter,
0.45|jm
25
385
944.7
2
High
A0611023-
009A
10
0.01122
MCE Filter,
0.45|jm
25
385
935.8
2
High
A0611023-
01 OA
10
0.01122
MCE Filter,
0.45|jm
25
385
811.1
2
High
A0611023-
003A
10
0.01122
MCE Filter,
0.8|jm
25
385
748.7
2
High
A0611023-
006A
10
0.01122
MCE Filter,
0.8|jm
25
385
713.0
2
High
A0611023-
007A
10
0.01122
MCE Filter,
0.8|jm
25
385
641.7
2
High
A0611023-
008A
10
0.01122
MCE Filter,
0.8|jm
25
385
606.1
2
High
A0611023-
011A
10
0.01122
MCE Filter,
0.8|jm
25
385
980.4
2
High
A0611023-
012A
10
0.01122
MCE Filter,
0.8|jm
25
385
891.3

-------
Batch
Loading
Sample ID
Grid
Openings
Counted
Grid
Opening
Size
(mm)
Filter Type
(MCE =
mixed
cellulose
ester)
Size
(mm)
Effective
Filter
Area
(mm2)
Total
Chrysotile
Asbestos
Structures
(s/mm2)
2
High
A0611023-
013A
10
0.01122
MCE Filter,
0.8|jm
25
385
641.7
2
High
A0611023-
014A
10
0.01122
MCE Filter,
0.8|jm
25
385
730.8
2
High
A0611023-
015A
10
0.01122
MCE Filter,
0.8|jm
25
385
935.8
2
High
A0611023-
016A
10
0.01122
MCE Filter,
0.8|jm
25
385
650.6
2
High
A0611023-
017A
10
0.01122
MCE Filter,
0.8|jm
25
385
730.8
2
High
A0611023-
018A
10
0.01122
MCE Filter,
0.8|jm
25
385
650.6
3
Low
A0611025-
001A
10
0.01122
MCE Filter,
0.45|jm
25
385
463.5
3
Low
A0611025-
002A
10
0.01122
MCE Filter,
0.45|jm
25
385
436.7
3
Low
A0611025-
003A
10
0.01122
MCE Filter,
0.45|jm
25
385
392.2
3
Low
A0611025-
004A
10
0.01122
MCE Filter,
0.45|jm
25
385
436.7
3
Low
A0611025-
005A
10
0.01122
MCE Filter,
0.45|jm
25
385
311.9
3
Low
A0611025-
007A
10
0.01122
MCE Filter,
0.45|jm
25
385
436.7
3
Low
A0611025-
006A
10
0.01122
MCE Filter,
0.8|jm
25
385
267.4
3
Low
A0611025-
008A
10
0.01122
MCE Filter,
0.8|jm
25
385
445.6
3
Low
A0611025-
009A
10
0.01122
MCE Filter,
0.8|jm
25
385
472.4
3
Low
A0611025-
01 OA
10
0.01122
MCE Filter,
0.8|jm
25
385
499.1
3
Low
A0611025-
10
0.01122
MCE Filter,
25
385
383.2

-------
Batch
Loading
Sample ID
Grid
Openings
Counted
Grid
Opening
Size
(mm)
Filter Type
(MCE =
mixed
cellulose
ester)
Size
(mm)
Effective
Filter
Area
(mm2)
Total
Chrysotile
Asbestos
Structures
(s/mm2)


011A


0.8|jm



3
Low
A0611025-
012A
10
0.01122
MCE Filter,
0.8|jm
25
385
383.2
3
Low
A0611025-
013A
10
0.01122
MCE Filter,
0.8|jm
25
385
267.4
3
Low
A0611025-
014A
10
0.01122
MCE Filter,
0.8|jm
25
385
508.0
3
Low
A0611025-
015A
10
0.01122
MCE Filter,
0.8|jm
25
385
463.5
3
Low
A0611025-
016A
10
0.01122
MCE Filter,
0.8|jm
25
385
356.5
3
Low
A0611025-
017A
10
0.01122
MCE Filter,
0.8|jm
25
385
303.0
3
Low
A0611025-
018A
10
0.01122
MCE Filter,
0.8|jm
25
385
303.0
4
High
A0611028-
001A
9
0.01122
MCE Filter,
0.45|jm
25
385
1029.9
4
High
A0611028-
002A
5
0.01122
MCE Filter,
0.45|jm
25
385
1836.0
4
High
A0611028-
003A
7
0.01122
MCE Filter,
0.45|jm
25
385
1311.4
4
High
A0611028-
007A
9
0.01122
MCE Filter,
0.45|jm
25
385
1029.9
4
High
A0611028-
008A
8
0.01122
MCE Filter,
0.45|jm
25
385
1225.5
4
High
A0611028-
01 OA
9
0.01122
MCE Filter,
0.45|jm
25
385
1099.2
4
High
A0611028-
011A
5
0.01122
MCE Filter,
0.45|jm
25
385
1978.6
4
High
A0611028-
012A
5
0.01122
MCE Filter,
0.45|jm
25
385
2014.3
4
High
A0611028-
014A
5
0.01122
MCE Filter,
0.45|jm
25
385
1853.8

-------
Batch
Loading
Sample ID
Grid
Openings
Counted
Grid
Opening
Size
(mm)
Filter Type
(MCE =
mixed
cellulose
ester)
Size
(mm)
Effective
Filter
Area
(mm2)
Total
Chrysotile
Asbestos
Structures
(s/mm2)
4
High
A0611028-
015A
6
0.01122
MCE Filter,
0.45|jm
25
385
1738.0
4
High
A0611028-
016A
6
0.01122
MCE Filter,
0.45|jm
25
385
1515.2
4
High
A0611028-
017A
7
0.01122
MCE Filter,
0.45|jm
25
385
1515.2
4
High
A0611028-
004A
8
0.01122
MCE Filter,
0.8|jm
25
385
1214.3
4
High
A0611028-
005A
7
0.01122
MCE Filter,
0.8|jm
25
385
1286.0
4
High
A0611028-
006A
6
0.01122
MCE Filter,
0.8|jm
25
385
1515.2
4
High
A0611028-
009A
10
0.01122
MCE Filter,
0.8|jm
25
385
873.4
4
High
A0611028-
013A
9
0.01122
MCE Filter,
0.8|jm
25
385
1099.2
4
High
A0611028-
018A
5
0.01122
MCE Filter,
0.8|jm
25
385
1836.0

-------
16
16
16
16
16
16
16
16
16
16
16
16
4
4
4
4
Loading
Sample ID
Grid
Openings
Counted
Grid Opening
Size (mm)
Filter
Type
(MCE =
mixed
cellulose
ester)
A0611024-
001A
0.0112
MCE Filter
0.45|jm
25
A0611024-
002A
0.0112
MCE Filter
0.45|jm
25
A0611024-
003A
0.0112
MCE Filter
0.45|jm
25
A0611024-
004A
0.0112
MCE Filter
0.45|jm
25
A0611024-
005A
0.0112
MCE Filter
0.45|jm
25
A0611024-
006A
0.0112
MCE Filter
0.45|jm
25
A0611024-
007A
0.0112
MCE Filter
0.45|jm
25
A0611024-
008A
0.0112
MCE Filter
0.45|jm
25
A0611024-
009A
0.0112
MCE Filter
0.45|jm
25
A0611024-
01 OA
0.0112
MCE Filter
0.45|jm
25
A0611024-
011A
0.0112
MCE Filter
0.45|jm
25
A0611024-
012A
0.0112
MCE Filter
0.45|jm
25
A0611026-
001A
0.0112
MCE Filter
0.45|jm
25
A0611026-
002A
10
0.0112
MCE Filter
0.45|jm
25
A0611026-
003A
10
0.0112
MCE Filter
0.45|jm
25
A0611026-
004A
0.0112
MCE Filter
0.45|jm
25

-------
4
4
4
4
4
4
4
4
2
2
2
2
2
2
2
2
Loading
Sample ID
Grid
Openings
Counted
Grid Opening
Size (mm)
Filter
Type
(MCE =
mixed
cellulose
ester)
A0611026-
005A
0.0112
MCE Filter
0.45|jm
25
A0611026-
006A
0.0112
MCE Filter
0.45|jm
25
A0611026-
007A
0.0112
MCE Filter
0.45|jm
25
A0611026-
008A
0.0112
MCE Filter
0.45|jm
25
A0611026-
009A
0.0112
MCE Filter
0.45|jm
25
A0611026-
01 OA
10
0.0112
MCE Filter
0.45|jm
25
A0611026-
011A
0.0112
MCE Filter
0.45|jm
25
A0611026-
012A
0.0112
MCE Filter
0.45|jm
25
A0611027-
001A
10
0.0112
MCE Filter
0.45|jm
25
A0611027-
002A
10
0.0112
MCE Filter
0.45|jm
25
A0611027-
003A
0.0112
MCE Filter
0.45|jm
25
A0611027-
004A
0.0112
MCE Filter
0.45|jm
25
A0611027-
005A
0.0112
MCE Filter
0.45|jm
25
A0611027-
006A
0.0112
MCE Filter
0.45|jm
25
A0611027-
007A
10
0.0112
MCE Filter
0.45|jm
25
A0611027-
008A
0.0112
MCE Filter
0.45|jm
25

-------
2
2
2
2
8
8
8
8
8
8
8
8
8
8
8
8
Loading
Sample ID
Grid
Openings
Counted
Grid Opening
Size (mm)
Filter
Type
(MCE =
mixed
cellulose
ester)
A0611027-
009A
0.0112
MCE Filter
0.45|jm
25
A0611027-
01 OA
0.0112
MCE Filter
0.45|jm
25
A0611027-
011A
0.0112
MCE Filter
0.45|jm
25
A0611027-
012A
0.0112
MCE Filter
0.45|jm
25
A0611028-
001A
0.0112
MCE Filter
0.45|jm
25
A0611028-
002A
0.0112
MCE Filter
0.45|jm
25
A0611028-
003A
0.0112
MCE Filter
0.45|jm
25
A0611028-
007A
0.0112
MCE Filter
0.45|jm
25
A0611028-
008A
0.0112
MCE Filter
0.45|jm
25
A0611028-
01 OA
0.0112
MCE Filter
0.45|jm
25
A0611028-
011A
0.0112
MCE Filter
0.45|jm
25
A0611028-
012A
0.0112
MCE Filter
0.45|jm
25
A0611028-
014A
0.0112
MCE Filter
0.45|jm
25
A0611028-
015A
0.0112
MCE Filter
0.45|jm
25
A0611028-
016A
0.0112
MCE Filter
0.45|jm
25
A0611028-
017A
0.0112
MCE Filter
0.45|jm
25

-------
&EPA
United States
Environmental Protection
Agency
(8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
EPA 600/R-08/046
November 2008
www.epa.gov
Recycled/Recyclable Printed with
vegetable-based ink on paper that
contains a minimum of
50% post-consumer fiber content
processed chlorine free

-------