EPA/600/R-94/134
METHOD 100.2
DETERMINATION OF ASBESTOS STRUCTURES OVER
10 im IN LENGTH IN DRINKING WATER
Kim A. Brackett, Ph.D. (IT Corp.)
Patrick 0. Clark (Risk Reduction Engineering Laboratory)
U.S. Environmental Protection Agency
James R. Millette, Ph.D. (MVA, Inc.)
June 1994
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
-------
METHOD 100.2
DETERMINATION OF ASBESTOS STRUCTURES OVER
10 im IN LENGTH IN DRINKING WATER
1.0 SCOPE AND APPLICATION
1.1 This test method is recommended for the determination of the
presence and quantitation of asbestos structures in drinking water
samples. The method allows for the quantitation of structures
greater than 10 0m in length.
O
1.2 This test method describes the equipment and procedures necessary
for the sampling and analysis of drinking water by transmission
electron microscopy (TEM).
1.3 The identification of asbestos by TEM is based on (a) morphology,
(b) selected area electron diffraction (SAED) and (c) energy
dispersive x-ray analysis (EDXA).
1.4 Applicable analytes and Chemical Abstract Service (CAS) Numbers:
Asbestos CAS Number
Chrysotile
Crocidolite
Amosite (Grunerite)
Anthophyllite
Tremolite
Actinolite
1.5 Data Quality Objectives
1.6
1.7
12001-29-5
12001-28-4
12001-73-5
77536-67-5
77536-68-6
77536-66-4
Method
TEM
Accuracy3
95%
Precision3
95%
Completeness
100%
a.
Confidence coefficient of a confidence interval for a
Poisson variable within which count ranges are expected to
fall.
Analytical Sensitivity. A sensitivity of 200,000 fibers per liter
(0.2 MFL) is required unless filter loading satisfies the stopping
rules in Sect. 11.31. See TABLE 1.
Only asbestos structures meeting the definitions set forth in the
Chatfield protocol are counted (1).
-------
TABLE 1. Limitation of Analytical Sensitivity by Volume
of Water Sample Filtered (1)
Volume of Sample Filtered (mL)
25 mm Diam. Filterb
0.1
0.5
1.0
2.0
5.0
10
15
25
50
100
47 mm Diam. Filter0
0.6
2.8
5.7
11
28
57
85
142 •
285
570
Analytical
Sensitivity (f/L)a
3.0 x 107
6.0 x 106
3.0 x 106
1.6 x 106
6.0 x 10s
3.0 x 105
2.0 x 105 "
1.2 x 103
6.0 x 104
3.0 x 104
* Concentration corresponding to 1 structure detected in
10 grid openings of approximately 0.008 mm2.
b Assuming active filter area of 1.99 cm2.
c Assuming active filter area of 11.34 cm2.
2.0 SUMMARY OF METHOD
2.1 Water is collected in a polyethylene or glass container and
shipped to the laboratory. Known aliquots of the sample are
filtered through a 0.1 to 0.22 urn pore mixed cellulose ester
(MCE). A carbon extraction replica is prepared from a portion of
the filter and is examined in the TEM at a magnification of 10,000
to 20,OOOX. Asbestos structures are identified by morphology,
selected area electron diffraction (SAED) and energy dispersive
x-ray analysis (EDXA). Structures are classified according to the
counting rules specified in the Chatfield polycarbonate filter
protocol (1). Only asbestos structures greater than 10 /zm in
length are counted. Some states may require identification and
measurement of all asbestos fibers, regardless of size. In this
case the use of a 0.1 [im pore-size polycarbonate or MCE filter
membrane is necessary to prevent loss of small fibers during
filtration.
-------
3.0 DEFINITIONS
3.1 ANALYTICAL SENSITIVITY — The waterborne concentration represented
by the finding of one asbestos structure in the total area of
filter examined. This value will depend on the fraction of the
sample filtered and the dilution factor (if applicable).
3.2 ASBESTIFORM — A specific type of fibrous habit which has greater.
flexibility and higher tensile strength than other habits of the
same mineral.
3.3 ASBESTOS — Generic term for a group of hydrated mineral
silicates.
3.4 ASPECT RATIO -- The ratio of the length of a fibrous particulate
to its apparent width (equivalent diameter).
3.5 BUNDLE — A structure composed of three or more fibers in a
parallel arrangement with each fiber lying less than one diameter
apart.
3.6 CLUSTER ~ A structure with fibers in a random arrangement such
that all fibers are intermixed and no single fiber is isolated
from the group.
3.7 EDXA — Energy dispersive X-ray analysis.
3.8 FIBER -- For the purposes of this method, a structure having a
minimum length of 10 pm and an aspect ratio (length to width) of
3:1 or greater with substantially parallel sides.
3.9 FIBRIL — The smallest crystalline fiber that can be separated
from a fiber bundle which cannot be subdivided without losing its
fibrous properties.
3.10 GRID — A 3 mm diameter 200-mesh copper lattice used to hold the
carbon extraction replica for observation in the TEM.
3.11 INTERSECTION -- Nonparallel touching or crossing of fibers, with
the projection having an aspect ratio >3:1.
3.12 MATRIX — Fiber or fibers with one free end and the other end
embedded in or hidden by a particulate.
3.13 MFL -- Million fibers per liter.
3.14 SAED or ED -- Selected area electron diffraction.
3.15 STRUCTURE — A microscopic bundle, cluster, fiber or matrix which
may contain asbestos.
3.16 TEM — Transmission electron microscope.
-------
4.0 INTERFERENCES
The minerals listed below can exhibit morphological, chemical or crystal
structure similarities to the asbestos minerals. The laboratory QA/QC
manual should describe routine techniques to differentiate them from
asbestos. In general, these techniques should be the same as those
required for accreditation by the National Institute of Standards and
Technology/National Voluntary Laboratory Accreditation Program
(NIST/NVLAP) for airborne asbestos.
4.1 Antigorite
4.2 Attapulgite (Palygorskite)
4.3 Halloysite
4.4 Horneblende
4.5 Pyroxenes
4.6 Sepiolite
4.7 Vermiculite scrolls
5.0 SAFETY
This test method may involve hazardous materials, operations and
equipment, and does not purport to address all of the safety problems,
if any, associated with its use. It is the responsibility of the user
of this method to establish appropriate safety and health practices and
determine the applicability of regulatory limitations prior to use.
Sample filtration should take place in a clean HEPA filtered positive
pressure hood to avoid possible contamination of the preparation.
Collapsing of the filter should be performed in an exhaust hood.
6.0 EQUIPMENT AND SUPPLIES
6.1 Transmission Electron Microscope capable of performing electron
diffraction, with a fluorescent screen inscribed with a calibrated
measuring scale. The TEM must have EDXA and be able to produce a
spot size, at crossover, less than 250 nm in diameter.
6.2 Energy dispersive X-ray analyzer
6.3 High vacuum carbon evaporator with rotating stage
6.4 Positive pressure HEPA filtered hood
6.5 Fume hood
6.6 Table-top low power ultrasonic bath
-------
6.7 Ozone generator capable of generating at least 400g of ozone per
day at a concentration of 1% by weight when supplied with dry
oxygen.
6.8 Quartz pipets
6.9 Submersible UV lamp (254 fun wavelength)
6.10 Waterproof marker
6.11 Forceps (tweezers)
6.12 Graduated pipettes (1, 5, 10 ml sizes), disposable glass
6.13 25 or 47 mm diameter filter funnel assembly (either glass or
disposable plastic)
6.14 1000 ml side arm vacuum filtration flask
6.15 25 or 47 mm diameter mixed cellulose ester (MCE) membrane filters
(<0.22 fan and 5 /an pore size)
6.16 Disposable petri dishes (or suitable equivalent) for storage of
filtration membranes
6.17 Glass microscope slides
6.is Curved scalpel blades
6.19 Low temperature oven or cabinet-type desiccator
6.20 Low temperature plasma asher
6.21 Jaffe washer
6.22 200 mesh copper TEM finder grids
6.23 Carbon rods
6.24 1000 mL glass or polyethylene sample bottles with screw-on caps
7.0 REAGENTS AND STANDARDS
7.1 Reagent grade chemicals shall be used in all tests. Unless
otherwise indicated, it is intended that all reagents shall
conform to the specifications of the Committee on Analytical
Reagents of the American Chemical Society (2).
7.2 Deionized particle-free water
7.3 Acetone
7.4 Dimethylformamide (DMF)
-------
7.5 Glacial acetic acid
7.6 Chloroform
7.7 1-methyl-2-pyrrolidone also called l-methyl-2-pyrolidinone or
n-methyl-2-pyrrolidone
(CAS 872-50-4)
7.8 NIST traceable asbestos standards
7.9 Laboratory standards for the interference minerals listed in Sect.
4.0.
8.0 SAMPLE COLLECTION. PRESERVATION AND STORAGE
8.1 The sample container will be an unused, pre-cleaned, screw-capped
bottle of glass or low density (conventional) polyethylene of at
least 1 liter capacity. It is recommended that the use of
polypropylene bottles be avoided since problems of particulate
being released into water samples have been observed. Before use,
the bottles should be rinsed twice by filling approximately one-
third full with fiber-free water and shaking vigorously for 30
seconds. After discarding the rinse water, the bottles should
then be filled with fiber-free water and treated in an ultrasonic
bath for 15 minutes, followed by several rinses with fiber-free
water.
8.2 Blank determinations should be made prior to sample collection.
When using polyethylene bottles, one bottle from each batch, or a
minimum of one from each 24, should be tested for background
level. When using glass bottles, four bottles from each 24 should
be tested. Additional blanks may be desirable when sampling
waters suspected of containing very low levels of asbestos, or
when additional confidence in the bottle blanks are desired. An
acceptable bottle blank level is defined as <0.01 MFL >10 jam.
8.3 SAMPLE COLLECTION — It is beyond the scope of this procedure to
furnish detailed instructions for field sampling; the general
principles of obtaining water samples apply. If tests are being
made of drinking water in a bulk storage supply, there may be a
vertical distribution of particle sizes. If a representative
sample of the water supply is required, a carefully designated set
of samples should be taken representing the vertical as well as
the horizontal distribution and then composited for analysis.
Compositing must be done in the laboratory, and not in the field.
When sampling from a distribution system a commonly used faucet
should be chosen. Remove all hoses or fittings from the faucet
and allow the water to run to waste for a sufficiently long period
to ensure that the sample collected is representative of the fresh
water supply. For most buildings this may be indicated by a
change in temperature of the water at the faucet. Faucets or
valves should not be adjusted until all samples have been
-------
collected. Samples should not be taken from hydrants or other
faucets at the deadend of a distribution system.
As an additional precaution against contamination, each bottle may
be rinsed several times in the source water being tested before
the final sample is taken. This procedure is not suitable when
taking depth samples from a storage tank, however.
8.4 QUANTITY OF SAMPLE — Two samples of approximately 800 ml should
be collected from each site. No preservatives should be added
during sampling. This sample volume will leave an air space to
allow efficient redispersal of settled material, by shaking,
before analysis. Each bottle should be labeled with the date,
time, place, field sample number and sampler's name using a
waterproof marker.
8.5 SAMPLE SHIPMENT — Samples must be transported to the analytical
laboratory as soon as possible after collection. The samples
should be shipped in a sealed container, separate from any bulk or
air samples. Samples should be shipped in a cooler with ice to
retard bacterial or algal growth in the samples. Do not freeze
the samples. Samples should be received and filtered in the
laboratory within 48 hours of collection. Samples must be
accompanied by a properly executed chain of custody document.
8.6 SAMPLE PRESERVATION — Samples should be filtered immediately
after arrival at the laboratory or stored in a refrigerator until
filtered,
8.7 SAMPLE COMPOSITING — Up to five samples may be composited after
receipt in the laboratory. The composite sample must be prepared
from the individual samples within 48 hours of collection, or, if
the samples have been stored for more than 48 hours, they must be
individually treated with Og-UV in the original containers.
Samples should be sonicated in the original container and equal
amounts extracted to make up the composite. It may also be
prudent to filter an aliquot of each sample for analysis in case
the composite sample exceeds l/5th the MCL (1.4 MFL >10 pi long).
If, later, the original samples are to be filtered separately,
they must be treated again with 03-UV in the original containers
and resonicated..
9.0 QUALITY CONTROL
9.1 The quality control checks required for this method generally
follow those specified in the Federal Register for AHERA analysis
of air samples (3) and the NISTIR document relating to airborne
asbestos analysis (4). These requirements are summarized in TABLE
2. The criterion for acceptability of bottle and process blanks
is <0.01 MFL >10 urn in length.
-------
10.0 CALIBRATION AND STANDARDIZATION
10.1 MAGNIFICATION CALIBRATION — Magnification calibration must be
done at the fluorescent screen and must be performed at the
magnification used for fiber counting, generally 10,000 and
20,OOOX. Calibration is performed using a grating replica (e.g.,
one containing at least 2,160 lines/mm). Define a field of view
on the fluorescent screen either by markings or physical
boundaries. The field of view must be measurable or previously
inscribed with a scale or concentric circles (all scales should be
metric). If the instrument contains a tilting stage (goniometer),
the z-axis must be adjusted to ensure that the stage is in the
eucentric position prior to performing any measurements. A
logbook must be maintained with the dates of the calibration
recorded. Frequency of calibration will depend on the service
history of the instrument. It is recommended that calibrations be
performed monthly to establish the stability of the magnification.
Also, the calibration should be checked following any maintenance
of the microscope involving adjustment of the lens or high voltage
power supplies or the disassembly of the electron optical column
apart from filament exchange.
10.2 CAMERA CONSTANT ~ The camera length of the TEM in the electron
diffraction (ED) mode must be calibrated before ED patterns of
unknown samples are observed. This can be achieved by using a
carbon-coated grid on which a thin film of gold has been sputtered
or evaporated. A thin film of gold can also be evaporated on the
specimen grids to obtain ED patterns superimposed on the ring
• pattern from the polycrystalline gold film* In practice, it is
desirable to optimize the thickness of the gold film so that only
one or two sharp rings are obtained on the superimposed ED
pattern. Thicker gold films will tend to mask weaker diffraction
spots from the fibrous particulates. Since the unknown d-spacings
of most interest in asbestos analysis are those which lie closest
to the transmitted beam, multiple gold rings from thicker films
are unnecessary. Alternatively, a gold standard specimen can be
used to obtain an average camera constant on a regular basis for
each TEM in the laboratory. The stage must be at the eucentric
position for this calibration. The camera constant calculated for
that particular instrument can then be used for ED patterns of
unknowns taken during the corresponding period.
10.3 SPOT SIZE — The diameter of the smallest beam spot at crossover
must be measured regularly. Photograph the beam at crossover at
20,000 to 25,OOOX at a short exposure setting (to avoid spreading
of the exposed spot on the film). Measure the diameter on the
negative and divide by the magnification used. The resulting
figure must be less than 250 nm.
8
-------
TABLE 2. SUMMARY OF LABORATORY DATA QUALITY OBJECTIVES
Unit Operation
Sample Receiving
Sample Prep
Sample Analysis
Performance
or 10%
QC Check
Review chain-of-
custody and
sampling data
Supplies and
reagents
Grid opening
size
Clean area
monitoring
Lab blank
Etcher blank
23 grids/sample
System check
Alignment check
Mag calibration
Camera constant
EDS Cu Ka line
Lab blank
Frequency
Expectation
Each sample
On receipt
20/20 grids/
lot of 1000
or I/sample
After service
I/series or
10%
1/20 samples
Each sample
Each day
Each day
Monthly or
after service
Weekly
Each day
1/prep series
Conformanee
95% complete
Meet specs.
100%
Meet specs.
Meet specs.
75%
>70% intact
openings
Each day
Each day
95%
95%
95%
Meet specs.
Replicate count
Duplicate count
Analysis of
standards
Analysis of SRM
Data entry
Record and verify
SAED patterns
>1/100 samples l.SXPoisson
std. dev.
>1/100 samples 2XPoisson
std. dev.
Training and
comparison with
, unknowns
1 per analyst
per year
Each sample
1/5 samples
100%
Calculations
Hand calculation of 1/100 samples
automated data
reductions or 2nd
analyst check of
manual calcs.
l.SXPoisson
std. dev.
95%
80%
accuracy
85%
-------
10.4 EDXA SYSTEM — The resolution and calibration of the EDXA must be
checked at least monthly and after service. Initially, the system
is calibrated by using two reference elements to calibrate the
energy scale of the instrument according to the manufacturer's
instructions. This can easily be done using a carbon-coated
copper grid upon which a thin film of aluminum has been evap-
orated. The Al and Cu Ka peaks should be centered at 1.48 KeV and
8.04 KeV respectively. The deviation from these energies should
be no more than ±10eV. The ability of the system to resolve the
sodium Ka line from the copper L line should be confirmed by
obtaining a spectrum from a standard crocidolite sample on a
copper grid. Additional resolution checks are usually found in
the manufacturer's instruction manual. The k-factors relative to
silicon should be calculated for Na, Mg, Al, Si, Ca and Fe using
NIST SRM 2063. The k-factor for Mg to Fe must be calculated; a
value of 1.5 or less is required. EDXA spectra should be obtained
from NIST traceable standards and kept on file in the laboratory
for comparison with published, standard spectra and unknown
spectra.
11.0 PROCEDURE
11.1 If the water samples are suspected to contain high levels of
organic contaminants, or have been stored for periods longer than
48 hours, oxidation of the organics by the ozone-ultraviolet
technique (1) may be necessary. Details for this procedure appear
in Appendix 1.
11.2 Wet-wipe the exterior of the sample bottle before entering the
clean area used for specimen preparation. The sample prep area
should be separate from the areas used for bulk sample or air
sample preparation to avoid possible cross-contamination. Sample
filtration should take place in a positive pressure HEPA filtered
hood.
11.3 The use of vertical sided 25 or 47 mm diameter glass filtration
systems with a sintered glass frit support is recommended to avoid
loss of fibers by settling on tapered sides of the reservoir.
Disposable plastic funnel assemblies may be substituted for glass
apparatus. A few precautions must be taken with reusable
glassware to ensure optimum sample preparations. All glassware
should be carefully washed in a detergent solution with a brush
before each use and rinsed several times in fiber-free water. Any
glassware that has contained asbestos in solution should be placed
in soapy water and scrubbed before it has had the opportunity to
dry. Sonication in a detergent solution is also recommended.
Frequent blanks should be run with fiber-free water to check
cleanliness of the apparatus.
11.4 Unwrap an unused disposable plastic filter funnel unit and remove
the tape around the base of the funnel. Remove the funnel and
discard the top filter membrane supplied with the unit. Do not
remove the coarse polypropylene support pad. Assemble the unit
with the adapter and a properly sized neoprene one-hole stopper,
and attach the assembly to the 1000 ml vacuum flask.
10
-------
11.5 Whether using a glass or plastic filtration unit, care must be
taken to ensure that the filter support and the mating surface of
the filtration base not be damp when the backing filter is placed
on the support. If this should occur, the vacuum across the
nitration surface will be uneven, resulting in uneven distribu-
tion of the filtered particulates. Either the filtration base
must be thoroughly dried before use, or completely wetted so that
the backing filter and filtration membrane are uniformly wet
before filtration is started.
11.6 DISPOSABLE FILTER UNIT - Wet the support pad with distilled
water, if desired, and place a 5 urn pore-size MCE backing filter
on the pad Place a <0.22 urn pore-size MCE filter membrane on the
backing filter. Ensure that both filter membranes are completely
wet, or dry depending upon the technique preferred. Apply a
vacuum to the flask and ensure that the filters are centered and
pulled flat with no entrapped air bubbles between the membranes.
If any irregularities are seen on the filter surface, discard the
filters and try again. Replace the funnel and seal the assembly
with tape.
11.7 REUSABLE GLASS FILTER UNIT
11.7.1 (Dry Filter) Apply vacuum to the flask and leave the
vacuum on until the filtration process is completed.
Place a <5 /«n pore-size MCE filter on the glass frit to
serve as a backing filter. Be sure that the filter is
not creased during installation. If the filter appears
to absorb water, it should be discarded and the frit
rinsed with methanol to speed drying. Place a <0.22 urn
pore-size MCE filter membrane on the backing filter
Make sure that the mating surface of the filter
reservoir is dry, then place it on the filters and clamp
in place. r
11.7.2 (Wet Filter) Wet the glass frit and place a 5 jum pore-
size MCE filter on the frit to serve as a backing
filter. Place a <0.22 /on pore-size MCE filter membrane
on the backing filter. Make sure that both membranes
are thoroughly soaked with water. Install the filter
reservoir and clamp in place.
11.8 A process blank sample consisting of fiber-free water should be
run before the first field sample. The quantity of water should
be >10 mL for a 25 mm diameter filter and >50 mL for a 47 mm
diameter filter.
11.9 Shake the capped sample bottle vigorously by hand and place in a
low power ultrasonic bath for 15 minutes. Shake the sample by
hand again before removal of any aliquots.
11.10 Remove the desired aliquot from the original sample. Large
volumes may be measured with a graduated cylinder, smaller
volumes should be taken with disposable glass pipettes. Samples
taken by pipette should be taken from the vertical center of the
original sample. No aliquot less than 1 mL should be taken from
11
-------
the original sample. The minimum volume that should be filtered
is 10 ml for a 25 mm diameter filter, or 50 ml for a 47 mm
diameter filter. If it is necessary to filter aliquots less
than these volumes, the aliquots should be brought up to these
levels with fiber-free water and shaken vigorously before
filtration. Obtaining a filter with the proper loading is a
matter of trial and error. It -is best to filter several volumes
of the sample. Samples with high particulate or asbestos
content may require volumes less than 1 ml to be filtered. In
this case, preparation of serial dilutions must be performed.
11.11 Agitate the aliquot and pour into the filter reservoir. Apply
vacuum to the flask. When volumes larger than the capacity of
the reservoir have to be filtered, the additional solution
should be carefully added while the reservoir is over half full
to avoid disturbing the particulates already deposited on the
filter. Do not rinse the sides of the reservoir.
11.12 If a reusable glass reservoir is used, immediately remove the
reservoir and place in soapy water.
11.13 Disassemble the filtration apparatus and remove the filters with
clean forceps. Carefully separate the working filter from the
backing filter and discard the backing filter. Place the
working filter in a pre-cleaned disposable petri dish and cover.
11.14 Allow filter to dry. Drying may take place in a HEPA filtered
hood, an asbestos-free oven, or a cabinet type desiccator. The
cover of the petri dish should be opened slightly to allow water
vapor to escape.
11.15 Using a clean scalpel remove a portion of the dry filters for
preparation of TEM grids by the direct transfer technique. Be
sure to avoid the outer ring of the filter that was covered by
the mating surface of the reservoir. Transfer the removed
portion to an unused petri dish.
11.16 A portion of an unused filter should also be prepared as a lot
blank.
11.17 MCE filters must be collapsed prior to low temperature etching.
Use of either the DMF-Acetic acid method or the acetone method
is acceptable. Samples should be transferred to an exhaust hood
for this step.
11.17.1 DMF-ACETIC ACID METHOD -- Place a drop of the clearing
solution (35% dimethylformamide [DMF], 15% glacial
acetic acid and 50% water [V/V] on a clean microscope
slide. Use just enough solution to saturate the
filter. DMF is a relatively toxic solvent and should
be used in a fume hood. Carefully lay the filter
segment, sample surface upward, on top of the solution,
bringing the filter and solution together at an angle
of about 20° to help exclude air bubbles. Remove any
solution not absorbed by the filter with lens paper.
Label the slide with a glass scribing tool or a
12
-------
permanent marker. Place the slide into an oven, or on
the surface of a hot plate, at 65-70°C for 10 to 30
minutes.
11.17.2 ACETONE METHOD — Place the filter section on a clean
microscope slide. Affix the filter section to the
slide with tape around the edges. Label the slide with
a glass scribing tool or permanent marker. Place the
slide in a petri dish containing several paper filters
soaked with acetone. Cover the dish and wait for the
sample to fuse and clear (approximately 5 minutes).
11.17.3 Plasma etching of the collapsed filter is required if
0.22 fm pore-size membrane filters are used. Plasma
etching is optional (but recommended) with 0.1 urn pore-
size filters. The microscope slide to which the
collapsed filter pieces are attached is placed in a low
temperature plasma asher. Because plasma ashers vary
greatly in their performance, both from unit to unit
and between different positions in the asher barrel, it
is difficult to specify the conditions that should be
used. Insufficient etching will result in a failure to
expose embedded fibrils, and too much etching may
result in loss of particulate from the surface. It is
recommended that conditions be used which will remove
about 10% of the filter mass. Additional information
about calibration of the plasma asher can be found in
the AHERA (3) and NISTIR (4) documents.
11.18 CARBON-COATING FILTER SEGMENTS
11.18.1 Coating must be performed with a high-vacuum evap-
oration unit equipped with a rotating tilting stage.
Units based on evaporation of carbon filaments in a
vacuum generated only by an oil rotary pump have not
been evaluated for this application and must not be
used. The carbon rods should be sharpened by a carbon
rod sharpener to necks of about 4 mm long and 1 mm in
diameter. The rods are installed in the evaporator in
such a manner that the points are approximately 10 cm
from the surface of the microscope slide.
11.18.2 Affix the glass slide to the rotating tilting table and
evacuate the chamber to a pressure of <0.013 Pa. The
evaporation must be performed in very short bursts
separated by some seconds to allow the electrodes to
cool to avoid overheating the surface of the filter
An experienced analyst can judge the thickness of the
carbon film to be applied, and some tests should be
made first on unused filters. If the film is too thin,
there will be few complete, undamaged grid openings on
the specimen and large particles may be lost. If the
coating is too thick, the TEM image will lack contrast
and the ability to obtain selected area electron
diffraction (SAED) patterns will be compromised. A
13
-------
carbon film that appears similar to a 15% gray scale is
usually adequate.
11.19 Prepare the Jaffe washer. The precise design of the Jaffe
washer is not important, so any one of the published designs may
be used (3,5). Place the washer in a fume hood and fill with
DMF, acetone or l-methyl-2-pyrrolidone to the level of the
screen on which the samples will be placed.
11.20 Place calibrated TEM grids in the Jaffe washer. Indexed
("Finder") grids or grids with a unique center mark should be
used. The area of the grid square openings must be determined
either by using the TEM at a calibrated magnification low enough
to measure the sides of the opening, or with phase contrast
microscopy at a calibrated magnification (usually 400X). If the
measurements are made by TEM, one grid opening per grid is
measured. For measurement by phase contrast, 20 grid openings
are measured on each of 20 grids and the average size calcu-
lated. The TEM grids are first placed on a piece of lens tissue
or filter paper so that individual grids can be picked up with
forceps. Grids should be placed on pieces of filter paper or on
individual screens and placed in the Jaffe washer. Three or
more grids should be prepared from each sample,
11.21 Using a clean curved scalpel blade, cut 3 mm square pieces of
the carbon coated filter from the glass slide. The point of the
scalpel should be placed on the filter and a rocking motion used
to cut the 3 mm square segments. Squares should be selected
from the center of the filter and at two points between the
outer periphery of the active surface and the center. The
excised filter segments are placed carbon-side up on the grids.
A map of the Jaffe washer should be drawn to keep track of the
samples. Place the lid on the Jaffe washer and allow to stand
until the filter is adequately dissolved (several hours).
11.22 Remove the grids from the Jaffe washer and allow to dry
thoroughly before placing them in marked grid storage boxes (or
other suitable containers).
11.23 Analyze the samples by TEM at an accelerating voltage of 80 to
120 kV and a screen magnification of 10,000 to 20.000X.
11.23.1 Use at least three grids from each filter to obtain the
necessary number of grid openings or structures to
reach the required analytical sensitivity.
11.23.2 Carefully load the grid into the sample holder. Orient
the grid so that the grid bars are parallel and
perpendicular to the long axis of the holder. This
orientation will align the grid bars with the X and Y
axes of the specimen translation controls.
11.23.3 Scan the grid at a magnification of 250 to 1000X to
determine its suitability for analysis. Reject the
grid if:
14
-------
• 1. Less than 70% of the grid openings covered by the
replica are intact.
2. The replica is folded or doubled.
3. The replica is too dark or has obviously visible
filter structure because of incomplete
dissolution.
4. The grid is too heavily loaded to obtain an
accurate count.
5. The distribution of structures on the grid is
obviously not uniform.
11.23.4 Reject individual grid openings with greater than 5% of
their areas occupied by holes or tears in the carbon
film. Reject openings with >25% covered by
particulates.
11.24 Examine the grid openings at 10-20,OOOX magnification.
11.25 Record the length and width of any grouping of particles in
which an asbestos fiber with an aspect ratio >3:1 and a lenqth
greater than 10 /tm is detected.
11.26 Asbestos structures will be classified as fibers according to
the following rules.
11.26.1 FIBER — A structure having a minimum length greater
than 10 fjan and an aspect ratio (length to width) of 3:1
or greater and substantially parallel sides without
rounded ends.
11.26.2 Count an asbestos bundle >10 pm long as a single fiber.
Assign a length equal to the maximum length of any
fiber within the bundle. If the bundle has stepped
sides assign a width equal to an estimate of the mean
width of the bundle.
11.26.3 Count a matrix as a single fiber if it contains a fiber
or fibers, meeting the length and aspect ratio require-
ments, with one free end and the other end embedded in
or hidden by a particulate. If two ends are visible
which appear to be the ends of a single fiber, the
distance between, the two ends is measured. If only one
end of the fiber is visible, the fiber will be assigned
a length equal to twice its visible length, except
where this would place the concealed end outside of the
particle. In this case, the length will be recorded as
the visible length plus the diameter of the portion of
the particle at the point of the fiber intersection.
If the structure is too complex to be dealt with in
this manner, record the overall dimensions of the
structure but do not include it in the fiber count.
15
-------
11.26.4 Count and record as single fibers the individual fibers
visible with a cluster as long as they meet the fiber
definition. If the aggregate is too complex, record
the overall dimensions but do not include it in the
fiber count.
11.26.5 Fibers which intersect the top and left sides of the
grid opening are counted and recorded as twice their
visible length. Fibers intersecting the bottom and
right sides of the grid opening are not recorded.
11.26.6 Count only one end of the fiber to avoid the possi-
bility of counting a single fiber more than once.
11.27 Structures classified as chrysotile must first be examined by
SAED. If the characteristic chrysotile ED pattern is observed,
the fiber will be counted. If no pattern is observed, or the
pattern is not distinctive, the fiber must be examined by
quantitative EDXA. If EDXA is characteristic of chrysotile, the
fiber will be counted. Chrysotile fibers identified by
morphology alone can be recorded but not counted towards the
regulatory limit. The analyst's count sheet must indicate the
method used to verify identity. A modified version of .the AHERA
count sheet may be used, which has columns to check off the
method of identification.
11.28 Structures which are suspected to be an amphibole must first be
examined by SAED. If a random orientation ED pattern with a
0.53 nm layer spacing is obtained, the fiber should be analyzed
by EDXA. If the elements and peak ratios of the spectra
correspond to those of a known amphibole, the fiber will be
counted. If the random orientation ED pattern cannot be
obtained, is incomplete or is not recognizable as a non-
amphibole pattern, but an EDXA spectrum corresponding to a known
amphibole is obtained, the fiber will be counted. Only fibers
classified in this manner or by single or double zone axis SAED
can be included in the regulated fiber count. The count sheet
should indicate method used for identification.
11.29 Record both a typical SAED pattern and x-ray spectrum for each
type of asbestos observed for each set of samples from the same
source, or a minimum of every fifth sample analyzed. Record the
micrograph number on the count sheet and attach the x-ray
spectrum to the back of the count sheet. If the x-ray spectrum
is stored on disk, record the file and disk number on the count
sheet.
11.30 Record NSD when no structures are detected in the grid opening.
11.31 STOPPING RULES — Counting can be stopped at the completion of
the grid opening in which an analytical sensitivity of 0.2 MFL
is reached, or at the completion of the grid opening which
contains the 100th structure, whichever occurs first. A minimum
of 4 grid openings must be analyzed even if this results in the
counting of more than 100 asbestos fibers over 10 fan in length.
16
-------
11.32 The grid openings examined must be drawn approximately equally
from the three grids used in the analysis.
11.33 After completion of analysis, remove the grid from the
microscope and replace in the labelled specimen storage box.
Sample grids must be stored for a minimum of three years from
the date of analysis. Sample filters may also be archived if
desired.
12.0 DATA ANALYSIS AND CALCULATIONS
12.1 Calculation of results. The concentration of asbestos in a
given sample is calculated using the following formula:
no str x efa x RD
GO x GOA x Vx 1000
where:
no str = number of asbestos fibers counted
efa = effective filter area of the sampling filter in mm2
RO = dilution ratio of original sample (if applicable)
60 = number of grid openings counted
GOA = area of grid openings in /im2
V = original volume of sample filtered in ml
12.2 The following information must be reported for each sample
analyzed:
12.2.1 Mean concentration of asbestos in million fibers per
liter. .
12.2.2 Upper and lower 95% confidence limits on the mean
concentration.
12.2.3 Aliquot used for analysis and dilution factor (if any).
12.2.4 Effective filter area.
12.2.5 Total area of filter examined.
12.2.6 Number of asbestos structures counted.
12.2.7 Analytical sensitivity.
12.2.8 Copies of the TEM count sheet, if requested.
17
-------
12.2.9 Number of structures which were too complex to
classify, and number of suspected chrysotile and
amphibole fibers which could not be positively
identified.
13.0 METHOD PERFORMANCE
13.1 Limitations of accuracy can result from errors in identification
of asbestos structures. Complete identification of every fiber
is not possible due to both instrumental limitations and the
nature of some of the fibers.
13.1.1 The requirement for a calibrated SAED pattern
eliminates the possibility of an incorrect
identification of chrysotile fibers. However, it is
not always possible to obtain a satisfactory
diffraction pattern from every fiber. The only
significant possibilities of misidentification occur
with halloysite, vermiculite scrolls or palygorskite,
all of which can be discriminated from chrysotile by
the use of EDXA and by observation of the 0.73 nm (002)
reflection of chrysotile in the SAED pattern (1).
13.1.2 Complete identification of all amphibole fibers is not
practical due to instrumental factors, the nature of
some of the fibers, and limitations of time and cost.
Particles of a number of other minerals having
compositions similar to those of some amphiboles could
be mistakenly classified as amphibole when zone axis
SAED is not used. However, quantitative EDXA
measurements on all fibers as support for the random
orientation SAED patterns makes misidentification
unlikely. The possibility of misidentification is
further reduced with increasing aspect ratio, since
many of the minerals with which amphibole may be
confused do not display its prominent cleavage parallel
to the c-axis (1).
13.2 Limitations of accuracy can also result from the overlapping of
structures by other (nonasbestos) particulates.
13.3 Inadequate dispersion of fibers can occur if enough organic
contaminants were present in the original sample to cause
adhesion of the fibers to the container walls or each other.
13.4 Contamination of the aliquots by asbestos during preparation in
the laboratory can lead to inaccuracy of results. This is a
particular problem with chrysotile, and should be carefully
monitored by preparation of blank samples.
13.5 This method has not yet been subjected to an inter!aboratory
test round. Precision measurements for EPA intralaboratory
comparisons of results from multiple operators using three TEMs
produced a relative standard deviation (RSD) of 26.5% for MCE
filters and 25.5% for PC filters for fibers over 10/im in length
in standard samples. For similar methods, precision
18
-------
measurements for intralaboratory comparisons have been found to
have an RSD of 13 to 22%rfor standard and environmental water
samples, with an RSD of 8.4 to 29% for interlaboratory com-
parisons (1). Statistical formulae for the establishment of
confidence limits on the laboratory results can be found in
Chatfield and Dillon (1). An earlier study found an inter-
laboratory reproducibility of 25 to 50% in standard samples (6).
Accuracy measurements from inter and intralaboratory studies
have demonstrated an RSD of 17% for standard chrysotile
suspensions and an RSD of 16% for standard crocidolite
suspensions (1).
13.6 The detection limit will depend upon the concentration of
asbestos in the original sample and the constraints of time and
cost of analysis. The detection limit can be improved by
increasing the amount of water filtered, increasing the number
of grid openings counted or decreasing the size of the filter
used (when practical). Samples containing a high level of
particulates will necessarily have a higher detection limit.
14.0 POLLUTION PREVENTION .
14.1 Pollution prevention encompasses any technique that reduces or
eliminates the quantity or toxicity of waste at the point of
generation. Numerous opportunities for pollution prevention
exist in laboratory operation. The EPA has established a
preferred hierarchy of environmental management techniques that
places pollution prevention as the management option of first
choice. Whenever feasible, laboratory personnel should use
pollution prevention techniques to address their waste
generation. When wastes cannot be feasibly reduced at the
source, the Agency recommends recycling as the next best option.
14.2 The quantity of chemicals purchased should be based on expected
usage during its shelf life and disposal cost of unused
material. Actual reagent preparation volumes should reflect
anticipated usage and reagent stability.
14.3 For information about pollution prevention that may be
applicable to laboratories and research institutions, consult
Less is Better: Laboratory Chemical Management for Waste
Reduction," available from the American Chemical Society's
Department of Government Regulations and Science Policy, 1155
16th Street N.W., Washington D.C. 20036, (202)872-4477.
15.0 WASTE MANAGEMENT
15.1 The U.S. Environmental Protection Agency requires that
laboratory waste management practices be conducted consistent
with all applicable rules and regulations. Excess reagents
samples and method process wastes should be characterized and
disposed of in an acceptable manner. The Agency urges
laboratories to protect the air, water and land by minimizing
and controlling all releases from hoods and bench operations
complying with the letter and spirit of any waste discharge
permit and regulations, and by complying with all solid and
19
-------
hazardous waste regulations, particularly the hazardous waste
identification rules and land disposal restrictions. For
further information on waste management consult the "Waste
Management Manual for Laboratory Personnel," available from the
American Chemical Society at the address listed in Sect. 14.3.
16.0 REFERENCES
1. E.J. Chatfield and M.J. Dillon, "Analytical Method for the
Determination of Asbestos in Water." EPA 600/4-83-043.
2. "Reagent Chemicals, American Chemical Society Specifications,"
American Chemical Society, Washington, D.C.
3. USEPA, Asbestos-Containing Materials in Schools: Final Rule and
Notice. Federal Register, 40 CFR Part 763, Appendix A to Subpart
E., October 30, 1987.
4. NIST/NVLAP Program Handbook for Airborne Asbestos Analysis.
NISTIR 89-4137, August, 1989.
5. G.J. Burdett and A.P. Rood, "Membrane-Filter DirectrTransfer
Technique for the Analysis of Asbestos Fibers or Other Inorganic
Particles by Transmission Electron Microscopy," Environ. -Sci.
Techno!. 17:643-648, 1983.
6. K.S. Chopra, HInter!aboratory Measurements of Amphibole and
Chrysotile Fiber Concentrations in Water," in National Bureau of
Standards Special Publication 506, Proceedings of the Workshop on
Asbestos: 'Definitions and Measurement Methods, 1977.
20
------- |