EPA Report No.
July 1984
METHODOLOGY FOR THE MEASUREMENT OP AIRBORNE ASBESTOS
BY ELECTRON MICROSCOPY
by
George Yamate
Satish C. Agarwal
Robert D. Gibbons
III Research Institute
Chicago, Illinois 60616
Contract No. 68-02-3266
Project Officer
Michael E. Beard
Quality Assurance Division
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY-
OFFICE .OF .RESEARCH. AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring Systems
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
FOREWORD
iii
-------
PREFACE
iv
-------
ABSTRACT
The provisional electron microscope methodology for measuring the
concentration of airborne asbestos fibers was refined. The methodology is
divided into separate protocols. The step-by-step procedures for each
protocol are nearly identical, so that cumulative data can be obtained and
uncertainties, especially in asbestos identification, can be clarified. The
operational steps encompass (1) type of sample, (2) collection and transport,
(3) sample preparation, (4) examination under the transmission electron
microscope (TEM) and data collection, (5) data reduction and reporting of
results, and (6) quality control-quality assurance.
The TEM analytical protocol is subdivided into three levels of anal-
ysis: Level I, for screening many samples; Level II, for regulatory action;
and Level III, for confirmatory analysis of controversial samples. Because
identification of asbestos structures is critical, the level of analysis is
directly related to the information sought:
Level I—morphology and visual selected area electron
diffraction (SAED) pattern recognition.
Level II—morphology; visual SAED; and elemental analysis.
Level III—morphology; visual SAED; a selected number of SAED
micrographs of zone-axis patterns; and elemental
.analysis.
This report was submitted in fulfillment of Contract No. 68-02-3266 by
IIT Research Institute under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period September 19, 1979, to
June 19, 1981, and work was completed as of September 30, 1981.
-------
CONTENTS
Disclaimer ii
Foreword iii
Preface iv
Abstract v
Tables ix
Figures x
List of Abbreviations xi
1. Introduction 1
2. Conclusions and Recommendations 3
3. Guidelines for Understanding the Methodology 4
Level of Analysis 5
Order of Analysis 5
Collection and Reporting 5
Costs 5
Application to Nonairborne Sources 6
Geographical Considerations 6
Laboratory Conditions 6
4. Level I Analysis . 8
Summary of Protocol 8
Equipment, Facilities, and Supplies 8
Description of Methodology 9
1. Type of Samples—Source 9
2. Sample Collection and Transport 10
3. Sample Preparation for Analysis—Grid Transfer 13
4. TEM Examination and Data Collection 16
5. Data Reduction and Reporting of Results ,21
6. Quality Control/Quality Assurance 22
5. Level II Analysis 24
Summary of Protocol 24
Equipment, Facilities, and Supplies 25
Description of Methodology 25
1. Type of Samples—Source 25
2. Sample Collection and Transport 26
3. Sample Preparation for Analysis—Grid Transfer 29
vi
-------
CONTENTS (continued)
Page
4. TEM Examination and Data Collection 33
5. Data Reduction and Reporting of Results 39
6. Quality Control/Quality Assurance 42
6. Level III Asbestos Analysis 44
Discussion of Protocol 44
Summary of Protocol 46
Equipment, Facilities, and Supplies 47
Description of Methodology 48
1. Crystallography and Morphological Properties 48
2. Chemical Properties—Elemental-Analysis by EDS 49
3. Selected Area Electron Diffraction (SAED) 49
4. Use of Tilting to Acquire Exact Zone-Axis SAED Patterns 50
5. Characteristics of SAED Patterns Encountered 51
6. Determination of Camera Constant and SAED Pattern Analysis 53
7. Determination of Camera Constant Using Gold Rings 54
8. Measurement of d-Spacings and Interplanar Angles 55
9. Identification of Unknown Fibers 58
7. Archival Samples 60
Discussion of Protocol 60
Description of Methodology 61
1. Samples with Adequate Loading 61
2. Samples with Heavy Loading 62
8. Bulk-Sample* Analysis 65
Discussion of Protocol 65
Description of Methodology 65
* 1. Polarized Light Microscopy 65
2. X-Ray Diffraction Analysis 66
3. Electron Microscopy .• 66
9. Numerical Relationships and Analytical Aids 67
1
Limits of Detection 67
Statistical Methodology 68
1. 95% Confidence Limits for a Poisson Variate 70
2. Comparison of Two Poisson Variates 70
Magnification Calibration 73
Preparation of Blanks 74
Use of Computers 74
References 75
vii
-------
CONTENTS (continued)
Page
Appendices
A. Figures 76
B. Computer Printout of Level I Analysis (Example) 96
C. Computer Printout of Level II Analysis (Example) 99
viii
-------
TABLES
Number Page
1 Profile Comparison of Asbestos Standards 40
2 Determination of Camera Constant (Example) 55
3 Determination of Spot Spacings (Examples) 57
4 Comparison of d-spacings' from SAED File and
Powder Diffraction File (Example) 59
5 Hypothetical Data 69
6 95 Percent Confidence Limits 71
7 Control and Test Sample Differences 72
ix
-------
FIGURES
Number
Al Vacuum evaporator 76
A2 Multiple coating arrangement in evaporator 77
A3 Close-up of multiple-coating arrangement 78
A4 Modified Jaffe wick washer method (sketch) 79
A5 Modified Jaffe wick washer 80
A6 Transmission electron microscope 80
A7 Morphology and counting guidelines used 1
in determining asbestos structure 81
A8 Level I data sheet Cexample) 82
A9 Scanning of full-grid opening... 83
A10 Transmission electron microscope with energy dispersive spectrometer...84
iff
All Spectra profiles of asbestos standards 85
A12 Level II data sheet (example) 86
A13 EM data report (example) 87
•*'
A14 Sample summary report (example) 88
A15 Effects of tilting and alignment of fiber 89
A16 Method of measuring two perpendicular diameters for each ring 89
A17 Method of recognizing a horizontal row of spots 90
A18 Relationship of di , 62 , 6l.2> and R 91
A19 Typical zone-axis SAED patterns from araosite standard specimen 92
. ••
A20 Typical zone-axis patterns from crocidolite standard specimen 93
A21 Typical zone-axis patterns from treraolite standard specimen 94
A22 Typical SAED patterns and EDAX spectra from anthophyllite
standard specimen 95
-------
LIST OF ABBREVIATIONS
AEM analytical electron microscope
EDS. energy dispersive spectrometer
EM electron microscope
JCPDS Joint Committee on Powder Diffraction Standards
LTA low-temperature ashing
NIOSH National Institute of Occupational Safety and Health
PLM polarized light microscopy
QC/QA quality control/quality assurance
SAED selected area electron diffraction
SEM scanning electron microscope
STEM scanning transmission electron microscope
TEM transmission electron microscope
TSP total suspended particulates
U1CC Union Internationale Contre le Cancer
XRD x-ray diffraction
XRF x-ray fluorescence
xi
-------
SECTION 1
INTRODUCTION
Asbestos Is recognized as a health hazard, especially if inspired into
the alveolar region of the respiratory tract. Asbestos may be present in air
samples, water samples, biological or clinical samples, and other miscel-
laneous bulk samples, such as ores and food. These various types of samples
require different collection methodologies and diverse preparation techniques.
Asbestos analysis methodologies may be categorized as bulk-material
analyses, or those providing concentration information, and single-fiber
analyses, or those providing morphology, size distribution, and concentra-
tion. BuJLJk-material_analy.sis_ techniques, which include infrared spectroscopy,"
("differential thermal analysis, and x-ray diffraction analysis (XRD), are
ilimited by an inability to analyze concentrations of less than 1 ug» and by an
'inability to differentiate between fibrous and nonfibrous forms of minerals.
i i
Sing_le^fiber.,anal.ysis. techniques include optical microscopy and electron
microscopy. Optical microscopy employing phase contrast has been promulgated
injto_a raonitorpTg~*inethod for the workplace environment "(NIOSH-rP&CAM 239). In
addition^promulgation of a monitoring method for bulk-material asbestos
samples (building insulation) using polarized light, microscopy (PLM) is
presently being considered. However, optical microscopic techniques cannot;
[determine fibers of less than approximately 1 ym in diameter, and phase
;contrast cannot differentiate between asbestos and nonasbestos fibers. :
The electron microscope (EM) provides particle morphology and size, and a
degree of identification. A comprehensive study of various EM procedures
(Samudra et al., 1977) was conducted in development of a provisional method-
ology manual, Electron Microscope Measurement of Airborne Asbestos Concentra-^
tions (Samudra, 1978). Three EM methods are available: the scanning electron
microscope (SEM), the transmissio.ix electron microscope (TEM), and the analyti-
cal electron microscope (AEM). The SEM, with an x-ray energy-dispersive
spectrometer (EDS), permits visual characterization (analogous to reflection
optical microscopy) and fiber identification by elemental analysis. The TEM,
providing an increased data-acquisition capability, permits visual characteri-
zation (in the transmitted mode) and fiber identification by crystal structure
analysis. The AEM is a TEM with an EDS, and with the added capability of
SEM/STEM (scanning transmission electron microscope) operation, which permits
visual characterization (morphology and size) as well as fiber identification
using both crystal structure by selected area electron diffraction (SAED) and
elemental analysis by EDS.
-------
j" The original EM methodology was developed for the U.S. Environmental /
iProtection Agency (EPA) for measuring airborne asbestos concentrations, j
ispecifically for ambient air and for use as a "screening" tool. Development
guidelines included attainable precision and accuracy of results; relative
rapidness in use; cost-effectiveness; applicability to a large number of
laboratories possessing a TEM (at that time, very few laboratories had TEM's
with x-ray analysis capability or an AEM); and procedural steps to be inde-
pendent of unique or exceptional in-house capabilities of a single laboratory
(that is, interlaboratory precision rather than intralaboratory).
In usage, the EM method was successful within its prescribed limita-
tions—that is, the precision and accuracy of results between laboratories
using the complete method was good. However, problems __that had been recogr
nized in the study developing the methodology (Samudra et at., .1977) arose in
"the areas of (1) interpretation of airborne, (2) sample collection, (3) need
for more exacting identification of asbestos, especially of amphibole type,
and (4) use of only part of the methodology.
The,.pr.esent .study..was .undertaken to.refine .the..methodology.. The problem
areas and related criticisms were addressed within the underlying goals and
guidelines set for optimizing the methodology. Protocols similar to a cook-
book were not possible since basic knowledge or training was required regard-
ing (1) sample collection, (2) preparation of samples for EM, (3) use of the
TEM-AEM, and (4) diffraction pattern analysis. The refined methodology is
based on an assumption that each intended user of a particular level of analy-
sis has the necessary background and training to use it.
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
The EM methodology for measuring the concentration of airborne asbestos
fibers has been refined and specified, and is recommended for field evalua- ._
,tion. The methodology is based on a TEM analytical protocol that is divided 1
jinto three levels of effort: Level I, for screening many samples; Level II, j
;,for regulatory action; and Level III, for confirmatory analysis of controver-.J
Isial samples. The three-level analytical methodology is cost-effective, and
will provide the required results for proper assessment of asbestos.
-------
SECTION 3
GUIDELINES POR UNDERSTANDING THE METHODOLOGY
The methodology is divided into separate protocols. The step-by-step
procedures for each protocol are nearly identical, so that cumulative data can
be obtained and uncertainties, especially in asbestos identification, can be
clarified. These operational steps are:
(1) Type of Sample—Source
(2) Sample Collection and Transport
(3) Sample Preparation for Analysis—Grid Transfer
(4) TEM Examination and Data Collection
(5) Data Reduction and Reporting of Results
(6) Quality Control/Quality Assurance (QC/QA).
The analytical, protocol under the TEM examination and data collection
procedure is subdivided into three levels of increasing analytical effort in
terms of requiring an instrument of greater capability, an electron
tnicroscopist with greater expertise, and a longer analytical time. Level I, a
monitoring or screening methodology, resembles the present EPA provisional"
methodology (Samudra et al., 1978; Anderson and Long, 1980). Ley.e.l_.II, is a
regulatory method requiring additional analytical criteria to establish
asbestos identification limits, and to provide guidance for Level I analyses
by confirming or clarifying visual SAED patterns, ^yel IIIi the most
sophisticated and the costliest of the methods, is intended for confirming ,
asbestos identification, especially in judicial controversies and other
special situations.
x ••
In Sections 4, 5, and 6, the protocols for each of the three levels of
.analysis are presented independently of each other, and thus procedures common
to each are repeated. All figures are presented in Appendix A.
Section 7 describes modifications for using the methodology on archival
samples, which are samples collected on nonprocedural filter substrates, or
samples collected without regard to filter loading levels. Section 8
describes analysis of inorganic sources in bulk-air samples or in bulk form.
Section 9 concludes the report with a discussion of analytical aids pertaining
to the limits of detection, preparation of blanks, use of computers,
magnification calibration, and statistical methodology.
-------
General guidelines for understanding the methodology are discussed in
the following paragraphs.
LEVEL OP ANALYSIS
Knowledge of the history, source, and location of the sample, and the
purpose and objective of the analysis aids in selecting the correct level of
analytical effort. Simply "grinding the samples out" neither is cost-
effective nor produces the best results, especially for Level II and Level III
analyses. Instead of all Level I, all Level II, or all Level III, the
majority of the analyses may be Level I, followed by some Level II. Level III
could be used in its entirety or only at the analytical phase. If the source
is known to contain no amphlbole-type interference, or if chrysotile is of
interest, gold-coating can be eliminated.
If a legal proceeding is anticipated, Level III analysis will be required
where a chain-of-custody record is kept from collection, transport to the
laboratory, preparation, analysis, data reduction, and reporting of results.
EM finder grids oust be used for grid transfer. In addition, for quality
assurance, a second laboratory must be available for analyzing a portion of
the sample using the same degree of custodial care. QC/QA protocols must be
observed and records kept.
Whenever possible, and especially for unknown source samples, 10 to 20%
of each set of samples should be analyzed by Level II analysis prior to using
Level I as a screening procedure.
Level I is a relatively rapid procedure, and can be used by many
laboratories with access to a conventional TEM. However, Level I results
should net be used in legal proceedings. If "positives" or "false positives"
are found, especially in areas where asbestos is known to be absent, and the
field blank and laboratory blank have been checked, Level II analysis, and
possibly Level III analysis, should be performed.
ORDER OF ANALYSIS ••'-
. . The -order of analysis is (1) field blanks, (2) laboratory blanks (if .--_
needed), and (3) field samples.
COLLECTION AND REPORTING
The counting rule, "minimum 100 fibrous structures per known area
(complete grid opening) or 10 grid openings, whichever is first," is a minimum
rule for cost limitation. For very low asbestos presence, or for asbestos
contamination studies, where particulate loading is high and asbestos presence
very low, counting 20 grid openings from each of 2 grids (10 per grid) is
recommended.
The EM magnification factor is very high or, conversely, the area of '
deposit examined is very small. Therefore, although the electron microscopist '
may report a zero count, the notation "Below Detectable Level" is more appro-._$
priate in the sample report. Along the same lines, the electron microscopist
-------
should report observations, measurements, and conclusions as objectively as
possible, realizing the subjective nature of his decision-making, such as
parallel-sided, 3:1 aspect ratio, number count, size measurements,
recognition-discrimination of SAED patterns, and categorizing of asbestos
structure.
Data reduction and reporting of results must be consistent and stated.
Dimensions of X-fibers (unknown length since complete fiber is not visible)
may be doubled, not counted at all, or presented separately. Doubling of the
visible portion is recommended, and should be so stated in the report.
Mass or conversion of size measurements to an assumed shape-volume-
density relationship, is calculated, and thus is the least reliable of the
data, especially for X-fibers, bundles, clusters, and matrices.
Although morphology, SAED, and XRF either singly or in combination will
provide identification of asbestos, not all structures will be identified.
The nature of the asbestos structure prevents analysis of all structures by
SAED and/or by elemental analysis with EDS. Such factors as specimen thick-
ness, orientation, and proximity to other particulates or to the grid wire
will prevent attainment of good SAED patterns and limit the effectiveness of
chemical analysis.
COSTS
Levels I, II, and III analyses are estimated to require 200, 400, and
1200 min per analysis, respectively. Additional costs will result from
collection, preparation, and reporting of results. The equivalent monetary
costs will depend on the laboratory rates of the personnel involved.
APPLICATION TO -NONAIRBORNE SOURCES
f~ Although the methodology has been developed for airborne asbestos, other~~i
\types of samples from different sources can be analyzed if the samples are
\ finely divided and placed with proper loading and uniform distribution either_
ion .a.polycarbonate membrane filter or on a carbon-coated EM grid. Of course,
the limitations of the collection and preparation steps must be known and ..
accounted for to prevent inaccuracies in comparing results.
GEOGRAPHICAL CONSIDERATIONS
In some parts of the country, such as the-Upper.Great Lakes area, the
possibility of misidentification is much greater.because some.nonamphibole
minerals have visual SAED patterns.that.closely resemble those of amphiboles.
Gold-coating and Level II analysis will help in differentiating^ between these
minerals. .
LABORATORY CONDITIONS
Asbestos analysis involves sustained microscopy for periods of 3 to
7 hours with unscheduled rest breaks. Subjective decisions regarding such
factors as morphology, size measurement, visual identification, and possible
6
-------
1, difficult to ore** a .anipul.tive
^r"o^V^»-i°«l^
redundant v>.^^ *-- • -
deadlines may contribute to poor precision,
-------
SECTION 4
LEVEL I ANALYSIS
SUMMARY OF PROTOCOL
Level I analysis is a monitoring or screening technique. It assesses the
amount and type of asbestos structures in the atmosphere through the following
steps:
; - •.«_.
(1) .A known volume of air is passed through a polycarbonate
;membrane filter (pore diameter, 0.4 ym; filter diameter,'
137 or 47 mm) to obtain approximately 5 to 10 yg of
Iparticulates per cm2 of filter surface.
(2) The particulate-laden filter is transported in its own
filter holder.
(3) The filter is carbon-coated in the holder.
(4) jThe particulates are transferred to an EM grid using a
Irefined Jaffe wick washer.
(5) ; The EM grid, containing the particulates, is gold-coated
jj-ightly.
(6) The EM grid is examined under low magnification (250X to
lOOOXffollowed by high-magnification (16.000X on the
fluorescent screen) search and analysis.
I »—•
:" (7) 'A known area (measured grid opening) is scanned, and the I
'; fibrous structures (fibers, bundles, clusters, and '
.•matrices) are counted, sized, and identified as to
•--y-K ~-. • • asbestos type (chrysotile, amphibole, ambiguous, or no
/ identity) by morphology and by observing the SAED pattern.-
(8) The observations are recorded—a minimum of 100 fibrous
structures or 10 grid openings, whichever is first.
(9) The data are reduced and the results reported.
EQUIPMENT, FACILITIES, AND SUPPLIES
The following items are required for Level I analysis:
(1) An 80__qr 100-kV TEM. with a fluorescent viewing screen
inscribed with graduations for estimating the length and
width of fibrous particulates.
-------
(2) JA vacuum evaporator with -a turntable for rotating 7
jspeciraens during coating, for such uses as carbon-coating;
[polycarbonate filters, gold-coating EM grids, and i
^preparing carbon-coated EM grids.
(3) An EM preparation room adjacent to the room housing the
EM. This room should either be a clean-room facility, or
contain a laminar-flow class-100 clean bench to minimize
contamination during EM grid preparation. Filter handling
and transfer to EM grids should be performed in a clean
atmosphere. Laboratory blanks should be prepared and
analyzed weekly to ensure quality of work.
Several refined Jaffe wick jwashers for dissolving membrane
filters.
(5) Miscellaneous EM supplies and chemicals, including carbon-
coated 200-mesh copper grids, grid boxes, and chloroform.
(6) Sample collection equipment, including 37-mm-diameter or
47-mm-diameter filter holders, 0.4-ym (pore size)
polycarbonate filters, 5.0-um (pore size) cellulose ester
membrane filters for back-up, a sampling pump with
ancillary equipment, a tripod, critical orifices or flow
meters, and a rain/wind shield.
DESCRIPTION OF METHODOLOGY
1. Type of Samples—Source
This protocol was' originally developed for the EPA for measuring airborne
asbestos_..(Saraudra et al., 1977; Samudra et al., 19.7.8). A broad interpretation
of airborne has been to apply the term to samples obtained from ambient air
(the original purpose), aerosolized source materials (such as the asbestos
workplace environment, and fugitive dust emissions), bulk-air material (such
as total suspended particulate (TSP) samples, dust, and powders) and any other
type of sample obtained by nonrestrictive use of (1) collection of a volume of
air, (2)'separation from the air, and (3) concentration of the particulates
onto a substrate. The airborne protocol has also been applied to samples
collected in the regulatory areas of the EPA, as compared with, for example,
the workplace environment (National 'Institute of Occupational Safety and
Health), mining activities (U.S. Bureau of Mines), and shipboard atmosphere
(Federal Maritime Administration).
The present methodology has been optimized for application specifically
to samples collected from a volume of air in which the asbestos concentration
is considered a minor component of the total particulate loading (other analy-
tical methods are available for samples known to contain high concentrations
of asbestos); and in which the particles are less than 15 jjm in diameter,
since particles~g"reater than 15 gm either are not inhaled or are deposited in
the upper respiratory tract and expelled, and preferably less than 10 ym in
diameter as recommended by the Clean Air Scientific Advisory Committee
(Hileman, 1981), since particles up to 10 pm can be absorbed by the alveolar
-------
region of the lung. These concentration and size restrictions will preclude
many air samples collected in an asbestos-processing environment and in bulk-
air material from the complete methodology. However, such samples can still
be examined with the TEM, within the limitations of the instrument, through
changes in preparation techniques—provided the effects on the final results,
such as fractionation of size and representativeness of the sample, are
carefully considered.
2. Sample Collection and Transport
Sample Collection—
Sampling procedures vary depending on the nature of the sample, purpose
of collection, analytical method to be used, sample substrate, and time and
cost of sample collection relative to the total analytical effort. Neverthe-
less, the primary objective of sample collection always is to obtain a repre-
sentative, unbiased sample.
Impingers, impaction devices, electrostatic precipitators, and thermal~1
precipitators have been used in sample collection, but each has limitations.j
, Presently, the preferred substrates are membrane filters^ which are manu- "~*
factured from different polymeric materials, including polycarbonate, mixed
esters of cellulose, polystyrene, cellulose acetate, and cellulose nitrate.
Polycarbonate membrane filters..differ from the others in being thin, strong,
and smooth-surfaced, and in having sieve-like construction (circular pores
from top surface to the bottom). The other membrane filters are thicker, have
irregular-surfaces, and have depth-filter construction (tortuous paths from
top surface to bottom).
Consequently, polycarbonate filters have been selected for airborne f
.asbestos analysis. The collection of small-sized particles (prefer less than";
!lO urn in diameter), the light loading of participates, the uniform distribu-
tion of particulates attainable using a depth-type backing filter, the smooth
surface and circular holes (which aid in determining size and instrument tilt
'.axis), and the relative ease in grid transfer (thin and strong) minimize • -.-:«. **. •.
disadvantages of lack of retention and/or movement of large particles during-'^SC^
handling. Other membrane materials,_ such as the cellulose.-ester....type,_.are__ ......
recommended for phase" contrast and PLMj'neavy"Part*-cle- loadings, and physical ,-,..
• . .. . ,_ - •-- _ .— - ->•- . . . . . . ... "!'i.Vrp,!V '
retention of large particles. ••".>,-'"
-r- - • -••>•
In microscopical analysis, uniformity of particulate distribution and
loading is critical to success. Air samples are taken on 37-mm-diameter or
47-mm-diameter, 0.4-ym (pore size) polycarbonate membrane filters using the
shiny, smooth side as the particle-capture surface. Cellulose ester-type
membrane filters (pore size, 5.0 ytn) are used to support the polycarbonate
filter on the support pad (37-mra-diaraeter personal sampler) or on the support
plate (47-mm-diameter holder).
Air monitoring cassettes (37-mm-diameter) of three-piece construction are
available from several manufacturers. As with the 47-ram-diameter filters,
loading the cassettes with the support pad, back-up filter, and 0.4 ym (pore
size) polycarbonate filter should be carefully performed on a class-100 clean
bench. Since the filters are held in place by pressure fit rather than by
10
-------
screw tightening, air must not enter from the sides of the unit; a plastic
band or tape (which can double as a label) should be used as a final seal.
Collecting airborne samples with proper loading requires experience.
Each of the following techniques is useful in collecting airborne samples for
direct microscopy, preserving representative sizes, without diluting
participate deposits:
(1) For long-terra sampling at a site, test samples should be
returned to the laboratory by express mail service, or air
express service or by being hand-carried, and should then
be analyzed by scanning electron microscopy.
(2) The estimated particulate loading (deposit is barely
visible to the naked eye) should be bracketed by varying
the filtration rate and using the same time, or by varying
the time and using the same filtration rate.
(3) An automatic particle counter, such as a light-scattering
instrument (0.3-ym detection) or a real-time mass monitor
(0.1-ym detection), should be used to obtain an
approximate particulate-loading level of the area.
Although any one of the three techniques will work, the suggested tech-
nique is to take the samples as a set, varying the sampling rates and using
the same time so as to obtain filter samples with different particulate
loadings. Each set is composed of a minimum of four 37-mm-diameter or 47-mm-
diameter filter units — three for different particulate loadings (low, medium,
high), and the fourth for a field blank. Suggested sampling rates are 0 for
the field blank, 2.48. L/rain for the low loading, 7.45 L/min for the medium,
and 17.62 L/min for the high, for a 30 min sampling period using a 47-mm-
diameter filter "holder. Simultaneous sampling will provide at least one
sample with a particulate loading suitable for direct EM analysis.
"• ''
TSP's range from 10 yg/m3 in remote, nonurban areas, to 60 yg/m3 in near-~7
J
urban areas, to 220 yg/m3 in urban areas. However, for heavily polluted
areas,,...TSP. levels may reach 2000 yg/m3. A loading of 5 to 10 _yg .per cm2 of
filter is adequate for EM analysis; values beyond 20 to 25 yg per cm"2 require' '.'-,.''v:
a dilution treatment. As an example, for 47-mm-diameter filters at face
velocities of 3.0 cm/s (2.48 L/min)t, 9.0 cm/s (7.45 L/min), and 21.2 cm/s
(17.62 L/min), respectively, air vo'lumes of 74.4 L, 223.5 L, and 528.6 L are
sampled in 30 min. For a TSP level of 200 yg/m3, 14.88 yg (1.07 yg/cm2),
44.7 yg (3.23 yg/cm2), and 105.7 yg (7.63 yg/cm2), respectively, would be
collected on 47-mm-diaraeter filters (which would have effective filtration
areas of 13.85 cm2). The sampling time could be increased to 60 rain for areas
having lower TSP levels, or reduced in a heavily polluted area (source
emissions).
Airborne samples from emission sources contain coarse particles (above
the respirable size) of large matrix structures, binder materials, road dust,
clay minerals, fillers, and other materials. For these samples, a fifth
filter unit can be added that has a size-selective inlet (cyclone, impactor, (
or elutriator) attached prior to the filter unit. The flow pattern and flow
11
-------
rates of the tandem sampling arrangement must be checked before use. A satis-
factory, tested combination presently used in California is a cyclone-filter
unit with a DSQ cut-off of 2.5 ym at 21.7 L/rain, and a DSQ cut-off of 3.5 u™
at 15.4 L/min (John and Reischl, 1980). Additional sampling devices, such as
impingers (used in biological sampling), impactors, and other designated
filter units (for TSP, XRD, or x-ray fluorescence (XRF), for example) can be
added to the system to obtain supplementary as well as interrelated data.
This expandable multifilter sampling unit, designated Hydra, offers the
following advantages:
(1) It is small, inexpensive, and compact, so that an adult
can easily handle it.
(2) It is efficiently designed, and includes a tripod,
sampling pump, manifold, critical orifices, and a row of
preloaded 37-inm-diameter or 47-mm-diameter filter
holders. A rain/wind shield, size-selective cyclone-
filter units, tubing, and other extras can be added as
needed.
(3) Its sample preparation steps and handling are minimized.
(4) It allows complementary as well as supplementary analysis
(TSP, size fractionation, bacteria, and XRF, for example),
although additional air sampling capacity is required.
(5) It accommodates ambient air and source emission samples,
with or without a size-selective inlet.
(6) It allows synchronous sampling in several places in the
vicinity following the same sampling procedure, thereby
accommodating particulate concentration fluctuations.
•f
(7) It includes filter holders that serve as transport and
. . storage units.
$•&
Hydra's disadvantages are a short sampling period, which may record an
episode; a small sampling quantity or volume, which may not indicate the
"'presence""6f ' asbestos fibers; £nd__a_detection...limit.pf_ 2 x_ 101* fibers/in3 for ,..„..
sampling 1 m3 of air with the 47-mm-diameter filter". ..... ..... ." '" "''"r"iv v-'!'-4^'
_ __ »1 --- II* — -- •"-" *"*•-._ ........ -. • •.-•••«•. . __ . . . _ 1. .,!,'.,
Using 8 inch x 10 inch, or 102-mm-dlameter filter sizes, is not recom-
mended. The sampling units are designed for purposes other than microscopy.
Interchanging the type of sample substrate filter (glass fiber or paper to
polycarbonate) does not correct the inherent problems of filter size and
sampling unit.
Sample Storage and Transport —
Once the sample is acquired, its integrity must be assured, and contami-
nation and loss of fibers prevented, until it is examined under the EM. The
low cost and small size of the 37-mm-diameter and 47-mm-diameter filter
holders enables them to be used as combination storage and transport con-
tainers. The filter holders should be maintained in a horizontal position
12
-------
during storage and transport to the laboratory so that the particulate-loaded
filters can be removed under optimally controlled conditions in the laboratory.
For 47-mm-dianieter holders (open-face) to be used in transport or stor-
age, the screw cap is carefully removed, and the shiny, waxy, stiff separator
paper used to keep the polycarbonate filters apart is carefully placed on the
retaining ring. The cap is then carefully screwed back on so that the sepa-
rator paper seals and protects the particulate-loaded filter without touching
it. The 37-mm-diameter, three-piece filter holder (aerosol monitor) is used
in its open-face position, and capped after usage for transport and storage.
When the more expensive 47-mra-diameter holder is to be re-used immedi-
ately, the particulate-loaded filter should be carefully removed and placed in
a 47-mm-diameter Petri-slide (such as that manufactured by the Millipore
Corp.*) This transfer takes place in the"field rather than in the laboratory,
so that the Petri-slide should be taken into the field. The 37-mm-diameter
filter holder or the 47-mm-diameter holder/Petri-slide should be secured and
all necessary sample identification marks and symbols applied to the holder.
3. Sample Preparation for Analysis—Grid Transfer
Carbon-Coating the Filter—
~ The polycarbonate filter, with the sample deposit and suitable blanks, f
should be coated with carbon as soon as possible after sampling is completed.;
To begin this procedure, the particulate-loaded 47-mra-diameter polycarbonate j
filter is removed from the holder and transferred carefully to an open-faced |
47-mm-diameter Petri-slide for carbon-coating in the vacuum evaporator (see —!
j_JFigure Al, Appendix A). If the 47-mm-diameter filter is already in the Petri-
^lide, the cover is replaced with an open-face cover, minimizing filter
disruption. The^37-mm-diameter filter is left in the holder, but the upper
lid is removed to create an open-faced filter. The open-faced holders are
placed on the rotating turntable in the vacuum evaporator for carbon-coating.
Figure A2 shows the multiple-coating arrangement in the evaporator; Figure A3
shows a close-up of the 37-mm-diaraeter and the modified 47-mm-diameter holders.
for carbon-coating.
'••"-'' For archival filters and those of'larger sizes, portions of about 2.5'cm
x 2.5 cm should be cut midway between the center and edge using a scalpel.
The portions are then attached 'withvcellophane tape to a clean glass micro-
scope slide and placed on the turntable'in the"vacuum .evaporator for coating.
Any high-vacuum carbon evaporator may.be used.to carbon-coat the.filters
(CAUTION: carbon sputtering devices should'not be'used), typically, the
electrodes are adjusted to a height of 10 cm above the level of the filters.
A spectrographically pure carbon electrode sharpened to a neck of 0.1 cm x
0.5 cm-is used as the evaporating electrode. _The sharpened electrode is
* Millipore Corp., 80-T Ashby Rd., Bedford, Mass. 01730
13
-------
• ' '
placed in its spring-loaded holder so that the 'rieck rests against the flat
surface of a second carbon electrode.
The manufacturer's instructions should be followed to obtain a vacuum of
about 1.33 x 10~3 Pa (1 x 10~5 torr) in the bell jar of the evaporator. 'With
the turntable in motion, the neck of the carbon electrode is evaporated by
increasing the electrode current to about 15 A in 10 s, followed by 20 to 25 A
for 25 to 30 s. If the turntable is not used during carbon evaporation, the
particulate matter may not be coated from all sides, resulting in an undesir-
able shadowing effect. The evaporation should proceed in a series of short
bursts until the neck of the electrode is consumed. Continuous prolonged
evaporation should be avoided, since overheating and consequent degradation of
the polycarbonate filter may occur, impeding the subsequent step of dissolving
the filter. The evaporation process may be observed by viewing the arc
through welders goggles (CAUTION: never 'look at the arc without appropriate
eye protection). Preliminary calculations show that a carbon neck of 5 mm3
volume, when evaporated over a spherical surface 10 cm in radius, will yield a
carbon layer that is 40 nm thick.
Following carbon-coating, the vacuum chamber is slowly returned to
ambient pressure, and the filters are'removed and placed in their respective
holders or in clean, marked Petri dishes for storage on a clean bench.
Transfer of the Sample to the EM Grid—
Transferring the collected particulates from the carbon-coated polycar-
bonate filter to an~EM~grid is accomplished in a clean room or on a class-100
clean bench". The transfer is made in a Jaffe wick washer, which is usually a
glass Petri dish containing a substrate to support the EM grid/carbon-coated
membrane filter combination. Solvent is added to a level to just wet the "~~1
fcombination and cause gentle dissolution of the membrane with minimum loss or/
.'dislocation of the particulates, resulting in a membrane-free EM grid with '
I particles embedded in the carbon film coating. The substrate support can be
;J:stainless steel mesh bridges, filter papers, urethane foams, or combinations
V-joT these. - .
- U •
TV-?---.-.-. The. ref ined.,.Jaf fe wick washer is_ described as follows:
(1) The glass Petri dish (diameter, 10 cm; height, 1.5 cm) is
made airtight by grinding the top edge of the bottom dish
with the bottom of the cover dish, with water and . .
Carborundum* powder (80 mesh); this creates a ground-
glass seal .(closer fit). ..and minimizes the need to-ref.ill
the Petri dish with added solvent. (The usual glass.....
Petri dish was found, .not to retain the solvent for long
periods of time, and unless the wicking substrate is kept
continuously wet, poor solubility of the membrane filter
results, leading to a poor-quality EM grid).
* Carborundum is a registered trademark of the Carborundum Co., Carborundum
Center, Niagara Falls, N.Y. 14302.
14
-------
(2) A combination of foam and a single sheet of 9-cm filter
paper is used as the substrate support. A 3-cm x 3-cm x
0.6-cm piece of polyurethane foam (the packing in
Polaroid film boxes) is cut and placed in the bottom
dish. A 0.5-inch V-shaped notch is cut into the filter
paper; the notch is oriented in line with the side of the
foam, creating a well for adding solvent.
Spectrographic-grade chloroform (solvent) is poured into
the Petri dish through the notch until it is level with
the top of the foam (also level with the paper). The
foam will swell, and care is needed to avoid adding
solvent above the filter paper.
(3) On top of the filter paper, pieces of 100-mesh stainless
steel screen (0.6 cm x 0.6 cm) are placed, usually in two
rows, to make several grid transfers at one time (for
such uses as replicas), and to facilitate maintenance of
proper identity of each transfer.
A 3-mm section (usually midway between the center and
edge) of the carbon-coated polycarbonate filter is cut in
a rocking motion with a scalpel. The section may be a
square, rectangle, or triangle, and should just cover the
3-mm EM grid.
(5) A section is laid carbon-side down on a 200-mesh carbon-
coated EM grid. (Alternatively, Formvar-coated* grids or
uncoated EM grids may be used. Here, the carbon coating
on the polycarbonate filter forms the grid substrate.)
Minor overlap or underlap of the grid by the filter
section can be tolerated, since only the central 2-mra
portion of the grid is scanned in the microscope. The EM
grid and filter combination is picked up at the edges
with the tweezers and carefully laid on the damp 100-mesh
stainless steel screen. The EM grid-filter combination
will immediately "wet out" and remain on the screen.
(6) Once all specimens are placed in the washer, more solvent
is carefully added through the notch" to maintain the
liquid level so that it just touches the top of the paper
filter. Raising the solvent level any higher may float
the EM grid off the mesh or. displace the polycarbonate
filter section. . - ...... ... . ..
(7) The cover is placed in the washer 'and oriented in place
over the specimen, and a map of 'the fiTter/grid'/s'creen
arrangement is made on the glass cover and in the
logbook. • '•-"•- •"-
* Forravar is a registered trademark of the Monsanto Company, 800 N. Lindbergh
Blvd., St. Louis, Mo.
15
-------
(8) Solvent (chloroform) is added periodically to maintain
the level within the washer mun_the_'jfiJLter_ i£_
completely dissolved by the wicking action (24 to 48 h).
* ,_-•—i»-iftu- •^^"•"** "• '•*»•- . v f ~ ' ^ |Ht,fci _ - *•* !»•*• -•*, ^ j ««- ., „, i , . • - •" r"»^
(9) The temperature in the room must remain relatively
constant to minimize condensation of solvent on the
bottom of the cover and subsequent falling of solvent
drops on the EM grid. Should day-night or other
temperature differentials occur, solvent condensation on
the under-surface of the cover can be minimized by
placing the Jaffe washer at a slight tilt (three glass
slides under one edge of the Petri dish parallel to the
row of grids) to allow the condensation drops to flow
toward the lower edge rather than fall on the EM grids.
At temperatures lower than 20°C (68°F), the complete
filter solution may take longer than 72 h.
(10) After the polymer is completely dissolved, the stainless
steel mesh screen with the EM grid is picked up while wet
and set on lens paper tacked to the bottom of a separate
Petri dish. The EM grid is then lifted from and placed
next to the screen to dry. When all traces of solvent
have evaporated, the grid is stored in a grid box and
identified by location and grid box in the logbook.
Figure A4 illustrates the Jaffe wick washer method; Figure A5 shows the
washer. The foam/filter combination is currently preferred, as is use of a
closely fitted (by means of the ground-glass seal) Petri dish.
Gold Coating—
Ajn_addi.t.i.o.nal...s,tep will aid jin_s_ubjectively_ evaluating theJSAED pattern.
This step is required for specimens from the uppeT~Gfeat" Lakes area and for
those of unknown*origins. After the particulates on the filter are
transferred to the EM grid, the grid is held to a glass slide with double
stick tape for gold-coating in the vacuum evaporator. Several EM grids may be.
taped to the glass slide for coating at one time. Approximately 10 mm of
0.015-cm-diameter (0.006-inch) pure gold wire is placed in a tungsten basket
(10.cm from.the rotating table holding the EM grids) and evaporated onto the
grid. ~~ •'"•-.T"
The thin gold-coating establishes an internal standard for SAED analysis.
For some mineral species, an internal standard will clarify visual identifica-
tion of the pattern of a fibrous particulate as being or not being an amphi-
bole species (for example, minnesotaite as opposed to amosite). With exper-
ience, differentiation in SAED patterns can be observed. For samples of known!
[geographic origins, gold-coating is optional, since the additional coating
Ihinders observation and identification of small-diameter chrysotile fibers. \
4. TEM Examination and Data Collection
. -.0.
v;s-.. .;
low Magnification Examination of Grids—
Figure A6 shows a modern TEM. The grid is observed in the TEM at
magnifications of 250X and 1000X to determine its suitability for detailed
16
-------
study at higher magnification. The grid is rejected and a new grid used if:
(1) the carbon film over a majority of the grid openings is damaged and not
intact; (2) the specimen is "dark due to incomplete dissolution of the
polycarbonate filter; or (3) the particulate loading is too light (unless a
blank) or too heavy with particle-particle interactions or overlaps.
TEN Analysis (Morphology and SAED)—
The following guidelines are observed for consistency in the analytical
protocol:
(1) Magnification at the fluorescent screen is determined by
calibration with a diffraction-grating replica in the
specimen holder.
(2) A field o'f view or "gate" is defined. On some
microscopes, the central rectangular portion of the
fluorescent screen, which is lifted for photographic
purposes, is convenient to use. On others, a scribed
circle or the entire circular screen may be used as the
field of view. The area of the field of view must be
accurately measurable.
(3) The grid opening is selected on a random basis.
(4) The analysis, morphology, and SAED are performed at a
tilt angle of 0°.
(5) The recommended instrument settings are: accelerating
voltage, 100 kV; beam current, 100 pA; film magnifi-
cation, 20JOOOX (which is equivalent to 16.000X on the
fluorescent screen for this instrument); and concentric
circles of radii 1, 2, 3, and A cm on the fluorescent
. , screen.
ii"*». '".. T-7.-iV '
"•"'',' •"• '• • • • .-.•"; >".'•.•£••>•
(6) The grid opening is measured at 1000X. ' ": ' " '
•:• :." "|-5-(7) 'Since asbestos fibers are found isolated" as well as with • .-:^K'"^.^.
each other or with other particles in varying arrange- "" " .''•': '
ments, the fibrous particulates are characterized as
asbestos structures: v
Fiber (F) is a particle with an aspect ratio of 3:1 or
greater, with substantially parallel sides.
Bundle (B) is a particulate composed of fibers in a
parallel arrangement, with each fiber closer than the
diameter of one fiber.
Cluster (Cl) is a particulate with fibers in a random
arrangement such that all fibers are intermixed and no
single fiber is isolated from the group.
17
-------
Matrix (M) is a fiber or fibers with one end free and the
other end embedded or hidden by a particulate.
Combinations of structures, such as matrix and cluster,
matrix and bundle, or bundle and cluster, are categorized
by the dominant fiber quality—cluster, bundle, or
matrix.
(8) Counting rules for single fibers are:
(a) Particulates meeting the definition of fiber are
isolated by themselves. With this definition, edge
view of flakes, fragments from cleavage planes, and
scrolls, for example, may be counted as fibers.
(b) Count as single entities if separation is equal to
or greater than the diameter of a single fiber.
(c Count as single entities if three ends can be seen.
(d) Count as single entities if four ends can be seen.
(e) In general, fibers that touch or cross are counted
separately.
(f) Two or more fibers are counted as a bundle if the
distances between fibers are less than the diameter
of a single fiber, or if the ends cannot be
resolved.
(g) Fibrils attached longitudinally to a fiber are
counted as part of the fiber and the size (width) is
estimated based on the fiber-to-fibril relationship.
(h) A fiber partially hidden by grid wires (one or two
ends) is counted, but labeled as an X-fiber. If the
number of X-fibers is more than 20% of the fibers
identified as asbestos, a larger-mesh EM grid should
be used, such as 100 mesh (about 200 pm wide).
(9) Sizing rules for asbestos structures are:
(a) For fibers, widths and lengths are obtained by
orienting the fibers to the inscribed circles on the
fluorescent screen. Since estimates are within
±1 mm, small-diamete-r fibers have greater margins of
error. Fibers less than 1 mm at the fluorescent
screen magnification level are characterized as
being 1 mm. A cylindrical shape is assumed for
fibers. X-fibers are sized by measuring their.
entire visible portions in the grid opening.
(b) Bundles and clusters are sized by estimating their
widths and lengths. The sum of individual diameters
is used to obtain the total width, and an average
length for the total length. A laminar-sheet shape
18
-------
is assuned, with the average diameter of the
individual fiber as the thickness.
(c) Matrices are sized by adding the best estimates of
individual fiber components. A laminar or sheet
structure is assumed for volume calculation.
(10) The method of sizing is as follows:
(a) An asbestos structure is recognized, and its
location in the rectangular "gate" relative to the
sides, inscribed circles, and other particulates, is
memorized.
(b) The structure is moved to the center for SAED
observation and sizing.
(c) Sizing is performed using the inscribed circles. If
the structure, such as a fiber, extends beyond the
rectangular "gate" (field of view), it is super-
imposed across the series of concentric circles
(several times, if necessary) until the entire
structure is measured.
(d) The structure is returned to its original location
by recall of the location, and scanning is
continued.
Figure A7 illustrates some of the counting and morphology guidelines used in
determining asbestos structures.
TEM Procedure—
The TEM procedure is as follows:
(1) EM grid quality is assessed at 250X.
(2) Particulate loading is assessed at 1000X.
(3) A grid opening is selected at random, examined at 1000X,
and sized.
(4) A series of parallel traverses is made across the grid
opening at the film magnification of 20.000X. Starting
at one corner, and using the tilting section of the
fluorescent screen as a "gate" or "chute," the grid
opening is traversed. Movement through the "gate" is not
continuous, but rather is a stop/go motion. On reaching
the end of one traverse, the image is moved the width of
one "gate," and the traverse is reversed. These parallel
traverses are made until the entire grid opening has been
scanned.
(5) Asbestos structures are identified morphologically and
counted as they enter the "gate."
19
-------
(6) The asbestos structure is categorized as fiber (with or
without X-) bundle, cluster, or matrix, and sized through
use of the inscribed circles.
(7) The structure (individual fiber portion) is centered and
focused, and the SAED pattern is obtained through use of
the field-limiting aperture.
(a) SAED patterns from single fibers of asbestos
minerals fall into distinct groups. The chrysotile
asbestos pattern has characteristic streaks on layer
lines other than the central line, and some
streaking also on the central line. Spots of normal
sharpness are present on the central layer line and
on alternate lines (that is, 2nd, 4th etc.) The
repeat distance between layer lines is about
0.53 nm.
(b) Amphibole asbestos fiber patterns show layer lines
formed by very closely spaced dots, and have repeat
distances between layer lines also of about
0.53 nm. Streaking in layer lines is occasionally
present due to crystal structure defects.
(c) Transmission electron micrographs and SAED patterns
obtained with asbestos standard samples should be
used as guides to fiber identification. An example
is the "Asbestos Fiber Atlas" (Mueller et al.,
1975).
(8) From visual examination of the SAED pattern, the
structure is classified as belonging to one of four
categories: (1) chrysotile, (2) amphibole group
(includes amosite, crocidolite, anthophyllite, tremollte,
and actinolite), (3) ambiguous (incomplete spot
^patterns), or (A) no identification. SAED patterns "~
""cannot be inspected for some fibers. Reasons for the
absence of a recognizable diffraction pattern include
contamination of the fiber, interference from nearby
particles, fibers that are too small or too thick, and
nonsuitable orientation of the fiber. Some chrysotile""*
^fibers are destroyed in t,he electron beam, resulting in
'patterns that fade away within seconds of being formed.
Some patterns are very faint and can be seen only under
the binocular microscope. In general, the shortest
available camera length must be used, and the objective
lens current may need to be adjusted to give optimum
pattern visibility for correct identification. A 20-cm
camera length and a 10X binocular are recommended for
inspecting the SAED pattern on the tilted screen.
20
-------
(9) Additional grid openings are selected, scanned, and
counted until either the total number of structures
counted exceeds 100 per known area or a minimum of 10
grid openings has been scanned, whichever is first.
( 10) The TEM data should be recorded in a systematic form so
that it can be processed rapidly. Sample information,
instrument parameters, and the sequence of operations
should be tabulated for ease in data reduction and
subsequent reporting of results. Figure A8 shows an
example of a data sheet used in Level I analysis.
Figure A9 illustrates the method of scanning a full-grid opening. The
"field of view" method of counting previously included in the provisional
methodology, which is based on randomly selected fields of view, has been
discontinued. Originally, the method was recommended for medium loading level
on the filter (50 to 300 fibers per grid opening). However, if samples are
collected at three different loading levels and the optimum is selected, this
medium loading on the filter will not be used. Samples with grid openings
containing 50 to 300 fibers may be used as laboratory fiber preparations or
selected source samples, but in field samples the particulate loading is
usually of much higher concentration than the fiber. Filter loading is
characterized by the particulate concentration, not by fiber concentration.
5. Data Reduction and Reporting of Results
Data Reduction—
From the data sheet, size measurements are converted to microns (16,OOOX~|
screen magnification), mass of asbestos structure is calculated, and other I
characterizing parameters are calculated through use of a hand calculator orI
\ computer. (Appendix B, an example of a computer printout from Level I
""analysis, shows reduced data—that is, what was found on the specified number
of grid openings or area examined.) These measurements are summarized and
?,related to the volume of air sampled and the total effective filtration area •
(area of deposit). Size measurements of X-fibers may be doubled and noted, or
kept as a separate category.
• ' Fiber number concentration is calculated from the equation
t L
„., . , Total no. of fibers
Flbers/ra3 = No. of EH fields
Total effective filter area, cm2
Area of an EM field, cm2
Volume of air sampled, m3
The number of X-fibers, bundles, clusters, and matrices are calculated in a
similar manner. X-fibers may be included with fibers if they are few in
21
-------
number. Similarly, their corresponding mass (from their size measurements)
may be included.
Fiber mass for each type of asbestos (chrysotile or amphibole) in the
sample is calculated by assuming that both chrysotiles and amphiboles have
circular cross-sections (cylindrical shape) and that the width measurements
are one diameter. The density of chrysotile is assumed to be 2.6 g/cm3, and
of amphiboles to be 3.0 g/cra3 . The individual mass is calculated from the
equation
Mass, yg = y x (length, ym) x (diameter,
x (density, g/cm3) x 10 6
The total mass concentration of fibers for each type of asbestos is then
calculated from the total mass of all the individual fibers of that type.
The individual masses of bundles, clusters, and matrices are calculated
by assuming a laminar or sheet-like structure with an average thickness of the
fiber make-up of the structure. Again, the density of chrysotile is assumed
to be 2.6 g/cm3, and of amphiboles to be 3.0 g/cm3. The individual masses are
calculated from the equation
Mass, yg = (length, ym) x (width, ym) x (thickness, ym)
x (density, g/cm3) x 10 6
The total mass for each type of structure for each type of asbestos is the sum
of all the individual masses.
Other characterizing parameters of the asbestos structures are: (1)
length and width distribution of fibers, (2) aspect ratio distribution of
fibers, and (3) relationships of fibers, bundles, clusters, and matrices.
Reporting of Results—
The data in their acquired and reduced forms are reported as summarized,
or, depending on the purpose of the analysis, are further reduced to present
the interrelationships of the various characterizing parameters. Again, the
Level 1 methodology is a monitoring or screening technique, and its limita-
tions, such as the possibility of "false positives" and misidentification,
should be noted.
6. Quality Control/Quality Assurance
Sampling procedures will vary depending on the type of sample, objectives
of the sampling, and time/cost factors. The primary goals of sampling are to
obtain a representative sample at the location and time of sampling, and to
maintain sample integrity. The sampling team will have written sampling
22
-------
procedures, and the field chief and/or designated individual will be
responsible for all record-keeping (including sample identification, labeling,
logging of data, site description, and meteorological conditions), pre- and
post-collection checks, and continuous sample custody and sign-outs until the
sample is delivered to the laboratory and transferred to the appropriate
quality assurance officer (QAO). Verification of sampling times, flow rates,
equipment calibration, and taking of field blanks will be checked arid recorded
in the field logbook.
Samples are turned over to the QAO for logging into a project logbook.
Each sample is carefully examined for gross features, such as tears, breaks,
and overall condition of container. The QAO registers the as-received sample
number and other designated information, and assigns a simple internal code
number that will accompany the sample through the preparation stage, grid
transfer, grid analysis, data reduction, and reporting of results.
After being logged into the project logbook, the sample is transferred to
the custody of the electron microscopy staff, where every precaution is taken
to maintain sample integrity and to prevent contamination and loss of
collected particulates. During storage and transport, the filters in their
respective holders are maintained in a horizontal position at all times.
The sample logging, handling, and storing procedures ensure that all
samples can be readily located and identified throughout the course of a
program. The QAO has divisional responsibility for QC/QA activities, and must
see that the laboratory maintains high standards. He must be aware of current
standards of analysis, and must ensure that internal quality control
standards, instrument calibration, and records of samples and completed
analyses are kept for ease of later retrieval and use.
For quality*control, internal laboratory blanks are analyzed at least
once a week, which may or may not coincide with a sample batch blank. In
addition, a magnification calibration of the EM using a carbon grating replica.—. ,
(2,160 lines per mm) is performed once a week. The results are recorded in-aiy£.;; ,
EM instrument log, along with other routine instrumental performance checks.
All photographs, TEM, SEM, and STEM images are recorded in a photo log. These^...' ..
.QCI"fesu'lts are documented for inspection by the QAO. 'v'v%;'-. •
23
-------
SECTION 5
LEVEL II ANALYSIS
SUMMARY OF PROTOCOL
I Level II analysis is a regulatory technique consisting of Level Ij
[analysis plus chemical elemental analysis. Morphology, size, SAED pattern,
~and chemical analysis are obtained sequentially. By a process of elimination,
mineral fibers are identified as chrysotile, amphibole, ambiguous, or "no-
J.dentity" by morphology and SAED pattern. X-ray elemental analysis is used to
| categorize the amphibole fibers, identify the ambiguous fibers, and confirm or
{validate chrysotile fibers.
Level II analysis is summarized as follows:
(1) A known volume of air is passed through a polycarbonate
membrane filter (pore diameter, 0.4 um; filter diameter,
37 or 47 mm) to obtain approximately 5 to 10 ug of
particulates per cm2 of filter surface.
(2) The particulate-laden filter is transported in its own
filter holder.
(3) The filte* is carbon-coated in the holder.
.-
(4) The particulates are transferred to an EM grid using a
refined Jaffe wick washer.
(5) The EM grid, containing the particulates, is gold-coated
lightly.
(6) The EM grid is examined under low magnification (250X to
pv.'lOOOX)- followed by high-magnification (16,OOOX on the
^iv fluorescent screen) search and analysis.
(7) A known area (measured grid opening) is scanned, and the
fibrous structures (fibers, bundles, clusters, and
matrices) are counted, sized, and identified as to
asbestos type (chrysotile, amphibole, ambiguous, or no
identity) by morphology and by observing the SAED pattern;
and finally by elemental analysis using EDS.
(8) The observations are recorded—a minimum of 100 fibrous
structures or 10 grid openings, whichever is first.
(9) The data are reduced and the results reported.
24
-------
EQUIPMENT, FACILITIES, AND SUPPLIES
The following items are required for Level 1 analysis:
(1) A modern 100-kV TEM equipped with an EDS. A scanning
accessory as found in a STEM will increase the versatility
and analytical capability for very small fibers and for
fibers adjacent to other particulate matter. The
microscope should be equipped with the fluorescent viewing
screen inscribed with graduation of known radii to
estimate the length and width of fibrous particulates.
(2) A vacuum evaporator with a turntable for rotating
specimens during coating, for such uses as carbon-coating
polycarbonate filters, gold-coating EM grids, and
preparing carbon-coated EM grids.
(3) An EM preparation room adjacent to the room housing the
EM. This room should either be a clean-room facility, or
contain a laminar-flow class-100 clean bench to minimize
contamination duing EM grid preparation. Filter handling
and transfer to EM grids should be performed in a clean
atmosphere. Laboratory blanks should be prepared and
analyzed weekly to ensure quality of work.
(A) Several refined Jaffe wick washers for dissolving membrane
filters.
(5) Miscellaneous EM supplies and chemicals, including carbon-
coated 200-mesh copper grids, grid boxes, and chloroform.
(6) Sample collection equipment, including 37-mm-diameter or
47-mm-diameter filter holders, 0.4-ym (pore size)
polycarbonate filters, 5.0-ym (pore size) cellulose ester
membrane filters for back-up, a sampling pump with
1 ,; ancillary equipment, a tripod, critical orifices or flow
••'"- '- meters, and a rain/wind shield.
DESCRIPTION OF METHODOLOGY - •-- "?*;:•••; -V'>V-;>:
• '..-'",'•;' '.' ^ '' '.•
1. Type of Samples—Source
This protocol is an expansion of the method originally developed for the
EPA for measuring airborne asbestos (Samudra et al., 1977; Samudra et al.,
1978). A broad interpretation of airborne has been to apply the terra to
samples obtained from ambient air (the original purpose), aerosolized source
materials (such as the asbestos workplace environment, and fugitive dust
emissions), bulk-air material (such as total suspended particulate (TSP)
samples, dust, and powders) and any other type of sample obtained by nonre-
strictive use of (1) collection of a volume of air, (2) separation from the
air, and (3) concentration of the particulates onto a substrate. The airborne
protocol has also been applied to samples collected in the regulatory areas of
the EPA, as compared with, for example, the workplace environment (National
25
-------
Institute of Occupational Safety and Health), mining activities (U.S. Bureau
of Mines), and shipboard atmosphere (Federal Maritime Administration).
The present methodology has been optimized for application specifically
to samples collected from a volume of air in which the asbestos concentration
is considered a minor component of the total particulate loading (other analy-
tical methods are available for samples known to contain high concentrations
of asbestos); and in which the particles are less than 15 ym in diameter,
since particles greater than 15 ym either are not inhaled or are deposited in
the upper respiratory tract and expelled, and preferably less than 10 ym in
diameter as recommended by the Clean Air Scientific Advisory Committee
(Hileman, 1981), since particles up to 10 ym can be absorbed by the alveolar
region of the lung. These concentration and size restrictions will preclude
many air samples collected in an asbestos-processing environment and in bulk-
air material from the complete methodology. However, such samples can still
be examined with the TEM, within the limitations of the instrument by changes
in preparation techniques—provided the effects on the final results, such as
fractionation of size and representativeness of the sample, are carefully
considered.
2. Sample Collection and Transport
Sample Collection—
Sampling procedures vary depending on the nature of the sample, purpose
of collection, analytical method to be used, sample substrate, and time and
cost of sample collection relative to the total analytical effort. Neverthe-
less, the primary objective of sample collection always is to obtain a
representative, unbiased sample.
Impingers, impaction devices, electrostatic precipitators, and thermal
precipitators ha~ve been used in sample collection, but each has limitations.
Presently, the preferred substrates are membrane filters, which are manufac-
tured from different polymeric materials, including polycarbonate, mixed
esters-of cellulose, polystyrene, cellulose acetate, and cellulose nitrate.. •.>;•;.!
Polycarbonate membrane filters differ from the others in being thin, strong,
and..smooth-surfaced, and in having sieve-like construction (circular pores ......
fr-om .top/ surface to the bottom). The other membrane "filters "are' thicker; have'y.'
irregular-surfaces, and have depth-filter construction (tortuous paths from
top surface to bottom).
., &
Consequently, polycarbonate filters have been selected for airborne
asbestos analysis. The collection of small-sized particles (prefer less than
10 ym in diameter), the light loading of particulates, the uniform distribu-
tion of particulates attainable using a depth-type backing filter, the smooth
surface and circular holes (which aid in determining size and instrument tilt
axis), and the relative ease in grid transfer (thin and strong) minimize
disadvantages of lack of retention and/or movement of large particles during
handling. Other membrane materials, such as the cellulose ester type, are
recommended for phase contrast and PLM, heavy particle loadings, and physical
retention of large particles.
26
-------
In microscopical analysis, uniformity of particulate distribution and
loading is critical to success. Air samples are taken on 37-mm-diaraeter or
47-mm-diameter, 0.4-ym (pore size) polycarbonate membrane filters using the
shiny, smooth side as the particle-capture surface. Cellulose ester-type
membrane filters (pore size, 5.0 um) are used to support the polycarbonate
filter on the support pad (37-mm-diameter personal sampler) or on the support
plate (47-mm-diameter holder).
Air monitoring cassettes (37-mm-diameter) of three-piece construction are
available from several manufacturers. As with the 47-mm-diaraeter filters,
loading the cassettes with the support pad, back-up filter, and 0.4 urn (pore
size) polycarbonate filter should be carefully performed on a class-100 clean
bench. Since the filters are held in place by pressure fit rather than by
screw tightening, air must not enter from the sides of the unit; a plastic
band or tape (which can double as a label) should be used as a final seal.
Collecting airborne samples with proper loading requires experience.
Each of the following techniques is useful in collecting airborne samples for
direct microscopy, preserving representative sizes, without diluting
particulate deposits:
(1) For long-term sampling at a site, test samples should be
returned to the laboratory by express mail service, or air
express service or by being hand-carried, and should then
be analyzed by scanning electron microscopy.
(2) The estimated particulate loading (deposit is barely
visible to the naked eye) should be bracketed by varying
the filtration rate and using the same time, or by varying
the time and using the same filtration rate.
(3) An automatic particle counter, such as a light-scattering
instrument (0.3-um detection) or a real-time mass monitor
,. . (0.1-ym detection), should be used to obtain an
.';:-,;r.--;:•• approximate particulate-loading level of the area.
Although any one of the three techniques will work, the suggested
'technique is to" take the samples as a set, varying ..the. .sampling rates and ''":.-'
using the same time so as to obtain filter samples with different particuiafe:V>v
loadings. Each set is composed of a minimum of four 37-mm-diameter or 47-min- -l;-
diameter filter units—three for different particulate loadings (low, medium,
high), and the fourth for a field blank. Suggested sampling rates are 0 for
the field blank, 2.48 L/min for the low loading, 7.45 L/min for the medium,
and 17.62 L/min for the high, for a 30 min sampling period using a 47-mra-
diameter filter holder. Simultaneous sampling will provide at least one
sample with a particulate loading suitable for direct EM analysis.
TSP's range from 10 ug/m3 in remote, nonurban areas, to 60 vig/ra3 in near-
urban areas, to 220 ug/m3 in urban areas. However, for heavily polluted
areas, TSP levels may reach 2000 ug/m3. A loading of 5 to 10 ug per cm2 of
filter is adequate for EM analysis; values beyond 20 to 25 ug per cm2 require
a dilution treatment. As an example, for 47-mm-diameter filters at face
velocities of 3.0 cm/s (2.48 L/min), 9.0 cm/s (7.45 L/min), and 21.2 cm/s
27
-------
(17.62 L/rain), respectively, air volumes of 74.4 L-; 223.5 L, and 528.6 L are
sampled in 30 min. For a TSP level of 200 yg/m3, 14.88 yg (1.07 yg/cm2),
44.7 yg (3.23 Ug/cm2), and 105.7 vg (7.63 yg/cm2), respectively, would be
collected on 47-mra-diameter filters (which would have effective filtration
areas of 13.85 cm2). The sampling time could be increased to 60 min for areas
having lower TSP levels, or reduced in a heavily polluted area (source
emissions).
Airborne samples from emission sources contain coarse particles (above
the respirable size) of large matrix structures, binder materials, road dust,
clay minerals, fillers, and other materials. For these samples, a fifth
filter unit can be added that has a size-selective inlet (cyclone, impactor, .
or elutriator) attached prior to the filter unit. The flow pattern and flow
rates of the tandem sampling arrangement must be checked before use. A
satisfactory, tested combination presently used>in California is a cyclone-
filter unit with a DSQ cut-off of 2.5 ym at 21.7 L/min, and a DSQ cut-off of
3.5 ym at 15.4 L/min (John and Reischl, 1980). Additional sampling devices,
such as impingers (used in biological sampling), impactors, and other
designated filter units (for TSP, XRD, or x-ray fluorescence (XRF), for
example) can be added to the system to obtain supplementary as well as inter-
related data.
This expandable multifilter sampling unit, designated Hydra, offers the
following advantages:
(1) It is small, inexpensive, and compact, so that an adult
can easily handle it.
(2) It is efficiently designed, and includes a tripod,
sampling pump, manifold, critical orifices, and a row of
preloaded 37-mm-diameter or 47-mm-diameter filter
holders. A rain/wind shield, size-selective cyclone-
filter units, tubing, and other extras can be added as
.."''needed. ' . ...,....,.....-,._.x
'•••'•• ' '• ' ' •• ,-.' .;••>-: Tr-.'-i-
(3) .Its sample preparation steps and handling are minimized. -••''•^•.'.•i1'.1
- • '-* "•' •
-:-(-4).r.;*ic.-allows complementary, as well as supplementary^ analysis
. '" ' . '::(TSP, size fractionation, bacteria, and XRF, for example), "
-------
Using 8 inch x 10 inch, or 102-mm-diameter filter sizes, is not recom-
mended. The sampling units are designed for purposes other than microscopy.
Interchanging the type of sample substrate filter (glass fiber or paper to
polycarbonate) does not correct the inherent problems of filter size and
sampling unit.
Sample Storage and Transport—
Once the sample is acquired, its integrity must be assured, and contami-
nation and loss of fibers prevented, until it is examined under the EM. The
low cost and small size of the 37-mm-diameter and 47-mm-diameter filter
holders enables them to be used as combination storage and transport con-
tainers. The filter holders should be maintained in a horizontal position
during storage and transport to the laboratory so that the particulate-loaded
filters can be removed under optimally controlled conditions in the labora-
tory.
For 47-mm-diameter holders (open-face) to be used in transport or
storage, the screw cap is carefully removed, and the shiny, waxy, stiff
separator paper used to keep the polycarbonate filters apart is carefully
placed on the retaining ring. The cap is then carefully screwed back on so
that the separator paper seals and protects the particulate-loaded filter
without touching it. The 37-mm-diameter, three-piece filter holder (aerosol
monitor) is used in its open-face position, and capped after usage for
transport and storage.
When the more expensive 47-mm-diameter holder is to be reused immedi-
ately, the particulate-loaded filter should be carefully removed and placed in
a 47-ram-diameter Petri-slide (such as that manufactured by the Millipore
Corp.*). This transfer takes place in the field rather than in the labora-
tory, so that the Petri-slide should be taken into the field. The 37-mm-
diameter filter holder or the 47-mm-diameter holder/Petri-slide should be
secured and all necessary sample identification marks and symbols applied to
the holder. .. ••• .--j
" ""*••'•.
3. Sample Preparation for Analysis—Grid Transfer
Carbon-Coating the Filter— ""'"-''
The polycarbonate filter, with the sample deposit and suitable blanks,
should be coated with carbon as soon as possible after sampling is
completed. To begin this procedure,* the' particulate-loaded 47-mm-diameter
polycarbonate filter is removed from the holder and transferred carefully to
an open-faced 47-mm-diameter Petri-slide for carbon-coating in the vacuum
evaporator (see Figure Al, Appendix A). If the 47-mm-diameter filter is
already in the Petri-slide, the cover is replaced with an open-face cover,
minimizing filter disruption. The 37-mm-diameter filter is left in the
holder, but the upper lid is removed to create an openrfaced filter. The
open-faced holders are placed on the rotating turntable in the vacuum
evaporator for carbon-coating. Figure A2 shows the multiple-coating
* Millipore Corp., 80-T Ashby Rd., Bedford, Mass. 01730
29
-------
I
arrangement in the evaporator; Figure A3 shows a close-up of the 37-mm-
diameter and the modified 47-mm-diameter holders for carbon-coating.
For archival filters and those of larger sizes, portions of about 2.5 cm
x 2.5 cm should be cut midway between the center and edge using a scalpel.
The portions are then attached with cellophane tape to a clean glass.
microscope slide and placed on the turntable in the vacuum evaporator for
coating.
Any high-vacuum carbon evaporator may be used to carbon-coat the filters
(CAUTION: carbon sputtering devices should not be used). Typically, the
electrodes are adjusted to a height of 10 cm above the level of the filters.
A spectrographically pure carbon electrode sharpened to a neck of 0.1 cm x
0.5 cm is used as the evaporating electrode. The sharpened electrode is
placed in its spring-loaded holder so that the neck rests against the flat
surface of a second carbon electrode.
The manufacturer's instructions should be followed to obtain a vacuum of
about 1.33 x 10~3 Pa (1 x 10~5 torr) in the bell jar of the evaporator. With
the turntable in motion, the neck of the carbon electrode is evaporated by
increasing the electrode current to about 15 A in 10 s, followed by 20 to 25 A
for 25 to 30 s. If the turntable is not used during carbon evaporation, the
particulate matter may not be coated from all sides, resulting in an undesir-
able shadowing effect. The evaporation should proceed in a series of short
bursts until the neck of the electrode is consumed. Continuous prolonged
evaporation should be avoided, since overheating and consequent degradation of
the polycarbonate filter may occur, impeding the subsequent step of dissolving
the filter. The evaporation process may be observed by viewing the arc
through welders goggles (CAUTION: never look at the arc without appropriate
eye protection). Preliminary calculations show that a carbon neck of 5 mm3
volume, when evaporated over a spherical surface 10 cm in radius,-will yield a
carbon layer that is 40 nm thick.
'•/Following carbon-coating, the vacuum chamber is slowly returned to
ambient .pressure, and the filters are removed and placed in their respective
.ho,lde.rs.,or .in, clean, marked Petri dishes for storage on a clean bench.
Transfer of the Sample to the EM Grid—
Transferring the collected particulates from the carbon-coated polycar-
bonate filter to an EM grid is accomplished in a clean room or on a class-100
clean bench. The transfer is made in a Jaffe wick washer, which is usually a
glass Petri dish containing a substrate to support the EM grid/carbon-coated
membrane filter combination. Solvent is added to a level to just wet the
combination and cause gentle dissolution of the membrane with minimum loss or
dislocation of the particulates, resulting in a membrane-free EM grid with
particles embedded in the carbon film coating. The substrate support can be
stainless steel mesh bridges, filter papers, urethane foams, or combinations
of these.
30
-------
The refined Jaffe wick washer is described as follows:
(1) The glass Petri dish (diameter, 10 cm; height, 1.5 cm) is
made airtight by grinding the top edge of the bottom dish
with the bottom of the cover dish, with water and
Carborundum* powder (80 mesh); this creates a ground-
glass seal (closer fit) and minimizes the need to refill
the Petri dish with added solvent. (The usual glass
Petri dish was found not to retain the solvent for long
periods of time, and unless the wicking substrate is kept
continuously wet, poor solubility of the membrane filter
results, leading to a poor-quality EM grid).
(2) A combination of foam and a single sheet of 9-cra filter
paper is used as the substrate support. A 3-cm x 3-cra x
0.6-cra piece of polyurethane foam (the packing in
Polaroid film boxes) is cut and placed in the bottom
dish. A 0.5-inch V-shaped notch is cut into the filter
paper; the notch is oriented in line with the side of the
foam, creating a well for adding solvent.
Spectrographic-grade chloroform (solvent) is poured into
the Petri dish through the notch until it is level with
the top of the foam (also level with the paper). The
foam will swell, and care is needed to avoid adding
solvent above the filter paper.
(3) On top of the filter paper, pieces of 100-mesh stainless
steel screen (0.6 cm x .0.6 cm) are placed, usually in two
rows, to make several grid transfers at one time (for
such uses .as replicas), and to facilitate maintenance of
proper identity of each transfer.
(4) A 3-mm section (usually midway between the center and
edge) of the carbon-coated polycarbonate filter is cut in
a rocking motion with a scalpel. The section may be a
square, rectangle, pr triangle, and should just cover the
3-mm EM grid.
r - •;'..t-"-"."N (5') A section is laid carbon-side down on a 200-mesh carbon-"
coated EM grid. (Alternatively, Formvar-coatedt grids or
uncoated EM grids may b§ used. Here, the carbon coating
on the polycarbonate filter'forms the grid substrate.)
Minor overlap or underlap of the grid by the filter
section can be tolerated, since only the central 2-mm
portion of the grid is scanned in the microscope. The EM
grid and filter combination is picked up at the edges
with the tweezers and carefully laid on the damp 100-mesh
* Carborundum is a registered trademark of the Carborundum Co., Carborundum
Center, Niagara Falls, N.Y. 14302.
t Formvar is a registered trademark of the Monsanto Company, 800 N. Lindbergh
Blvd., St. Louis, Mo.
31
-------
stainless steel screen. The EM grid-filter combination
will immediately "wet out" and remain on the screen.
(6) Once all specimens are placed in the washer, more solvent
is carefully added through the notch to maintain the
liquid level so that it just touches the top of the paper
filter. Raising the solvent level any higher may float
the EM grid off the mesh or displace the polycarbonate
filter section.
(7) The cover is placed in the washer and oriented in place
over the specimen, and a map of the filter/grid/screen
arrangement is made on the glass cover and in the
logbook.
(8) Solvent (chloroform) is added periodically to maintain
the level within the washer until the filter is
completely dissolved by the wicking action (24 to 48 h).
(9) The temperature in the room must remain relatively
constant to minimize condensation of solvent on the
bottom of the cover and subsequent falling of solvent
drops on the EM grid. Should day-night or other
temperature differentials occur, solvent condensation on
the under-surface of the cover can be minimized by
placing the Jaffe washer at a slight tilt (three glass
slides under one edge of the Petri dish parallel to the
row of grids) to allow the condensation drops to flow
toward the lower edge rather than fall on the EM grids.
At temperatures lower than 20°C (68°F), the complete
filter solution may take longer than 72 h.
(10) After- the polymer is completely dissolved, the stainless
steel mesh screen with the EM grid is picked up while wet
and set on lens paper tacked to the bottom of a separate
Petri dish. The EM grid is then lifted from and placed
next to the screen to dry. When all traces of solvent
have evaporated, the grid is stored in a grid box and
identified by location and grid box in the logbook.
Figure A4 illustrates the Jaffe wick washer method; Figure A5 shows the
washer. The foam/filter combination" is currently preferred, as is use of a
closely fitted (by means of the ground-glass seal) Petri dish.
Gold Coating—
An additional step will aid in subjectively evaluating the SAED pattern.
This step is required for specimens from the upper Great Lakes area and for
those of unknown origins. After the particulates on the filter are trans-
ferred to the EM grid, the grid is held to a glass slide with double-stick
tape for gold-coating in the vacuum evaporator. Several EM grids may be taped
to the glass slide with double-stick tape for gold-coating in the vacuum evap-
orator. For comparison, one-half of the EM grids may be coated and the other
one-half not coated; recognition of the gold-coating is helpful in searching
and x-ray analysis. Several EM grids may be taped to the glass slide for
32
-------
coating at one time. Approximately 10 mm of 0.015-cm-diaraeter (0.006-inch)
pure gold wire is placed in a tungsten basket (10 cm from the rotating table
holding the EM grids) and evaporated onto the grid.
The thin gold-coating establishes an internal standard for SAED analysis.
For some mineral species, an internal standard will clarify visual identifi-
cation of the pattern of a fibrous particulate as being or not being an
amphibole species (for example, minnesotaite as opposed to amosite). With
experience, differentiation in SAED patterns can be observed. For samples of
known geographic origins, gold-coating is optional, since the additional
coating hinders observation and identification of small-diameter chrysotile
fibers.
4. TEH Examination and Data Collection
Figure A10 shows a modern TEM jwi_th_capabilities for elemental., arialysis^
wiih__an_EDS.. The grid is observed in the TEM""a"f"magnifications of ^50X and
1000X to determine its suitability for detailed study at higher magnifica-
tion. The grid is rejected and a new grid used if: (1) the carbon film over
a majority of the grid openings is damaged and not intact; (2) the specimen is
dark due to incomplete dissolution of the polycarbonate filter; or (3) the
particulate loading is too light (unless a blank) or too heavy with particle-
particle interactions or overlaps.
TEM Analysis (Morphology, SAED, and X-Ray Analysis)—
The following guidelines are observed for consistency in the analytical
protocol:
(1) Magnification at the fluorescent screen is determined by
calibration with a diffraction-grating replica in the
specimen holder.
(2) A field of view or "gate" is defined.' On some .... .,
microscopes, the central rectangular portion of the .-.^-.?'
fluorescent screen, which is lifted for photographic
... ,:,',-., .purposes, is convenient, to use. On others, a scribed „ ";
; ;- circle or the entire circular screen may be used as the ' " " ' r;;iu-;'"
field of view. The area of the field of view must be
accurately measurable. ' -•••--
(3) The grid opening is selected on a random basis.'
(4) The analysis, morphology,.and SAED are performed at a
tilt angle of 0". -- -- -.•...: =
(5) The recommended instrument settings"are: accelerating
voltage, 100 kV; beam current,_100_yA; film.magnifi-
cation, 20.000X (which is equivalent to U.OOOX on' the
fluorescent screen for this instrument); and concentric
circles of radii 1, 2, 3, and A cm on the fluorescent
screen.
33
-------
(6) The grid opening is measured at low magnification (about
1000X).
(7) Since asbestos fibers are found isolated as well as with
each other or with other particles in varying arrange-
ments, the fibrous particulates are characterized as
asbestos structures:
Fiber (F) is a particle with an aspect ratio of 3:1 or
greater with substantially parallel sides.
Bundle (B) is a particulate composed of fibers in a
parallel arrangement, with each fiber closer than the
diameter of one fiber.
Cluster (Cl) is a particulate with fibers in a random
arrangement such that all fibers are intermixed and no
single fiber is isolated from the group.
Matrix is a fiber or fibers with one end free and the
other end embedded or hidden by a particulate.
Combinations of structures, such as matrix and cluster,
matrix and bundle, or bundle and cluster, are categorized
by the dominant fiber quality—cluster, bundle, and
matrix.
(8) Counting rules for single fibers, which are illustrated
in Figure A7 are as follows:
(a) Particulates meeting the definition of fiber are
isolated by themselves. With this definition, edge
view of flakes, fragments from cleavage planes, and
scrolls, for example, may be counted as fibers.
(b) Count as single entities if separation is equal to
or greater than the diameter of a single fiber.
1 ' (c) Count as single entities if three ends can be seen.
(d) Count as single entities if four ends can be seen.
'•' ' (e) In general, fibers that touch or cross are counted
separately.
(f) Two or more fibers-are counted as a bundle if the
distances between fibers are less than the diameter
of a single fiber, or if the ends cannot be
resolved.
(g) Fibrils attached longitudinally to a fiber are
counted as part of the fiber and the size (width) is
estimated based on the fiber-to-fibril relationship.
-------
(h) A fiber partially hidden by grid .wires (one or two
sides of the grid opening) is counted, but labeled
as an X-fiber (X-F) in the structure column. If the
number of X-fibers is high enough to affect the size
distribution (mass, etc.), a large-mesh EM grid
should be used, such as 100 mesh (about 200 ym
wide).
(9) Sizing rules for asbestos structures are:
(a) For fibers, widths and lengths are obtained by
orienting the fibers to the inscribed circles on the
fluorescent screen. Since estimates are within
±1 mm, small-diameter fibers have greater margins of
error. Fibers less than 1 mm at the fluorescent
screen magnification level are characterized as
being 1 mm. A cylindrical shape is assumed for
fibers. X-fibers are sized by measuring their
entire visible portions in the grid opening.
(b) Bundles and clusters are sized by estimating their
widths and lengths. The sum of individual diameters
is used to obtain the total width, and an average
length for the total length. A laminar-sheet shape
is assumed, with the average diameter of the
individual fiber as the thickness.
(c) Matrices are sized by adding the best estimates of
individual fiber components. A laminar or sheet
structure is assumed for volume calculation.
(10) The method of sizing is as follows:
(a) "An asbestos structure is recognized, and its
location in the rectangular "gate" relative to the
sides, inscribed circles, and other particulates is
-,'. memorized.
(b) The structure is moved to the center for SAED
.-•.:.• -observation and .sizing. „
" •—••-...,J*
(c) Sizing is performed using the inscribed circles. If:;
the structure, such as a fiber, extends beyond the
rectangular gate (field of view), it is superimposed
across the series of concentric circles (several
times, if necessary) until the entire structure is
measured.
(d) The structure is returned to its original location
by recall of the location, and scanning is
continued.
35
-------
Analytical Procedure—
The analytical procedure is as follows:
(1) EM grid quality is assessed at 250X.
(2) Parciculate loading is assessed at 1000X.
(3) A grid opening is selected at random, examined at 1000X,
and sized.
(A) A series of parallel traverses is made across the grid
opening at the film magnification of 20.000X. Starting
at one corner, and using the tilting section of the
fluorescent screen as a "gate" or "chute," the grid
opening is traversed. Movement through the "gate" is not
continuous, but rather is a stop/go motion. On reaching
the end of one traverse, the image is moved the width of
one "gate," and the traverse i's reversed. These parallel
traverses are made until the entire grid opening has been
scanned.
(5) Asbestos structures are identified morphologically and
counted as they enter the "gate."
(6) The asbestos structure is categorized as fiber (with or
without X-) bundle, cluster, or matrix, and sized through
use of the inscribed circles.
(7) The structure (individual fiber portion) is centered and
focused, and the SAED pattern is obtained through use of
the field-limiting aperture.
(a) SAED patterns from single fibers of asbestos
minerals fall into distinct groups. The chrysotile
^asbestos pattern has characteristic streaks on layer
lines other than the central line, and some
streaking also on the central line. Spots of normal
sharpness are present on the central layer line and
on alternate lines (that is, 2nd, 4th etc.) The ''•-.'
repeat distance between layer lines is about 0.53 nra.
(b) Amphibole asbestos fiber patterns show layer lines -.
formed by very closely spaced dots, and have repeat.;
distances between layer lines also of about 0.53 nm.
Streaking in layer lines is occasionally present due
to crystal structure defects.
(c) Transmission electron micrographs and SAED patterns
obtained with asbestos standard samples should be
used as guides to fiber identification. An example
is the "Asbestos Fiber Atlas" (Mueller et al.,
1975).
36
-------
(8) From visual examination of the SAED pattern, the struc-
ture is classified as belonging to one of four cate-
gories: (1) chrysotile, (2) amphibole group (includes
araosite, crocidolite, anthophyllite, tremolite, and
actinolite), (3) ambiguous (incomplete spot patterns), or^
(4) no identification. SAED patterns cannot be inspected
•for some fibers. Reasons for the absence of a recog-
,nizable diffraction pattern include contamination of the
fiber, interference from nearby particles, fibers that
are too small or too thick, and nonsuitable orientation
'of the fiber. Some chrysotile fibers are destroyed in
.the electron beam, resulting in patterns that fade away
within seconds of being formed. Some patterns are very
faint and can be seen only under the binocular micro-
scope. In general, the shortest available camera length
must be used, and the objective lens current may need to
be adjusted to give optimum pattern visibility for
correct identification. A 20-cm camera length and a 10X
binocular are recommended for inspecting the SAED pattern
on the tilted screen.
(9) The specimen holder is tilted for optimum x-ray detection
(40° tilt for the JEOL* 100C instrument's Tracer
Northern! NS 880 analyzer and Kevex* detector). The
categorized asbestos structure is maintained in its
centered position for x-ray analysis by means of the Z-
control.
(10)
(11)
(12)
The spot size of the electron beam is reduced and
stigmated to overlap the fiber. As an option for STEM
instruments, the electron beam may be used in the spot
mode 'and the x-ray analysis performed on a small area of
the structure.
• •«
The EDS is used to obtain a spectrum of the x-rays~~f
generated by the asbestos structure. \ '">""'"
=- —** i'; -".":
The profile of the spectrum is compared with prof ilesl ..VT^.-;
obtained from .asbestos standards; the best (closest) .;v<'; ,
match identifies and categorizes the structure. The"" '"'''
image of the spectrum may be photographed, or the peak
heights (Na, Mg, Si, Ca.'Fe) recorded for normalizing at
a later time. No background spectra or constant acquisi-
tion time is required since the shape of the spectrum
(profile) is the criteria. Acquisition of x-ray counts
may be to a constant time; to a constant peak height for
a selected element, such as silicon (1.74 keV); or just
!* !
* JEOL (U.S.A.) Inc., 11 Dearborn Road, Peabody, Mass. 01960
t Tracor Northern Inc., 2551-T.W. Beltway Hwy., Middleton, Wis. 53562
* Kevex Corp., Chess Dr., Foster City, Calif. 94404
37
-------
long enough to get an adequate idea of the profile of the
spectra, and then aborted. Figure All illustrates
spectra obtained from various asbestos standards and used
as referenced profiles.
(13) The specimen holder is returned to 0° tilt to examine
other asbestos structures.
(14) Scanning is continued until all structures are
identified, measured, analyzed, and categorized in the
grid opening.
(15) Additional grid openings are selected, scanned, and
counted until either the total number of structures
counted exceeds 100 per known area, or a minimum of 10
grid openings has been scanned, whichever is first.
(16) The TEM data should be recorded in a systematic form so
that they can be processed rapidly. Sample information,
instrument parameters, and the sequence of operations
should be tabulated for ease in data reduction and
subsequent reporting of results. Figure A12 shows an
example of a data sheet used in Level II analysis.
Figure A9 illustrates the method of scanning a full-grid opening. The
"field of view" method of counting, which is based on randomly selected fields
of view, has been discontinued. Originally, the method was recommended for
medium loading level on the filter (50 to 300 fibers per grid opening). How-
ever, if samples are collected at three different loading levels and the opti-
mum is selected, this medium loading on the filter will not be used. Samples
with grid openings containing 50 to 300 fibers may be used as laboratory fiber
preparations or selected source samples, but in field samples, the particulate
loading is usually of much higher concentration than the fiber. Filter load-
ing is characterized by particulate concentration, not by fiber concentration.
EDS is relatively time-consuming, and becomes redundant, if. used. as"-. " |
repetitive analysis for a confirmatory check on chrysotile fibers. Chrysotile
identity by morphology., and visual SAED analysis is not as controversiai^as :
amphibole identification and categorization. •'."."' ;'*"'. . ' "'
The following rules are recommended for EDS analysis (Level II):.
•,
(1) For chrysotile structure identification, the first five
are analyzed by EDS, then one out of every 10.
(2) For amphibole structure identification, the first 10 are
analyzed by EDS, then one out of every 10.
(3) For amphibole structure identification and categorization,
all confirmed amphiboles are analyzed by EDS.
(4) For ambiguous structure identification and categorization,
all are analyzed by EDS.
38
-------
Energy dispersive x-ray analysis as used in asbestos analysis is
seraiquantitative at best. X-ray analyzer manufacturers may claim quantitative
results based on calibration standards and sophisticated computer software,
but such claims are based on stoichioraetric materials and extension of work
with XRF instrumentation. Asbestos has a varying elemental composition. The
electron beam in an EM is of varying size, and not all instruments are
equipped to measure the beam current hitting the specimen. The size of the :
"specimen has an effect on the x-ray output, and nearby materials may fluoresce
and add to the overall x-ray signals being generated. Moreover, specimen
tilting results in a loss of x-ray acquisition from particles hidden by grid
wires or by other particles.
••«i
The only consistency in x-ray analysis is that the intensity of the
output, within restrictions, is proportional to the mass, therefore providing
the semiquantitative analytical possibility. Asbestos minerals .have been
found to have a characteristic profile, although not an exact duplicate of
each other. For example, the Mg:Si ratio of chrysotile may vary from 5:10 to
10:10, averaging about 7:10. The ratio can be used to confirm the morphology j
and visual SAED analysis. "
^—
Table 1 illustrates the phenomena of variability with resemblance for
some of the amphibole fibers. Peak heights and profile measurements were
taken.
To aid in the visual perspective of the spectrum profile, the peak
heights were normalized to a silicon value of 10, resulting in a five-number
series that is relatively easy to visualize—as in the following examples:
chrysotile ~ 0-7-10-0-0
treraolite ~ 0-4-10-3-
-------
TABLE 1. PROFILE COMPARISON OF ASBESTOS STANDARDS
Asbestos Type
Amoslte (GF-38A)
Anthophylllte (AF-45)
Crocldollte (CR-37)
,. •* '.T • '(- i-' • _'•'
Treoollte (T-79)
Size, u
0.19 x 1 .44 (stlpnate)
0.19 x 0.75 (STE»)
0.19 x 1.25
0.19 x 0.88 (100 s)
0.25 x 1.81 (100 s)
0.12 x 1.56
0.31 x 2.38
0.19 x 1.56
Repeat
0.56 x 2.38 (stlgmate)
0.31 x 2.38 (stlgmate)
0.31 x 5.19 (stlgmate)
0.19 x 1.56 (stipnate)
0.19 x 1 .88 (stlgmate)
0.19 x 0.81 (stleroate)
0.06 x 0.50 (stignate)
0.06 x 0.69 (stlpnate)
0.12 x 1 .00 (stlgmate)
Repeat (STEM)
0.12 x 0.62 (stlgaate)
0.12 x 1.12 (self-mate)
0.19 x 1 .56 (stlpraate)
0.06 x 1.69 (stlgmate)
Repeat (STEM)
Repeat (STEM)
.Repeat (STEM)
0.38 x 2.19 (stlpmate)
0.38 x 2.19 (spot) v
0.25 x 1.75 (stipmate)
0.25 x 1.75 (spot)
Repeat (stiRmate)
(STE«-inn S)
(STEM-inO s)
( STEM- 1 no s)
(STEM-100 s)
(STEM-100 s)
(STEM-40 s)
(STEM-40 s)
Na Kg
182
18ft
18]
226
57ft
253
256
27ft
• 477
631
64n
1064
507
787
131 inn
28 2R
37 35
44 53
70 64
56 65
53 5ft
76 83
45 4R
72 85
35 42
16 22
138
114
an
95
70
37ft
135
14S4
64
1072
4ft
12.1
Si
497
521
352
870
42n7
2049
2127
169ft
2945
2577
1670
3ftlO
2191
2286
885
205
17]
37«
612
479
326
735
290
892
373
166
368
327
197
252
211
1118
364
4Rin
1«1
3114
113
333
Ca Fe
38ft
387
289
674
3338
1515
1M3
lllft
1959
349
71
46ft
309
257
50,
115
9ft
204
333
2ftO
16ft
421
159
.463. .
237 .
104
03 ' - '\
8"
f>5
62
M
245
' 72
1235
48
BR2
27
41
Profile
0-4-10-0-R
0-4-]0-n-7
n-s- 10-0-1
p_3_,n_n-c
n-l-io-o-fi
p. j_j 0-0-7
o-i-in-n-8
0-2-10-0-7
0-2-10-0-7
0-2-10-0-1
n-4-m-p-n
o-3-m-o-i
o-2-in-o-i
O-3-in-n-i
2-1-10-O-ft
i-i-io-n-ft
2-2-in-n-ft
l-l-m-o-5
1-,-io-n.',
1-1-10-0-5
2-2-10-0-5
l-l-l o-n-f,
2-2-in-n-f,
l-l-in-O-5
l-l-!0-n-ft'.':.. ' '
i-i-io-n-6 , ., .
:: ,.$',*•- •;
o-i-io-2-n
O-i-jn-3-o
p-4-]o-2-n
n-3-10-;-n
i-3-in-2-n
n_i-m-2-n
n-3-in-3-n
n-3.]n-2-.T
O-l-in-3-n
n-i-io-:-o
ft-i-in-i-n
-------
(area of deposit). Size measurements of X-fibers may be doubled and noted, or
kept as a separate category.
Fiber number concentration is calculated from the equation
Fibers/m3 = Total no. of fibers
No. of EM fields
Total effective filter area, cm2
X Area of an EM field, cm2
1
Volume of air sampled, m3
The number of X-fibers, bundles, clusters, and matrices are calculated in a
similar manner. X-fibers may be included with fibers if they are few in
number. Similarly, their corresponding mass (from their size measurements)
may be included.
Fiber mass for each type of asbestos (chrysotile or amphibole) in the
sample is calculated by assuming that both chrysotiles and amphiboles have
circular cross-sections (cylindrical shape) and that the width measurements
are one diameter. The density of chrysotile is assumed to be 2.6 g/cm3, and
of amphiboles to be 3.0 g/cm3. The individual mass is calculated from the
equation
Mass, ug = -7 x (length, vm) x (diameter, utn)2
•T 1
x (density, g/cm3) x 10 6
The total mass concentration of fibers for each type of asbestos is then
calculated, .from .the. total .mass.of all the individual ..fibers of that type. , ... M
" The individual masses of bundles, clusters, and matrices are calculated
by assuming a laminar or sheet-likexstructure with an average thickness of the
fiber make-up of the structure. Again, the density of chrysotile is assumed
to be 2.6 g/cm3, and of amphiboles to be 3.0 g/cm3. The individual masses are
calculated from the equation
Mass, u§ = (length, ym) x (width, um) x (thickness, um)
x (density, g/cm3) x 10~6
The total mass for each type of structure for each type of asbestos is the sun
of all the individual masses.
-------
Other characterizing parameters of the asbestos structures are: (1)
length and width distribution of fibers, (2) aspect ratio distribution of
fibers, and (3) relationships of fibers, bundles, clusters, and matrices.
Reporting of Results—
The data and their subsequent reduction are reported as summarized, or
can be further reduced to present the interrelationships of the various
characterizing parameters. Figure A13 is an example of the EM data report;
Figure A14 is an example of the sample summary report.
The methodology can establish the limits of identity for unknown samples,
act as a QC/QA method for Level I analysis, and satisfy most of the
identification criteria for asbestos.
6. Quality Control/Quality Assurance
Sampling procedures will vary depending on the type of sample, objectives
of the sampling, and time/cost factors. The primary goals of sampling are to
obtain a representative sample at the location and time of sampling, and to
maintain sample integrity. The sampling team will have written sampling
procedures, and the field chief and/or designated individual will be respon-
sible for all record-keeping (including sample identification, labeling,
logging of data, site description, and meteorological conditions), pre- and
post-collection checks, and continuous sample custody and sign-outs until the
sample is delivered to the laboratory and transferred to the appropriate
quality assurance officer (QAO). Verification of sampling times, flow rates,
equipment calibration, and taking of field blanks will be checked and recorded
in the field logbook.
Samples are turned over to the QAO for logging into a project logbook.
Each sample is carefully examined for gross features, such as tears, breaks,
and overall condition of container. The QAO registers the as-received sample
number and other designated information, and assigns a simple internal code
number-that will accompany the sample through the preparation stage, grid --*-.
transfer, grid analysis, data reduction, and reporting of results.
After 'being logged' into the project logbook, the sample is "transferred do
the custody of the electron microscopy staff, where every precaution is taken..
to maintain sample integrity and to prevent contamination and loss of
collected particulates. During storage,and transport, the filters in their
respective holders are maintained in a horizontal position at all times.
The sample logging, handling, and storing procedures ensure that all
samples can be readily located and identified throughout the course of a
program. The QAO has divisional responsibility for QC/QA .activities, and must
see that the laboratory maintains high standards. He must be aware of current
standards of analysis, and must ensure that internal quality control
standards, instrument calibration, and records of samples and completed
analyses are kept for ease of later retrieval and use.
-------
For quality control, internal laboratory blanks are analyzed at least
once a week, which may or may not coincide with a sample batch blank. In
addition, a magnification calibration of the EM using a carbon grating replica
(2,160 lines per mm) is performed once a week. The results are recorded in an
EM instrument log, along with other routine instrumental performance checks.
All photographs, TEM, SEM, and STEM images are recorded in a photo log. These
QC results are documented for inspection by the QAO.
-------
SECTION 6
LEVEL III ASBESTOS ANALYSIS
DISCUSSION OF PROTOCOL
^
The Level III protocol is an extension of the Level II analysis proce-
dures described in Section 5. This extension may be necessitated by the need
for positive identification of. the specific amphibole species in situations
where (1) fundamental disagreements between parties involved in a litigation
require further clarification; (2) for identification purposes, e.g., as
causative agents in medical diagnosis or studies; (3) for quality control of
Level II analysis in special situations, and/or; (A) for source samples
whether as bulk material or bulk-air type where a legal judgment is antici-
pated.
Since an SAED pattern may be considered as a signature of the crystal
structure of the diffracting crystal (mineral fiber or particulate), the
mineral giving the pattern can be identified by comparison of measured and
standard sets of d-spacings and interplanar angles (0) from SAED patterns
obtained in near-exact zone axis orientations. Such identification, however,
may not be absolute without the provision of SAED patterns from more than one
zone-axis orientation.
The Level III analysis is an objective, confirmatory-type analysis and
consists of Level II analysis plus quantitative SAED analysis from two
different near-exact zone-axis orientations on a selected number of fibers
identified for detailed SAED analysis during the course of Level II analysis.
The Level III analytical procedure consists of locating the selected
fibers contained in gold-coated grid openings (for internal calibration);
photographing the fibers under bright-fsld illumination; obtaining (by
tilting) and recording two zone-axis SAED patterns from each selected fiber;
and obtaining (recording and photographing) representative EDS spectra from
the subject fiber.
The present Level III protocol is based on the following guidelines:
(1) Maintenance of procedural continuity so that results of
Level II analytical effort will aid in conducting the Level
III effort.
(2) Since detailed SAED analysis on all the fibers measured in
Level II analysis is not possible due to time and cost
restraints, a selection criterion is needed to assure
representative analyses.
44
-------
(3) The primary emphasis in Level III analysis is on the
positive identification of the amphibole type.
(4) The present protocol is designed to allow greater flexibil-
ity and freedom of decision for the microscopist in deter-
mining the selection criteria since, due to practical
constraints (position, orientation, contamination, etc.)i
all fibrous particulates may not be suitable for detailed
SAED work.
(5) It is recommended that approximately 20% (at least 10%) of
the fibers examined in Level II analysis be selected for
Level III SAED analysis. Fibers which would be classified
as "amphiboles" or "ambiguous" in Level II analysis should
be more often included for Level III analysis as compared
to those fibers which could be readily identified as "not
asbestos." In cases where the majority of the fibers in
Level II belong to a single, easily-identifiable species
(e.g., chrysotile), fibers that are different should be
more often selected for detailed Level III analysis. This
flexibility in selection criteria will maximize the gain
(meaningful information) from Level III effort beyond what
would be achieved from the analysis of 10-20% randomly
selected fibrous particulates.
(6) The electron microscope grids used in Level III analysis
(also Level II if Level III is anticipated) should be
finder grids so that location of fibers examined could be
referenced for quantitative SAED and for future re-
checking.
(7) Level III analysis should always be conducted by or under
the close supervision of a professional electron micro-
scopist knowledgeable in crystallography, SAED analysis,
mineralogy, plus Level I and Level II asbestos analyses.
If such expertise is not available in-house, an outside
consultant should be retained.
(8) If enforcement proceedings and possible legal involvement
may be part of the analytical procedure, the sample collec-
tion procedure entails additional record-keeping to
maintain sample integrity. The field crew chief or a
designated individual initiates, in addition to normal
QC/QA activities, a chain-of-custody record. The sample is
collected by the field team or by a representative of the
adversary party in the presence of each other, and is
sealed and signed for with the date and time. The desig-
nated individual acknowledges receipt of the collected
sample. In transferring the sample, the designate signs a
release of the sample in the presence of the new recipient,
who notes the date and time, and signs for acceptance in
the designate's presence. The chain of custody ensures
that only responsible pers-nnel have access to and control
of the sample, thereby avoiding the possibility of
45
-------
contamination before and after transport to the labora-
tory. At the laboratory, the QAO has first access to the
sealed sample container, and signs for it after obtaining a
signed release by the hand-carrier.
SUMMARY OF PROTOCOL
(1) An EM grid is prepared as directed in Level II analysis
using finder or locator grids instead of regular 200-mesh
grids.
(2) The particulate-loaded grid is then one-half or completely
coated with a thin layer of gold.
(3) The gold-coated grid is placed in a tilt-rotation or a
double-tilt specimen holder, and examined in the AEM or
STEM.
(A) At low magnification the specimen grid is examined, and a
grid opening is selected and identified for reference.
(5) Fibers identified for detailed Level III SAED work during
Level II analysis, employing the selection criteria
described under Level III guidelines, are now examined one
at a time.
(6) A bright-field image of the fiber is taken at 0° tilt and
at the magnification of analysis (20.000X).
(7) With the tilt-rotation or double-tilt combination, well-
defined SAED patterns of two different zone-axis orienta-
tions are observed and photographed. The fiber location
with respect to the edges of the grid opening or to other
particulates may prevent more than one zone-axis orienta-
tion from being obtained for some fibers.
(8) X-ray elemental analysis is taken of the fiber after the
SAED patterns. The EDS analysis also may be affected by
proximity of the fiber to the edge of the grid opening or
to other particles if tilting of the specimen is required
for efficient use of the EDS. An image of the spectra is
taken along with a record of the peak heights (the
presence of grid peaks, such as Cu or Ni, as well as gold-
coating may serve as markers).
(9) As explained earlier, due to time and cost considerations,
at least 10% (preferably 20%) of the fibers examined in
Level II are analyzed in Level III work.
(10) Those fibers whose EDS elemental analysis points to a
possible amphibole identification are selected for SAED
pattern indexing.
-------
(11) Parameters of interest obtained form zone-axis SAED
patterns are: the camera constant, CC (obtained from the
gold ring); the diffraction spot spacing dj (along the
slant vector), dj (along a row); the inter-row spacing, R;
and the interplanar angle 61,2. See Figure A18 for
details.
(12) The reciprocal lattice values of the d-spacings dj and d2
and the inter-row spacing (R) are converted into direct
lattice spacings and then dj, d2, R, and 61,2 are compared
to those of standard amphibole species listed in JCPDS
Powder Diffraction Files, values computed from lattice
parameters and crystal structures, or SAED Standard
Pattern File developed internally from known amphibole
minerals regulated by EPA.
EQUIPMENT, FACILITIES, AND SUPPLIES
Essential items required for a Level III analysis are:
• A 100-kV AEM equipped with the fluorescent viewing screen
inscribed with graduations of known radii to estimate the
lengths and widths of fibrous particulates; or a modern
100-kV TEM equipped with an EDS. A scanning accessory as
found in an STEM will increase the versatility and
analytical capability for very small fibers or for fibers
adjacent to other particulate matter. This microscope
should also be equipped with the fluorescent viewing
screen inscribed with graduations of known radii to
estimate the lengths and widths of fibrous particulates.
• A specimen holder with tilt-rotation or double-tilt
capability to obtain diffraction patterns at different
zone-axis orientations.
• Darkroom facilities for developing negatives, making
enlarged prints of patterns, and facilitating measurement
of distances, spots, lines, and circles.
• A vacuum evaporator with a turntable for rotating
specimens during coating, 'for such uses as carbon-coating
polycarbonate filters, gold-coating EM grids, and
preparing carbon-coated EM grids.
• An EM preparation room adjacent to the room housing the
EM. This room should either be a clean-room facility, or
contain a laminar-flow class-100 clean bench to minimize
contamination during EM grid preparation. Filter handling
and transfer to EM grids should be performed in a clean
atmosphere. Laboratory blanks should be prepared and
analyzed weekly to ensure quality of the work. In addi-
tion, a sample preparation room with a laminar-flow class-
100 clean bench should be available for handling bulk-air
47
-------
samples, ashing procedures, sedimentation, ultrasonifi-
cation, filtration, and other prefilter activities.
• Several modifed Jaffe wick washers for dissolving membrane
filters.
• Miscellaneous supplies and chemicals, such as membrane
filters, EM grids, films, gold wire, chloroform, and
carbon rods.
• Sample collection equipment, such as filter holders,
sampling pumps, critical orifices, and tripods.
DESCRIPTION OF METHODOLOGY
A detailed discusson of the morphology, crystallography and chemistry of
asbestos minerals, electron microscopy, and SAED analysis is outside the scope
of the present protocol. Basic knowledge in these areas and an adequate level
of comprehensive knowledge of TEM and SAED are prerequisites for the micro-
scopists participating in asbestos analysis, especially at Level III stage.
Since Level III analysis is an extension of Level II analysis, common
methodological details dealing with type of samples (source), sample collec-
tion and transport, sample preparation, TEM examination and data collection,
data reduction and reporting of results, and quality control/quality assurance
(QC/QA) program, which were discussed in detail in Section 5 (Level II
Asbestos Analysis) will not be repeated here and the users are advised to
refer to Section 5 for details in these areas. Differences, if any, between
Level II and Level III protocols in common areas have been dealt with earlier
under "Guidelines" and "Summary of Protocol."
The following provides brief descriptions of some of the essential areas
of the Level III protocol that were not covered under Level II protocol.
1. Crystallography and Morphological Properties
Both crystallographic and morphological characteristics of asbestos
minerals can help considerably in asbestos indentification and analysis.
Chrysotile displays a unique narrow tubular morphology. The amphibole
asbestos minerals have very similar morphologies—they are elongated along the
z-axis (the chain direction) and generally lie with (100) planes approximately
perpendicular to the electron beam. All varieties of amphiboles exhibit these
Wadsley faults parallel to the length of the fiber.
Chrysotile possesses a cylindrical lattice which produces a unique SAED
pattern. All the amphiboles, except anthophyllite, which is orthorhombic,
have a monoclinic crystal structure. The amphiboles are double-chain
silicates in which the fiber axis, z, has a repeat of 0.53 nm (inter-row
spacing 'R1 in real space, Figure A18). Since the other lattice parameters
are also very similar, detailed zone-axis SAED analysis in more than one
orientation is needed for positive identification. The non-asbestos forms of
amphiboles have properties very similar to their asbestos counterparts, thus
they must be distinquished from asbestos on the basis of morphology alone.
48
-------
2. Chemical Properties—Elemental Analysis by EDS
Araphiboles are nonstochioraetric minerals and often contain substitutional
cations in varying amounts. Therefore, precise determination of their chem-
istry is difficult and positive identification based on chemistry alone is not
reliable. This may be particularly pertinent when dealing with asbestos
minerals present as minor constituents in mineral samples.
Elemental ratios, which are sometimes used to distinquish between
asbestos types, often vary over wide ranges even in standard samples. The
presence of gold coating, which would tend to preferentially absorb x-rays
from lighter elements more than heavier elements, may make the situation even
worse. In view of these ambiguities, and due to inherent practical difficul-
ties in obtaining representative quantitative EDS elemental analyses from
submicroscopic fibers, the present Level II and Level III protocols specify
the use of only qualitative EDS spectra, which are often very valuable for
screening purposes in the Identification procedure. For example, in distin-
guishing between tremolite and actinolite type of amphibole, actinolite
usually contains Fe, but tremolite does not.
3. Selected Area Electron Diffraction (SAED)
The method of obtaining an SAED pattern of a randomly oriented specimen
is usually described in the EM instruction manual. The general directions for
using the instrument to obtain and photograph SAED patterns are:
(1) Select the image magnification for the selected area.
(2) Bring the desired field of view to the center of the
screen.
(3) Insert the appropriate field-limiting aperture (according
to the desired field of view) into the beam path.
(A) Obtain the sharpest field-limiting aperture shadow.
(5) Confirm that the desired field of view is in the field-
limiting aperture.
(6) Focus the specimen image; a photograph of the selected
area image can be taken.
(7) Obtain the SAED pattern, remembering to retract the
objective lens aperture from the beam path. The SAED
pattern will be observed on the fluorescent screen.
(8) Select the desired camera length (the shorter the length,
the better for SAED patterns of asbestos taken at high -.-..
magnification).
(9) Focus the SAED pattern sharply. The beam stopper is used
to intercept the bright center spot.
-------
(10) For photography, the illumination is expanded (condenser.
reduced) after focusing the pattern, so that the pattern
becomes barely visible (indistinct). A manual time
exposure of approximately 20 to 30 s (maybe more
depending on such factors as specimen and film) is
required. The beam stopper can be left in place or
removed from the beam path 1 to 2 s before closing the
shutter. A double exposure of the specimen image and the
SAED pattern can be taken if particle-to-particle spacing
is adequate.
4. Dse of Tilting to Acquire Exact Zone-Axis SAED Patterns
Determination of the Tilt Axis—
In the side-entry type electron microscopes, the instrument tilt axis is
always fixed. However, the position of the tilt axis on the viewing screen
shifts with magnification. Also, there is always an angular rotation between
the image and the SAED pattern. It is highly desirable to know the location
of the tilt axis on the viewing screen and its relationship vis-a-vis SAED
pattern under the operating conditions to make effective use of specimen
tilting for obtaining exact zone-axis orientations. The following steps can
be used to locate the position of the tilt axis:
(1) A gold-coated EM grid with a standard asbestos mineral
specimen on a polycarbonate replica film is placed in a
tilt-rotation or double-tilt holder and inserted at 0°
tilt into an aligned TEM set at 100 kV, 100 uA, 20.000X
magnification, and 20-ym camera length operation.
(2) The image is focused on the fluorescent screen, which is
at approximately 16.000X magnification.
(3) A circular hole in the polycarbonate replica is positioned
in the center of the field of view.
(4) On tilting, the circular feature changes to an ellipse
with the major axis unchanged, and indicates the position
(direction) of tilt axis at that magnification. The minor
axis shows the perpendicular direction to the tilt axis.
A high tilt angle defines, the tilt axis more accurately
than a small tilt angle. Figure A15 illustrates the effect '
of tilt.
•
(5) A double-exposure photograph at 0° tilt and at some high
tilt angle, such as 30°, is taken of the focused circular
hole for reference.
Tilting—for zone-axis SAED Patterns—
Quantitative SAED requires knowledge of crystallography to obtain useful
zone axis diffraction patterns from which precise measurements can be made for
comparison with known asbestos standards on file. Thus the method of obtain-
ing the visual SAED pattern of randomly oriented specimens, as in Level I and
II analysis, is modified for quantitative SAED pattern analysis. It requires
tilting of the specimen to align major crystallographic directions with the
50
-------
electron beam. The zone axis is a line parallel to a set of intersecting
crystal planes and nearly parallel to the electron beam. A zone-axis pattern
thus gives regular repeat distances and even intensities of spots throughout
the pattern.
Either a double-tilt or a tilt-rotation type specimen holder can be used
for obtaining zone-axis patterns. A double-tilt holder is often preferred
because tilt-rotation combination involves translational movement of the fiber
during tilting, necessitating constant adjustment of the specimen-positioning
controls to keep the specimen centered in the SAED aperture. On the other
hand, it is much easier to obtain an accurate measure of the degree of tilt
and perform systematic tilting with the tilt-rotation specimen holder. It is
only necessary to rotate the specimen (fiber) until the tilt axis (as deter-
mined earlier) coincides with a major row of spots and then tilt until a major
zone axis is parallel to the incident electron beam. Alternately, fiber axis
of the fiber can be oriented either parallel or perpendicular to the tilt axis
and then further tilting is used to obtain exact zone-axis orientations.
In order to avoid flip-flopping between image and diffraction modes while
tilting, a recommended procedure is to defocus the diffraction pattern (the
aperature becomes visible and the specimen/fiber can be seen in it) so that a
double image of fiber in aperture can be seen with a poorly focused diffrac-
tion pattern. The movement of the fiber can then be tracked in relation to
the spot pattern during tilting and kept centered in the SAED aperture by use
of the specimen-positioning controls (knobs) of the microscope. Sometimes a
larger aperture aids in the tracking-pattern recognition process.
An experienced electron microscopist can readily recognize the geometri-
cal features like Klkuchi lines or Laue zones in the SAED pattern and use
these to obtain the exact zone-axis SAED patterns. A detailed discussion of
Kikuchi patterns and Laue zones and their utility in tilting experiments may
be found in any standard text book on electron microscopy. Use of the double-
tilt specimen holder is very helpful and less tedious in tilting experiments.
However, all laboratories may not have both types of specimen holders avail-
able. A skilled microscopist can use either specimen holder without much
difficulty. Experience and skill are more important factors in SAED analysis
than the type of specimen holder used.
5. Characteristics of SAED Patterns Encountered in Asbestos Analysis
Successful application and exploitation of SAED analysis in asbestos
analysis needs prior knowledge of the general appearance and distinguishing
characteristics of other SAED patterns which are often encountered. . The
following discussion summarizes some of the observed SAED features of asbestos
and other related minerals. This discussion is by no means comprehensive and
assumes that the reader is familiar with general crystallography and the
nomenclature pertaining to varous aspects of SAED patterns.
Minnesota!te and Stilpnomelane—
These iron-rich non-asbestos layer minerals are often encountered in
asbestos analysis of specimens from certain geographic locations. Particu-
lates of these minerals lie near their basal (001) planes. Stilpnomelane and
51
-------
minnesotaite both possess large superlattices and their commonly observed SAED
patterns are easily distinguishable from araphibole patterns. The spacing (in
reciprocal space) is about half (for minnesotaite) or less than that for most
amphiboles. These minerals can be readily distinguished in Level I or Level
II analyses if a gold coating (optional) is applied to the specimen grids. A
visual inspection of the number of rows of spots inside the (111) gold ring is
sufficient to distinguish minnesotaite and stilpnomelane from amphiboles.
Chrysotile—
Due to the cylindrical lattice of chrysotile the SAED pattern is unique.
The SAED pattern observed is symmetrical about the cylinder axis, x, and the
spacing of the rows of spots is proportional to I/a, where a_ is 0.53 nm. The
most distinguishing features of the pattern are the flared spots of the type
(130) which occur in the firt layer line. The flaring is due to the cylin-
drical lattice. A typical EDS spectra shows the presence of only Mg and Si
(Figure All).
Amphiboles—Systematic Absences, Twinning, and Double Diffraction—
The most commonly observed row of diffraction shots found in SAED
patterns in araphiboles is in the y* or b* direction, representing the shortest
reciprocal spacing between the spots (18.4 A in real space). There are many
strong zone axis orientations containing the y* row of spots. The lattice of
amosite, crocidolite, tremolite, and actinolite is c-centered, and for such a
lattice the h + k odd spots are absent along the y* or b* row. In practice,
however, weak spots may be present in forbidden positions due to the presence
of thin multiple twinning on (100), which cause streaking parallel to a*.
Often, reciprocal nets from both twins are present in the same SAED pattern.
In a twinned crystal, the number of important diffraction nets containing b*
is doubled, leading to the observation that the diffraction patterns appear
insensitive to tilt.
In some cases SAED patterns can contain spots from both twin individuals
which overlap. However, not all the spots present in the composite SAED
patterns are generated by the overlapping nets; some spots may be present
because of double diffraction where a diffracted beam from one twin becomes
the transmitted beam when it enters the other twin.
The purpose of the above discussion is to point out that although many
complications exist in the analysis of SAED patterns, these can be overcome;
in a good goniometric tilting stage most amphiboles can be identified by SAED
analysis.
Amosite—
The nearest reciprocal lattice section to the (100) direct lattice plane
in amosite is (301)* and it is also the most commonly observed section. Due
to the presence of the thin (100) twins, this section closely resembles (100)*.
Typical EDS spectra from amosite fibers (Figure All) show mainly Si and
Fe with smaller amounts of Mg and Mn. Mn is frequently observed as a substi-
tutional cation in amosite.
52
-------
Crocidolite—
Most of the commonly observed patterns are asymmetrical and cannot be
indexed easily. However, they all show rows of spots separated by a
reciprocal repeat (R) corresponding to the fiber axis (0.53 nm).
The main elements observed in typical EDS analysis are Mg, Si, Ca, and
Fe. Na, which is usually present in crocidolite, may not be detected in gold-
coated specimens because of absorption, or because of overlapping secondary
peaks from the copper grid.
Tremollte-Actinolite—
Temolite and actinolite show a variety of SAED patterns which have very
similar appearances. In actinolite some of the Mg is replaced by Fe, with the
result that interplanar d-spacings of actinolite are slightly larger than
tremolite. In both tremolite and actinolite, the main elemental constituents
are Mg, Si, and Ca. Actinolite also contains some Fe.
Anthophyllite—
Even though anthophyllite has an orthorhombic crystal structure, its
commonly observed patterns are similar to the monoclinic amphiboles. Antho-
phyllite fibers dehydrate more easily in an electron beam and are, therefore,
more difficult to study.
EDS elemental analysis shows the main constituents to be Si and Mg with a
small amount of Fe.
6. Determination of Camera Constant and SAED Pattern Analysis
As mentioned earlier, a thin film of gold is evaporated on the specimen
EM grid to obtain zone-axis SAED patterns superimposed with a ring pattern
from the polycrystalline gold film. Since d-spacings corresponding to identi-
fiable gold rings are known, these can be used as an internal standard in
measuring unknown d-spacings on an SAED pattern from a fiber. The precision
of measurement is as good as the quality of the photograph (or negative) and
usually the measurements should be in the order of 0.1-0.2 mm with an angular
tolerance of 0.5-1.5 degrees. The measurements can be made by several
methods: manually with a ruler, with a mechanical aid, or a densitometer,
etc. The patterns can be read directly on the developed negative or on an
enlarged non-glossy print.
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 SAED
pattern. Thicker gold film would normally give multiple gold rings, but it
will tend to mask weaker diffraction spots from the unknown fibrous particu-
lates. Since the unknown d-spacings of most interst in asbestos analysis are
those which lie closest to the transmitted beam, multiple gold rings are
unnecessary on zone-axis SAED patterns.
7. Determination of Camera Constant Dsing Gold Rings
An average camera constant using multiple gold rings can be determined as
explained below. However, in practice, in most cases determination of the
53
-------
average camera constant is not necessary and thicker gold films are not
desirable. The camera constant, CC, is 1/2 the diameter, D, of the rings
times the interplanar spacing, d, of the ring being measured and is expressed
as :
CCdnm-JL) = x d(0
The value of d for each ring can be obtained from the JCPDS file.
(a) Measure the diameters (two perpendicular locations) of the
gold rings in mm as precisely as possible (see Figure
A16).
(b) Measure as many distinct rings as possible to minimize
systematic errors.
(c) Example: if the measured values in mm are Dj , T>i, 03, D^ ,
and DS , these will represent, respectively, d-spacings of
4.079 A. 079 A. 079 4.079 4.079
, — » ; » , — > / — » and / — A
/3 2 /8 /u /12
(d) The camera constants will be:
l 4.079 l
CC, = -=- x -^2- = ^- x 2.355
D2 4.079 °2
CC2 = — x -^~ = — x 2.04
D3 4.079 °3 , ...
CC3 = -=- x = -=— x 1.442
/ff
.
CCU = ^-x = x 1.23
2 /H 2 -•
D5 .
CC5 = .^-x i = _x 1.178
(e) The camera constant f or the SAED patterh'is the average of
CCi , CC2 , CCa , CCi, , and CC$ . Table 2 presents an example
of camera-constant determination.
54
-------
TABLE 2. DETERMINATION OP CAMERA CONSTANT (EXAMPLE)
Ring
No.
1
2
3
4
Dj readings
(am)
23.0,
27.4,
37.8,
44.6,
22.0
27.6
38.2
45.4
Mean Dj
(mm)
22.5
27.5
38.0
45.0
d-spacing, di
(A)
2.355
2.04
1.44
1.23
Camera constant
Cj = Dt/2 x dt
26.5
28.0
27.4
27.7
Mean Value of L i 26.5 + 28.0 + 27.4 + 27.7 my/ IN
Camera Constant = ~ = 5 = 27>4 (mm~^
8. Measurement of d-Spacings and Interplaaar Angles
The gold film, because of its small, randomly oriented crystallites,
produces a ring pattern superimposed on the SAED pattern from the fibers. The
diameters of the gold rings correspond to known values of d-spacings, and this
provides an internal standard to correct for inherent uncertainties present
due to variations in instrumental and/or operating conditions. Since the d-
spacings of interest on SAED patterns are usually the ones that lie closest to
the center spot (transmitted beam), a camera constant measured from the first
gold ring in the direction of measurement of d-spacings will usually give
better accuracy in computed spacings than the use of an average camera
constant. This method will account for any distortions in the symmetry of the
gold ring pattern. The zone-axis SAED pattern usually has several rows of
spots within the circular pattern of the gold rings. These rows of spots
contain information about the two sets of planes in the crystal structure and
the angle between them. The following procedure outlines the steps necessary
to obtain the distances between planes (d-spacings) and the corresponding
interplanar angle, 6 (see Figure A17):
(1) From the spot pattern, determine the row with spots most
closely spaced, and designate this as a horizontal row.
Draw a fine line to show the row through the origin, and
designate this the zeroeth row. Draw fine lines to show
the first and succeeding horizontal rows. For a few
horizontal rows, measure the mean spacing between adjacent
spots (or the minimum vector):
Distance between spots m units apart
A, = _ _ _ —- - - - - - _ _
i m
where m is chosen as an optimum number to minimize
measurement errors. The mean horizontal spot distance, X,
55
-------
equals the summation of Xi divided by the number, n, of
rows measured. The d-spacing in A corresponding to this
vector is the camera constant divided by X, and is labeled
d2• Table 3 presents an example of spot spacing
measurement within a horizontal row.
(2) The perpendicular distance between two adjacent horizontal
rows is similarly measured. This interrow spacing, Z, is
the mean separation between horiziontal rows, and equals
the distance between a number of rows divided by the
number of spaces. This distance is an additional vector
for comparison that coincides with the slant vector, di~
spacing, when angle 81,2 is 90°. The row-spacing 00
equals the camera constant divided by Z. Table 3 presents
an example of perpendicular spacing between horizontal
rows; Figure A17 illustrates spot and row spacing.
(3) To obtain the d\-spacing and corresponding angle 9i,2» a
perpendicular is drawn to the zeroeth horizontal row
through the origin. A line is drawn to the first spot to
the right of the perpendicular in the first row and
extended through the succeeding rows. This line, called
the slant vector, forms the acute angle 9i,2« The mean
spacing, Y, between spots on the slant vector can be
measured by dividing the maximum distance between spots by
the number of spaces between them, or by calculating from
the interrow spacing:
sin G! ,2
The d-spacing in A corresponding to this vector is the
camera constant, CC, divided by Y and labeled dj.
CC x sin 61 ,2
Figure A18 illustrates the relationship of dj, d2, 61,2
and R. In some cases, the interplanar angle 81,2 nay be
more than 90 degrees (not shown in Figure A18).
Summary of Data from Each SAED Pattern:
(a) The camera constant, CC, as determined from the gold
rings, normalizes the distances on the SAED pattern
regardless of such factors as magnification and tilting.
56
-------
TABLE 3. DETERMINATION OF SPOT SPACINGS (EXAMPLES)
Separation Mean spacing,
Reading (mm) Units XL (a)
Spot spacing within a horizontal row, dj :
1 49 16 3.006
2 42.7 14 3.05
3
3.028 = Mean
27 4
d-spacing = '- = 9<05
Perpendicular spacing between horizontal rows, R:
43 8 5.0375
5.0375 = Mean
d-spacing, R = ' = 5.44 A
Note: It is preferable that the camera constant
values used in computing d-spacings are measured
from the first one or two gold ring diameters in
the direction of d-spacing measurement.
57
-------
(b) The parameters of interest are:
0 d-spacing of spots in a horiziontal row: CC/X = 62
• d-spacing of spots in the slant vector: CC/Y = di
• angle 01>2 formed between a horizontal row and slant
vector
• d-spacing corresponding to row separation as an
additional parameter of interest: CC/Z = R.
It should be noted that the use of camera constant in the form used here
in calculating di, 62, and R, which are measured in reciprocal space on SAED
patterns, automatically converts the calculated numbers into real space
spacings, which are then compared to those from a suitable standard file.
9. Identification of Unknown Fibers
Unknown d-spacings (di and d2>, interrow spacing (R), and interplanar
angles (9) measured from zone-axis SAED patterns of unknown fibers are
compared with corresponding known values tabulated in JCPOS powder diffraction
files, or those computed using lattice parameters and crystal structures of
candidate asbestos minerals, or with the values contained in an internally
developed file from standard specimens of candidate minerals. Table 4 is an
example of the TITRI standards file (Jones et al., 1981). Figures A19 to A22
are examples of zone-axis SAED patterns.
Unknowns are matched as closely as possible to the file parameters for
positive identification. However, considerable care and competent judgment
are required in Level III confirmatory analysis. For example, amphiboles are
usually nonstochiometric minerals, and thus a perfect match may not be possi-
ble between the d-spacings and interplanar angles determined from unknown
fibers and those available from standard minerals. JCPDS Powder Diffraction
files do not list interplanar angles. Since amphiboles have low-symmetry
crystal structures, tabulated values of d-spacings and interplanar angles
would be extensive and very expensive to generate, and to get an accurate
match may not be possible because these tables are derived assuming certain
lattice parameters which may not be the same as those of the unknown fibers
being analyzed. Given these inherent uncertainties, it would seem that use of
internally developed SAED files consisting of several readily accessible
orientations (by virtue of natural habit of amphibole fibers) from standard
amphibole species could eliminate a lot of tedious unnecessary work and yet
provide reliable data for comparison and identification of unknown fibers.
In practice, SAED analysis combined with qualitative EDS analysis may
help resolve certain cases where a close match in d-spacings and interplanar
angles is not possible. For difficult specimens or SAED patterns of contro-
versial nature, a second opinion may be necessary, especially if a legal case
is involved. . ."".
58
-------
TABLE 4. COHPARISION OP d-SPACINGS FROM SAED FILE
AND POWDER DIFFRACTION FILE (EXAMPLE)
Internal Standard File Data
AraphLbole
type
Amosite
Crocidolite
Tremolite
Anthophyllite
Zone
axis
flOO]
[307]
[101]
[Toil
[310]
[100]
[101]
[Tio]
[30T]
[3ioj
[100]
[101]
[loT]
[30T]
[100]
[T42]
dl
(A)
5.3
1.79
4.88
4.14
5.22
5.22
4.94
4.79
1.75
5.12
5.04
4.83
2.59
1.72
—
4.56
d2
(A)
9.14
9.26
9.23
9.11
5.13
8.97
9.05
8.19
8.97
5.12
9.03
9.03
8.97
8.98
—
4.56
9
(deg)
90.0
84.0
74.0
78.0
95.0
90.0
75.0
79.0
83.5
96.0
90.0
75.0
80.5
83.5
90.0
60.0
Interrow
spacing, R
(«)
5.3
—
5.17
4.21
—
5.22
5.19
5.23
—
—
—
—
—
—
5.24
—
Powder Diffraction
File Data (1975)
dl
(A)
5.22
1.76
4.84
4.10
5.22
5.20
5.89
4.89
1.76
—
5.07
4.87
2.59
1.69
5.28
4.50
d2
(A)
9.20
9.20
9.20
9.20
5.12
9.02
9.02
8.40
9.02
—
8.98
8.98
8.98
8.98
8.90
4.50
File
index
no.
17-725
17-725
17-725
17-725
17-725
19-1061
19-1061
19-1061
19-1061
19-1061
13-437
13-437
13-437
13-437
9-455
9-455
59
-------
SECTION 7
ARCHIVAL SAMPLES
DISCUSSION OF PROTOCOL
Samples that have been collected on filter substrates other than poly-
carbonate, or that have been collected without regard to filter loading
levels, are referred to as archival samples. These samples were usually
collected for other analytical objectives, such as for optical microscopy or
gravimetric analysis, for defined sampling periods without regard to concen-
tration levels in the air, or for collection of particles larger than 10 um in
diameter. Such samples were historically collected, and are of value and
interest in determining the presence of asbestos fibers and/or structures.
Filter substrates designated as archival samples include glass fiber filters;
cellulose or modified paper filters; cellulose ester filters; other organic
polymeric membranes, such as polystyrene, nylon, and polyvinyl chloride; and
all overloaded organic polymeric membrane filters.
The purpose of the preparation step is to transfer particles from a
filter surface to an EM grid with a minimum of distortion in morphology and
size distribution. The nature of non-polycarbonate filter substrates or
'particle loading makes it sometimes necessary to transfer a satisfactory
quantity of particles to a polycarbonate filter prior to transfer to the EM
grid. At present, only transfer to an EM grid from a polycarbonate filter has
been standardized.
A modified preparation technique is recommended for archival samples,
followed by the analytical methodology using Level I, Level II, or Level III
effort—with the understanding that these samples will indicate the presence
of asbestos, and secondarily the number, size, distribution, and morphology.
The results from sample to sample are less precise due to problems in
standardizing the preparation procedures used for archival samples.
The archival filter samples are prepared for analysis based on the
information sought, type of filter material, and particle loading on the
filter. The various preparation techniques for these filters include:
(I) Individual particle picking and/or reverse washing of the
filter, with subsequent filtration of the filtrate using a
polycarbonate filter.
(2) Collapsing the membrane filter structure by exposure to
solvent vapor (surface fusion), to produce a more uniform
substrate for replication and grid transfer.
60
-------
(3) Solubilizlng filter material in selected solvents,
followed by separation of participates.
(4) Low temperature ashing (LTA).
Two preparation methods are recommended based on filter loading and type of
filter material: surface fusion, and LTA.
DESCRIPTION OF METHODOLOGY
Because the greatest number of archival samples have cellulose ester
substrates, this type of filter material is used in examples describing the
methodology.
1. Samples with Adequate Loading
Discussion—
As an example, samples collected on cellulose ester filters have been
received by a laboratory. Direct transfer of the particulates to the EM grid
is possible using acetone as the solvent in a modified Jaffe wick washer.
However, a question arises concerning indeterminate particle loss in the
transfer. Carbon-coating the cellulose ester filter prior to grid transfer
minimizes particle loss. However, this Improvement in particle count is
offset by difficulty in visually observing and counting the fibrous particles
against a replica background of the uneven surface topography of the cellulose
ester filter, and by indeterminate loss of very small particles hidden in the
crevices of the uneven filter surface. The NIOSH method of surface fusion
(Zurawalde and Dement, 1977) is relatively reliable, and produces a more
consistent result, although the question of loss of the very small particles
has not been resolved. LTA, described later in this section, may also be used
for these samples.
Procedure—
The NIOSH technique, a modification of a particle-transfer technique
developed at Los Alamos Scientific Laboratory (Ortiz and Isom, 1974), is
described as follows:
(1) A section of the membrane filter is cut with a scalpel, and
placed on a clean microscope slide with the sampled side
facing up.
(2) The cut section is fastened on all sides to the slide with
narrow strips of transparent tape.
(3) The slide, with the cut section, is exposed to acetone vapor
(not liquid) for approximately 10 min. The acetone vapor
collapses the structure of the filter and produces a fused,
relatively smooth-surfaced film. The size of the acetone
vapor bath and time of filter response to the vapors are
critical in obtaining the desired smooth, fused surface;
each laboratory must determine its own optimum conditions.
(4) The fused filter section is placed on the rotating stage of
the vacuum evaporator for carbon-coating.
61
-------
(5) A 3-mm-diameter portion of the carbon-coated filter is
transferred to a carbon-coated EM grid in the modified
Jaffe wick washer.
(6) Acetone is used in dissolving the fused membrane filter.
(7) Transfer to the grid and options for analytical efforts
were described previously.
2. Samples with Heavy Loading
Discussion—
As an example, samples collected with a heavy deposit of particulates
have been received by a laboratory. These particulates may be organic in
nature (for example, pollen or soot), or of mineral matter. LTA is used to
remove the organic material (filter as well as particulates), leaving the
inorganic residue. The residue is gently resuspended and dispersed in
filtered distilled water by low-wattage, short-time ultrasonification. The
resuspension is then filtered onto a O.l-um (pore size) polycarbonate
filter. The dry, particulate-loaded polycarbonate filter is then carbon-
coated and transferred to EM grids for analysis as described previously.
Low temperature ashers are available with one, two, or four chambers.
The following modifications minimize contamination in using these units: /-,
(1) A single chamber is dedicated for ashing samples for EM
analysis.
(2) An in-line filter is placed in the oxygen supply between
the regulator and entry into the asher. ,;
(3) For models with direct access to ambient laboratory air on
completion of ashing and return to ambient pressure, a
filter is placed in the inlet line to prevent Laboratory
air from .being sucked into the chamber. . . . •
In using the single-chamber method, a blank test tube and the sample tubes (up
to four, for a total of five in a IQr-cm-diaraeter. chamber.) are placed in-the •••' :.-• ..>
chamber "lengthwise, with the opening facing-the door. • .'.V ,'..".".
4 ^ •
The filtration step is also used in diluting the initial heavy
particulate loading. Filtration of aliquots is not recommended to obtain
different levels of loading on the new filters; a representative sample from
each aliquot in the filtration of suspensions is difficult to obtain.
Instead, for heavy loadings, different known areas of filter segments (one-
eighth, one-fourth, or one-half of the filter) should be ashed so that the
entire contents of the resuspension tube can be filtered onto either a 25-mm-,
37-mm-, or 47-mra-diameter polycarbonate filter for the desired dilution.
Distilled water is filtered through a 0.1-um (pore size) polycarbonate
filter prior to use. All glassware is washed with soap and water, rinsed with
acid, and then rinsed with particle-free distilled water. The dedicated asher
chamber is carefully wiped with damp lens paper.
62
-------
LTA Procedure—
The LTA manufacturer's instructions are followed since the power required
for one, two, or four chambers, the mass (glassware plus sample) placed in the
chamber, and the desired rate of ashing all vary. In general, the following
steps are performed:
(1) Each filter segment with a known deposit area is carefully
placed in a clean test tube (13 mm x 80 mm) using a clean
tweezer.
(2) With forceps, the tubes containing the sample, and one lab
blank (unused filter segment of the same size and type of
filter material as the sample) are placed lengthwise, side
by side in the chamber, with the mouths of the tubes facing
the open end (door) of the asher chamber. The tubes are
' laid in the center of the chamber within the region of the
coils surrounding the chamber. Up to four sample tubes and
one blank can be laid like logs inside the chamber.
(3) The power is slowly and carefully increased to prevent
"flashing" of the filter, which would result in loss of
sample.
(4) The filter membrane vanishes in about 30 rain; ashing is
continued for another 2 to 3 h to ensure complete ashing.
The chamber is slowly allowed to reach ambient pressure.
(5) The test tubes are carefully removed and placed in a
beaker, covered, and stored on a class-100 clean bench for
resuspension.
Resuspensioa (SonlficaClon) Procedure—
Ultrasonification is used in resuspending and redispersing the ash
resulting from "the LTA. The superiority of a probe-type ultrasonic device
over a bath-type device has not been demonstrated. However, the criteria of
low energy and minimum Bonification time appear valid. The probe-type
instrument is more readily calibrated (desired reproducibility), but requires
a larger volume of suspension to work with. The bath-type instrument is more
difficult to calibrate, and is usually of fixed wattage. A generalization is
that probes are used for dispersing, baths for cleaning. . Recently, a bath-
type ultrasonic unit (Ladd Research Industries, Inc., Burlington, Vermont)
with a variable power source and timer has become available that appears to
have the advantages of both types of ultrasonic devices.
The resuspension procedure is as follows:
(1) 10 mL of filtered distilled water is added to each test
tube.
(2) Each tube is placed in a 100-mL beaker containing 50 mL of
water.
(3) The beaker, with the tube, is placed in the low-energy
ultrasonic bath.
63
-------
(4) Ultrasonic energy of about 50-to 60 w'(60%) is applied for
3 rain.
Filtration Procedure—
Liquid filtration of suspensions for EM examination using polycarbonate
filters is one of the more difficult procedures to standardize. Variations in
the nature of the filter material, geometry and distribution of pores, and
method of manufacturing make it difficult to obtain a uniform deposit of
particulates on the filter. The following procedure is used for consistency
in the filtration procedure:
(1) A filtering apparatus having a filter size adequate for the
desired dilution—preferably 25-mm or 47-mm diameter—is
assembled. A polycarbonate filter (O.l-gm pore size) is
used shiny side up for the deposit, with a cellulose ester
filter (5~um pore size) as a backing filter on the glass
frit.
(2) While dry, the filters are centered and suction is applied.
The filter funnel is mounted on the centered, perfectly
flat filters with the vacuum on.
(3) The vacuum is then turned off. A 2 mL amount of particle-
free distilled water is added to the filter funnel,
followed by careful addition of all the water in the test
tube containing the dispersed ash. The test tube is rinsed
twice with particle-free distilled water, and the contents
are carefully added to the filter funnel.
(4) Suction is then applied; neither rinsing i:he filter funnel
nor adding extra liquid is permitted during the entire fil-
tration process.
(5) At the end of filtration, suction is stopped.
•(6) If possible, the filter is dried on a glass slide or holder
that can be placed directly in the vacuum evaporator for
carbon-coating.
1 (7) The dry filter is stored, in a .disposable Petri dish (taped': ' • '/
on a glass slide), or in the special holder, until ready
for carbon-coating, grid ^transfer, and EM analysis.
(8) The effective area of the redispersion filter and the area
of original filter deposit cut for ashing must be recorded
(ashing factor) for inclusion in analytical data reduction
and reporting.
64
-------
SECTION 8
BULK-SAMPLE ANALYSIS
DISCUSSION OF PROTOCOL
Bulk samples may be air samples collected in large volumes using electro-
static precipitators, bag collectors, or high-volume samplers, for example; or
chey may be original pieces of source material containing asbestos, such as
insulation, asbestos paper products, and asbestos cement products.
Efficient usage of the three levels of analysis requires effective com-
munication between those requesting an analysis and those responsible for con-
ducting the analysis. Personnel requesting an analysis must understand the
limitations of each level of analysis by EM. For example, requesting EM
analysis of a bulk-material (solid) sample, where there is marked disagreement
regarding the presence of asbestos (amphibole), and using Level I (screening)
analysis, are incompatible. Bulk-material samples require grinding for analy-
sis; grinding requires care to minimize such problems as contamination, change
in size of the asbestos fiber, increase in fragments that morphologically meet
the criteria of a fiber, possible change in the relationship of asbestos to
nonasbestos components, and possible destruction of asbestos fiber crystal-
Unity.
Following grinding, bulk-material samples should first be analyzed by
PLM, followed by XRD, if necessary. XRD provides information on samples
having asbestos concentration levels of at least' 2%. PLM provides information
on asbestos and nonasbestos components, as well as on the size of the asbestos
fibers in the solid-bulk phase. The additional information aids in EM . • ..•..;:.."£
analysis of these samples at the selected, level. of .analysis. . . ...','•':«
DESCRIPTION OF METHODOLOGY
1. Polarized Light Microscopy
Analysis of bulk samples, such as insulation material, for component
identification and for determination of the type and concentration of asbestos
present is best accomplished by PLM. With the polarized light microscope,
particle properties—such as color, morphology, refractive index, bire-
fringence (which indicates a crystalline substance rather than an amorphous
substance), surface texture, reflectivity, and magnetism—can be observed and
determined. Determination of such a large number of particle properties
allows identification of specific particle types, in most cases. For example,
amorphous slags and crystalline minerals are common nonfibrous filler compon-
ents of insulation materials that can be easily distinguished by PLM.
65
-------
2. X-Ray Diffraction Analysis
XRD has been successfully used to measure asbestos concent In both
aerosol samples and bulk samples. The reported values for sensitivity and
accuracy vary depending on the exact technique, but recent reports quote
sensitivity values of about 1% chrysotile, whereas 5% was more common when the
technique was first demonstrated. Much of the improved sensitivity derives
from sample preparation techniques, which are important in XRD, but digital
data collection and use of such accessories as x-ray monochromators to reduce
background are also important.
While 1? sensitivity has not been demonstrated for asbestos materials
other than chrysotile, the same sample preparation procedures are applicable
to other mineral forms, with comparable sensitivities expected in cases with
serious interfering lines. Interferences would hinder the analysis for
chrysotite as well.as for other asbestos minerals. Considering that the high-
sensitivity procedures have been only partially demonstrated at the 17. level,
a sensitivity of, say, 2% is probably a more realistic expectation for
asbestos minerals in general.
3. Electron Microscopy
For bulk-air samples, asbestos analysis by EM entails an addition to the
sample preparation procedure to attain a representative powder sample at a
suitable concentration level to be placed on the EM grid. This additional
step is similar to the method used in preparing standards of known asbestos.
The finely divided powder samples are split into representative fractions, and
a small, weighed portion is suspended in a known volume of filtered distilled
water containing.0.1% Aerosol OT.* A mild ultrasonic treatment is used to
disperse the particles. Different known volumes of suspension are filtered
through a 0.1-ura (pore size), 25-mm-diaraeter Nucleporet membrane filter. The
dried Nuclepore^filter is then carbon-coated and transferred to an EM grid
using the refined Jaffe wick technique described previously.
Bulk-solid samples are gently and slowly ground to a powder for EM
analysis to minimize localized heating; the powder is then prepared for the EM
grid by the method described for bulk-air samples.
*"* * _.- • AI.I- «•!. •• ' '
For EM analysis of bulk-air samples, a weighed portion is suspended in
filtered distilled water, deagglomerated in an ultrasonic bath, transferred to
a volumetric flask, and brought to 'volume with filtered distilled water. An
aliquot is then filtered onto a 0.1-um (pore size) polycarbonate filter using
a 5.0-ym (pore size) cellulose ester filter as a back-up filter on the
filtration apparatus. The dried polycarbonate filter is then carbon-coated.
A 3-tmn x 3-mm portion of the carbon-coated filter is then directly transferred
to a 200-mesh carbon-coated copper EM grid using the refined Jaffe wick washer
technique. EM analysis based on Level I, II, or III effort is then performed.
* Fisher Scientific Co. (Cat. no. 50-A-292), 711 Forbes Ave., Pittsburgh, Pa.
t Nuclepore Corporation, 7035 Commerce Circle, Pleasanton, Calif.
66
-------
SECTION 9
NUMERICAL RELATIONSHIPS AND ANALYTICAL AIDS
The fibrous structures (fibers, bundles, clusters, and matrices) in an
air sample are to be counted, sized, and identified as asbestos or non-
asbestos. An air sample ranging from 1 to 5 m3, depending on its total
suspended particulates (TSP) content (in'ambient air, the average TSP is
between 30 and 300 ug/m3) is drawn through a 37-mra filter (effective
filtration area of 8.6 cm2) or a 47-mm filter (effective filtration area of
9.6 cm2). The asbestos content, unlike a prepared laboratory standard or
sample, is a very small percentage (less than 1%) of the particulate loading
(TSP content) collected on the filter surface.
Two small circular sections of the filter of approximately 3-mm diameter
are transferred to EM grids for transmission electron microscopy. Either one
or both EM grids are examined for asbestos content; 10 random grid openings
are examined for each EM grid. Each grid opening measures approximately 85 urn
x 85 urn. At 20.000X magnification, a field of view of 4.5 gm x 5.0 ym is used
in the examination. This approximates to about 300 fields per grid opening or
3000 fields of view per grid examined (6000 fields for two EM grids) if less
than 100 asbestos structures had been found. Unlike a field blank or
laboratory blank, the statistical significance of obtaining a low asbestos
count in the midst of atmospheric clutter needs to be recognized.
LIMITS OF DETECTION ' •••"
"
The minimum detection limit of the EM method for counting airborne
asbestos fibers varies depending on the amount of total extraneous, particulate,.,!
.matter.'.In • the sample, and on the contamination level in the laboratory - ' 'v-'
environment. This limit also depends on the air sampling parameters, loading
level, and EM parameters used. For^example, assuming that a fiber count has
an accuracy of ±1 fiber, when 10 full-grid openings are scanned, each grid
opening having an average area of 0.72 x 10~u cm2, the detection limit is
determined from the equation
n , . . 1 Area of filter (cm2) 1
Detection limit = — x - x
10 0.72 x 10'* (cm2) Volume of air (m3)
The minimum detection limit, then, is lower for very dilute samples.
Examining full-grid openings leads to a lower value of the minimum detection
limit because of the large area scanned, as compared with the field of view
method. With a given sample, the detection limit can be lowered considerably,
but the required experimental effort increases. The guideline of using 10
. 67
-------
"full-grid openings represents a judicious compromise between a reasonable
experimental effort and a fairly low value of the detection limit. However,
using two or more TEM grids reduces the detection limit further and improves
che precision of the estimates.
STATISTICAL METHODOLOGY
Several statistical strategies have been used to characterize airborne
asbestos distributions and estimate the abundance of asbestos fibers in a
given sample. These methods range from simple tabulation of observed frequen-
cies of fibers across grid opening samples to the fitting of statistical prob-
ability distribution such as the Poisson. This statistical section outlines a
general methodology for fitting observed data to a statistical probability
distribution (either Poisson or normal depending on the fit of the former).
The mean and 95X confidence interval are- then estimated and used for the
purpose of sample description and drawing inference regarding the abundance of
airborne asbestos fibers in the environment in which the samples were
obtained.
As an illustration, consider the hypothetical data in Table 5.
The expected number of grid openings with no fibers is:
Ne~" = (98) (e-2""^*) = 4.7806
The expected member of grid openings with 1, 2, 3, ... fibers are found by
multiplying Neu by u/r , i.e., 4.7806 x 3.0204, 4.7806 x 3.0204/2, 4.7806 x
3.0204/3, successively.
«*•
To test the fit of the Poisson distribution to the observed data we
compute a chi-square statistic, x2 = £ (observed - expected)2/expected = 8.26.
Since there are nine different observed frequencies (i.e., numbers of
fibers) there are 9-2=7 degrees of freedom (since we estimate one
parameter). The probability of xa2 = 8.26 is p = 0.4; therefore, we conclude - j
that the Poisson distribution fits the observed data.
In certain cases, the observed1frequencies will not have a Poisson
distribution (as determined by the 2rev*-°usly described chi-square statistic).
In this case we estimate the mean (X), variance S2 , and 95% confidence limits
68
-------
.TABLE 5. HYPOTHETICAL DATA
No. of fibers
on grid 1
opening Observed* Expected Observed
(r) frequency (f) frequency - expected
0
1
2
3
4
5
6
7
8
9
10
1
1 or more
Total
3 4.78
17 14.44
26 21.81
16 21.96
18 16.58
9 10.03
3 5.04
5 2.18
0 0.82
It 0.27
0 0.08
_0 0.03,
98 98.0
-1.78
+2.56
+4.19
-5.96
+ 1.42
-1.02
-2.04
+2.84
1.20 -0.20
Probability
(r)
.049
.149
.224
.224
.168
.101
.050
.022
.003
* The number of grid openings showing that number of fibers.
T We combine adjacent frequencies to get a minimum of 1 fiber per group.
Assuming a Poisson distribution, the mean is u = E fr/Ef = 296/98 = 3.0204.
That is, the sum of the product of the observed frequencies and number of
fibers divided by the sum of the frequencies.
69
-------
assuming normality. The 95% confidence limits are computed as a function of
the sample variance and t distribution.
n I X2 - [E X ]2
n(n - 1)
where k is the number of grid openings, n is the total number of fibers found,
and X^ is the number of fibers found in grid opening i.
The 952 confidence limits are given by
where t is the value of the two-tailed t distribution for probability p <.025
and n - 1 degrees of freedom.
1. 95Z Confidence Limits for a Poisson Varlate
To generate a level of confidence regarding our estimate of the number of
asbestos fibers per grid opening, a 95% confidence limit can be derived. For
counts of 0 through 20, Table 6 may be used.
For example, if out of 20 grid openings 15 fibers are found, the 95%
confidence limits are obtained by taking the lower and upper bounds from Table
6 as 8.40 and 24.74. Then per grid opening, the 95% confidence limit is
8.40/20 to 24.74/20, or 0.42 tn 1.237 fibers per grid opening.
For counts, greater than 20, a simple normal approximation is computa-
tionally convenient. The normal approximation is X - L ± x * (S^) where L is
the observed count, SL is /L , and x* *s 1.96 or 2.58 for the 95% or 99%
confidence limits, respectively.
For example, if 35 fibers were observed from inspection of 20 grid
openings, L = 35,-Sy = /35 = 5.91, A = 35 ±1.96 (5.91) = 23.4 to 46.6 or .
23.4/20 = 1.38 to 46.6/20 = 2.74 fibers per grid opening.
2. Comparison of Two Poisson Variates .
In certain cases, a new test sample is compared to a "blank" or "control"
sample. Table 7 gives those differences between control and test samples that
are significant at the 5% level.
Inspection of Table 7 reveals that the minimal detectable difference
between test and control samples is 5 fibers in the test sample and 0 fibers
in the blanks. Typically, inspection of 20 grid openings for a blank control
reveals between 0 and 5 fibers. At the upper bound (i.e., 5 fibers in a
70
-------
TABLE 6. 95 PERCENT CONFIDENCE LIMITS
No. of Fibers
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
95%
Lower
0.000
0.0253
0.242
0.619
1.09
1.62
2. -20
2.81
3.45
4.12
4.80
5.49
6.20
6.92
7.65
8.40
9.15
9.90
10.67
11.44
12.22
Limits
Upper
3.69
5.57
7.22
8.77
10.24
11.67
13.06
14.42
15.76
17.08
18.39
19.68
20.96
22.23
23.49
24.74
25.98
27.22
28.45
29.67
30.89
71
-------
TABLE 7. CONTROL AND TEST SAMPLE DIFFERENCES
Fiber
Control
0
0
0 -
I
I
1
2
2
3
3
3
4
4
5
5
5
Count
Test sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Fiber
Count
Control Test sample
6
6
7
7
7
8
8
9
9
10
10
10
11
11
12
12
13
13
13 •
14
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
72
-------
control sample), 18 fibers in the test sample are required for statistical
significance. This estimate of- 18 fibers may prove useful for establishing an
asbestos detection limit criterion.
MAGNIFICATION CALIBRATION
The following steps should be performed to calibrate the magnification of
the EM:
(1) Align the EM using the manufacturer's instructions.
(2) Insert mag-calibration grating replica* (with 54,864 lines per
inch, or 2160 lines per mm) in the specimen holder.
(3) Switch on the beam, obtain the image of the replica grating at
20,OOOX magnification (or at .the magnification at which the
asbestos samples will be analyzed), and focus.
(4) If the fluorescent screen has scribed circles of known
diameters, align one line tangentially to the circumference of
one circle using stage control. Count the number of lines in
a diameter perpendicular to the lines. In most cases, the
other end of the diameter will be between the n" and nc^ + 1
line. The fractional spacing can be estimated by eye.
Alternatively, the separation between lines can be estimated
using the scribed circles.
(5) If X line spacings span Y mm on the fluorescent screen using
this grating replica, the true magnification, M, is given by
M Y x 2160
M =
The readings should be repeated at different locations on the
replica, and the average of about six readings should be taken
as the representative or true magnification for that setting
of the EM, as in the following example:
* For example, Cat. no. 1002, E. F. Fullam Co., Schenectady, N.Y.
73
-------
Line Spacings, mm on Screen, Magnification,
X Y M
9.5 83
9.3 80
7.0 . 60
8.8 80
9.0 80
9.0 80
Average: 19000
On most EM's with large (18-cm diameter) fluorescent screens, the
magnification is substantially constant only with the central 8- to 10-cm-
diameter region. Therefore, calibration measurements should be made within
this small region and not over the entire screen.
PREPARATION OF BLANKS
Even after taking the utmost precautions to avoid asbestos contamination,
the possibility of some contamination cannot be ruled out. Contamination
should be checked periodically by running field blank samples in addition to
laboratory blanks. Field blanks should be analyzed prior to laboratory
blanks. A blank sample may consist of a clean filter subjected to all the
processing conducted for an actual air sample. This processing may include
ashing, resuspension, redeposition, carbon-coating, transfer to a TEM grid,
and TEM examination.,
When analyses of blank samples show significant background levels of
asbestos, these should be subtracted from the values obtained for field
samples. Also, the minimum detection limit may be calculated as twice or "';.
three times the standard deviation of the blank or background value. '
USE OP COMPUTERS - ' j'
i
Data reduction is facilitated by computers. Computer printouts can be
used in reports. Each laboratory. s,hould develop software suitable for its
needs as well as to maintain basic information, such as fiber, areas examined,
volume/mass of sample, and size distribution, for possible interlaboratory
comparison.
Appendixes B and C present sample printouts from Level I and Level II
analyses, respectively.
74
-------
REFERENCES
Anderson, C. H., and J. M. Long. 1980. Interim Method for Determining
Asbestos in Water. EPA-600/4-80-005, U.S. Environmental Protection
Agency, Athens, Georgia. 44 pp.
Hileman, B. 1981. Participate Matter: The Inhalable Variety. Environ. Sci.
Technol., 15(9):983-986.
John, Wi, and G. Reischl. 1980. A Cyclone for Size-Selective Sampling of
Ambient Air. APCA Journal, 30(8):872-876.
Jones, D. R., S. C. Agarwal, and J. D. Stockham. 1981. Asbestos Analysis of
Iron Ore Beneficiation Plant Samples. Final Report, Work Assignments No.
3 and No. 9, Contract No. 68-02-2617. U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina. 228 pp.
-,
Leidel, N. A., S. G. Bayer, R. D. Zurawalde, and K. A. Bursch. 1979.
USPHS/NIOSH Membrane Filter Method for Evaluating Airborne Asbestos
Fibers. U.S. Department of Health, Education, and Welfare. National
Institute for Occupational Safety and Health, Cincinnati, Ohio. 97 pp.
Mueller, P. K., A. E. Alcover, R. L. Stanley, and G. R. Smith. 1975.
Asbestos Fiber Atlas. EPA-650/2-75-036, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina. 58 pp.
Ortiz, L. W., and B. L. Isom. 1974. Transfer Technique for Electron
!•'• Microscopy of Membrane Filter Samples. Amer. Ind. Hyg. Assoc. J.
35(7):423-425.
••'. •, ••<.•• • • ' -
Samudra, A. V., F. C. Bock, C. F. Harwood, and J. D. Stockham. 1977.
- • Evaluating and Optimizing Electron Microscope Methods for Characterizing
Airborne Asbestos. EPA-600/2-.78-038, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina. 194 pp.
Samudra, A. V., C. F. Harwood, and J. D. Stockham. 1978. Electron Microscope
Measurement of Airborne Asbestos Concentrations: A Provisional
Methodology Manual. EPA-600/2-77-178, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina. 57 pp.
Zurawalde, R. D., and J. M. Dement. 1977. Review and Evaluation of Analytical
Methods for Environmental Studies of Fibrous Particulate Exposures.
DHEW(NIOSH) Publication No. 77-204. National Institute for Occupational
Safety and Health, Cincinnati, Ohio. 71 pp.
75
-------
APPENDIX A
FIGURES
Figure Al. Vacuum evaporator,
76
-------
'
(a) 37~mm diameter.
(b) kj-ntm diameter,
Figure A2. Multiple coating arrangement in evaporator.
-------
I
1
3B$Bj^^
(a) Modified ^7~mm-diameter Petri-slides
(b) 37~mro~diameter cassette.
Figure A3, Close-up of multiple coating arrangement.
78
-------
(a) Plan view.
Carbon-Coated
EM Grid
Wedge
Ground
Glass Dish
Edge Cover
Nuclepore
Grid
Stainless mesh
— T*. \ Stainless Mesh ^V * ' *' — —
\
o i
— i /
Filter
^ Petrl »•
_ JLft
L
Dish 1
— ' i / Foam /, \ ^,
t
\
i
1
l
" i 1 ter Paper
(b) Elevation view.
Nuclepore.
Carbon Coat
Carbon Substrate
Grid
Asbestos
Fibers
Ch loroform
Level
Asbestos
F i bers
~Gr id
Chloroform
Wash
(d) Principle of the Jaffe method.
Nuclepore C-coated
Particle Side Down
Stainless Mesh
Polyurethane Foam
(cj Details of placing the specimen for washing.
Figure A4. Modified Jaffe wick washer method (sketch)
-------
-J
00
o
Figure A5. Modified Jaffe wick washer,
Figure A6. Transmission electron microscope.
-------
Count as one fiber:
Count as two_fibers (space between fibers greater than the width of one fiber)
Count as three fibers:
Count as bundles:
Count as cluster/clump:
Count as matrix/debris:
Figure A7. Morphology and counting guidelines used
in determining asbestos structures.
81
-------
No.
- 3"
Filter Type.
Filter Area_
Grid Openi/iq Area
Nueto
Acc. Volta
11
Ift
19
2Q
£1
2Z
23
24
25
Struct.
P
P
F
P
M
P>
F
F
F
U
IU
M
kA
M
IU
P
IU
M
kJ
p
M
U
M
F
M
Dimension
Width
1
1
1
1
1
2
4
3
Icnqin
33
IQ
20
^
Q
il
?r\
»o
M
«.
%
2*?
23
7
12
H
U
15
19
17
3A
15
|5
7
13
SAED Observation
Cdyrs
v/
tX
•
*s
I/
x/
tX
•
I/
s
lX
v'
i/
S
St
•'•
:•
i
iS
Ai-gli
Ambiq
*>
»/
^
W>
•
»x
•
•10-ID
.
v^
,/
GO
4
5
to
7
Struct.
¥
2k
77
ZB
20
^
31
.^2^
33
54
JK
3fa
^7
.•^ft
.^0
40
41
42
4*
44
4.^
4h
47
48
49
50
Struct.
U
F
F
M
u
M
R
U
^A
U
U
F
M
P
R
F
P
P
LI
F
F
F
F
P
Dimension
Uirlth
-
1
4
5
i
Length
•7
»L
7
Zt
10
ii
4|
^Z
(,
r,,
1.^
17
2Q
in
24
2ft«7
IfT
U
17
0
?rt
Iff
13
9
4|
Cnrys
I/
*/
t/
iX
*/
t/
^
w/
K/
lX
v/
*/
S
S
SAE3 Obs-'-.-jv :-
Anph
iS
A»b:; -.T- :
,/
S
t/
S
'
S
v/
00
. .
Figure'-A8. Level I data sheet (example)
-------
CO
Tilt Area of Fluorescent Screen
.Figure A9. Scanning of full-grid opening.
-------
1
Figure ALO... Jransmission, electron microscope
with energy dispersive spectrometer.
84
-------
1
.
•
~ IF 48 LD2
—r.ii RD&veieKEv/CH 46. SE;;"
EuG= 2.eeeKEv, IB.
; :2811688 CHRVSO fl4
?Hflii flDD .016KEV/CH 40. SEC;
BUG= 2. fllKEYJ 27. •'
95 6. 64 8. 51
0-4-10-3-<1
95 8. 94 8. 91
0-7-10-0-0
TREHOLITE T 79 988
PHftli HDD .818KEV/CH
BUG= 1. 758KEV, 334.
48. SECS
. ::511680 RNTHOP C2
"":'Sll fiDD '. 816KEV/CH
5JG» 2. 90BKEV, 172.
18.238
0-4-10-3-<1
CROCIDOLITECR 37 11
PHflll flDD .818KEV/CH 48. SECS
0-3-10-0-1
?1?076 fiflOSITE 1
: flDD . 810KEV/CH
1 946<'E-',
95 8.04 8.91
1-1-10-0-6
95 8.84 8.51
0-2-10-0-7
Figure All. Spectra profiles of asbestos standards,
85
-------
Sample No.Pto4-7n-llQ-
/.>••?."/?'- *".;«.;.
.••'•''•'i ''::. •>.->">
ExlO •"""
Voltage_
Oate
Filter Type Mur Igno
re
Filter Area Q.fo Cmz
i-'''':!;-:'.yi' -'Beam Current inn I1.A
' ;':/r':' |;''--Mag"'« eat i on_2DJQOOx - Fi'lm
Grid Box
Grid Opening Area 72.25 X
Grid Location D-7
.• CommentsQond f)j.
7
e
9
10
Struct.
IT
1
2.
0
3
4
S
ft>
7
8
9
10
II
IZ
13
14
15
Ik.
n
Struct.
F
n
n
n
n
it
u
II
II
II
II
II
II
II
II
II
ll
Dimension -r
Width
2.
9
10
b
4
I
3
2 '
Z
IS'
10
5
b
3
7
4
4
•
•
.
.
Length..,
2(6 -I.
30 -:
• '
85 :
.fo5 ;
23
' 8
40
30, '
!l .-:••
80 :,
80J
.45 --.••'
43. SJ
25, -X;
zi*-:
z5- •-';;
-•*3fl':?
- ^.;!*-
• V." " "> "
. . •-' ''•
•^r- .
• ••• •
SAE Obserwatinn
Chyrs
r
-
."
' '.
*-' ' '."'' '
~
•
>i
( '• :-'•-.•
' :'>r .
: !" '
» ' '»
• - •'
• :
1
•• '
•<#.-.',
.
.•• •
- '
'
Amph
S
^
^
S
•
S
"
S
>.'•
•
' .
Ambig
•
',
<&>
.
<^>
(f^>
•
Non-A
/
/
^
t/
l/
y
,/
v/
l/
v^
No-P
•
FD/Imaqe
0°
. •
•
•
EOS
Ha
—
•
___
~
^~
i
— :
.•_
•
—
•
Hg
9
35
>9
fob
37
1*7
Si
19
HO
59
15?
I7O
5 1
Ca
T
3^5
2.2
4J?
'8
,
Fe
•7
IV
Z)
23
»fo
U>
•
10
r^>
TV^Ii 1
(^
^j
.
oo
:<•'•!.
Figure A12!. Level I.I data sheet (example)
: • f -iJ>,4 J; ^ ,:
i i J-! ^ i
-------
EM DATA REPORT
Sample Number: R09-2865 _
IITRI Sample No.: C010-1B59
Sample Type: Bulk, Air, Water. Misc.
Filter Type: Nuclepnre
Volume of Fluid Sampled: NA
3/26/81
1/30/81
t.o Anal yzoH:
Date Sample Received:_
(circle one)
Area of Filter Deposit (cm2):_
Mass Deposited: NA
8.6
r
Total Number of Structures:
Total Number of Asbestos Structures
2.1 Chrysotile 88
2.« Amphibole 5
Crocidolite 5
Tremolite
Amosite
101 |
: 93
Anthophyllite
Actinolite
Non- Identity
6.
2.3 Non-Identity 8_
3. Asbestos Structure Description
3.1
3.1
3.2
3.2
3.3
3.4
Area
Total Number of Fibers:
.1 Chrysotile
Fiber Length; Range (urn)
Fiber Diameter; Range (um)
Aspect Ratio; Range (um)
.2 Amphibole
•
Fiber Length; Range (um)
Fiber Diameter; Range (um)
Aspect Ratio; Range (um)
Total Number of Bundles:
85
80
.31 -
.06 -
3.5 -
5
.31 -
.06 -
5.0 -
Total Number of Clusters/Clumps:
Total Number of Matrix/Debris:
of Filter Sample Analyzed,
(cm2):
2.25
.25
31.0
2.19
.31
10.0
5
2
1
Mass
Mass
Mean
Mean
Mean
Mass
Mean
Mean
Mean
V
»
•
(ng)
(ng)
(um)
(um)
(um)
(ng)
(um)
(um)
(um)
Mass (ng)
Mass (ng)
Mass (ng)
0
0
11
1
0
6
.0016
.0010
.78
.07
.00
.0006
.00
.15
.40
.0007
.0002
negligible
.0007225
A.
5. Total Mass of Asbestos Analyzed (ng):
Number of Pictures Attached: 2
.0025
7. Qualitative Description of Non-Asbestos Particles Few, small particles--
non-descriptive ..-
8. Comments: Particulate loading OK. _________
Figure A13. EM data report (example)
87
-------
SAMPLE SUMMAKY KLI'OIM
Sample Number: R09-2865 Dale of Ruport:
IITRI Sample No.: C010-1859 Date Sample Received:
Sample Type: Bulk. ffiTj) Water, Misc. (circle one)
Filter Type: Nuclepore Area of Filter Deposit
4/1/81
1/30/81 •
(cmz): 8.6
Volume of Fluid Sampled: NA Mass Deposited: NA
1. Total Number of Structures: 1,202,215
2. Total Number of Asbestos Structures: 1,106,990
2.1 Chrysotile 1,047,474
2.2 Amphibole 59?516
Crocidolite 59,516 Ant hophyl lite
Tremolite Actinolite
Amosite Non-Identity
2.3 Non-Identity 95,225
3. Asbestos Structure Description
3.1 Total Number of Fibers: 1,011,765 Mass (ng)
3.1.1 Chrysotile 952.249 Mass {ng)
Fiber Length; Range (urn) .31 . 2.25 Mean f^1")
Fiber Diameter; Range (um).06. - ,25 Mean (urn)
Aspect Ratio; Range (um) 3.5 - 31, Q Mean (um)
3.2.2 Amphibole 59.516 Mass (ng)
Fiber Length; Range (um) ..31 - 2.19 Mean (um)
Fiber Diameter; Range (urn). 06 - .31 Mean (ym)
Aspect Ratio; Range (um) 5.0 - 10.0 Mean (um)
3.2.' Total Number of Bundles: 59,516 .Mass (ng)*
3.3 Total Number of Clusters/Clumps: 23,806 .Mass (ng)
3.4 Total Number of Matrix/Debris: 11,903 ..Mass (ng) ,
4. Area of Filter Sample Analyzed, (cm2): .0007225'"-'
5. Total Mass of Asbestos Analyzedt(ng): 29.5
6. Number of Pictures Attached: 2
19.00
11.90
0.78
0.07
11.00
7.14
1.00
0.15 -
6.40
8. 30
2.10
. , '0.10 .-
•
7. Qualitative Description^ Non-Asbestos Particles Few, small particles--
Non-descn'ptive
6. Comments: Particulate loading OK.
Figure A14. Sample summary report (example).
88
-------
xs,
(a) Effects of tiIting.
(b) Fiber alignment.
Figure A15. Effects of tilting and alignment of fiber.
Figure A16. Method of measuring two perpendicular diameters for each ring,
89
-------
Draw 0, 1st, 2nd, ... order of horizontal rows--
perpendicular separation between horizontal rows.
Figure A17. Method of recognizing a horizontal row of spots.
-------
Horizontal Row
Draw the Oth horizontal row.
Draw a perpendicular through the origin.
Join the origin to the first spot to the right
of the perpendicular in the 1st row and extend
the line to measure the acute angle 6 of this
line from the Oth row.
A18. Relationship of di, da, 61^2, and R.
-------
'.
• p
ti*
(a) Zone axis [lOO].
(b) Zone axis [301].
(c) Zone axis [101],
(d) Zone axis [101].
Figure A19. Typical Zone-axis SAED patterns from amosite standard specimen.
(Jones et al., 1981)
92
-------
1
(a) Zone axis [100]
(b) Zone axis [101].
(c) Zone axis [lio].
(d) Zone axis [301].
Figure A20. Typical zone-axis patterns from crocidolite standard specimen.
(Jones et al., 1981)
93
-------
(a) Zone axis [100].
(b) Zone axis [101J.
(c) Zone axis [201].
(d) Zone axis [301].
\
Figure A21. Typical zone-axis patterns from tremolite standard specimen.
(Jones et al., 1981)
94
-------
(a) Zone axis [100],
(b) Zone axis [H»2].
:£511680 ftNTHOP C2
•nil fiDD .618KEV/CH
-------
APPENDIX B
COMPUTER'PRiNTODT; OP LEVEL I ANALYSIS (EXAMPLE) '••
•>.'.•=?: --£ Y::v:V~" - ' •'? '.•
_ —•»* i ^ - ^ *
vO
a*
' '
I IT RESEARCH INSTITUTE. STRUCT.URfcfANALYSIS DATA
INDIVIDUAL OBJECT}. DATA :TA'DLE:>.".(F=F:IBER j . B=BlJNDL.E.f C
TABLE PREPARATION" DATE:^21^fifeR-i|i;;;>- -,1. j«5;;;^- /•.;'. -.
=CLUSTERf M=MATRIX) -->.
::= ss ^
=:a = 3S
SAMPLE
(3rd
Opn
.1
1
2
•p
2
3
3
3
4
4
7
7
7
8
8
9
12
12
12
13
13
13
14
14
14
15
15
1A
JA
ObJ
1
2
3
4
S
A
7
0
V
10
11
12
13
14
15
1A
17
18
19
20
21
22
23
24
= = = =:
CODE
Str
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F-'
F
F
25' F
2A
2?
28
29
F
F
F
)•"
0.000 0. 3 75 V-;.-j;2V.l-9 ."
0.000" 0.7SO•-£*$'••
0.000
0.000 0.. 937 /-3-..00 \
0.000 0.125.r-:4l'V2E
0.000 0.250 . "V-'.
0.000 0.812 •-:j,'CiS;'COO
.307
0.04A
1 . 300
5.464
0.000 0
o.ooo
0. OQO-/0 . 3.12 >:t;'l.i 07. :•
Not
As be
X
X
X
.X
•
X
X
X
x
'".X
X
f
X
«.
X
X
•
x
X
x
• x
: NO.
.;Patt -.-X-Ray
"_. *>" ^• * t . ~ " •• •
-------
APPENDIX B (Continued)
VO
TIT RESEARCH INSTITUTE STRUCTURE ANALYSIS DATA
SINGLE SAMPLE SUMMARY TABLETS
SAMPLE CODE: TABLE PREPARATION BATE: SI-
= = = = = = = = = =r= = as±:== = = = s
Aerosol Object Count
Object
Structure?
Tiber
rifHTlF'lC* Co]
Tape
Chrwsotile
Other
All Fiber
1 net ion find
fli r Vc>] ume -• 1 .
J.ieF-osit Ares -: 1.
A?ih*?d Are ft = I .
Kedoposit Aresi =: 1.
= K = = sssrsss = =: = ==: = =:=:=:=:si=i=: = =ri3:j5-:
And Calculated Object Mass C
Actual
Object
Count
0.
1 2 .
22.
34.
Prepnrat
00 Cu M
00 So Cm
('0 So Citi
00 So C,m
Number
Conc:r?n.
( Number
Per Cu M)
0.
SJ304.
15223.
23529.
inn Iiatu
haracterj sties
Mass
Concen. Avprjisle
(Pi cos! ram Width
Per i;u M) (Micron)
0
19155
Grjd H
1 nd j vi
U'jmber
I'llm ;-i
.0 0.00 :k 0.00
.4 0.38 .1. 0.21
0.46 J: 0.37
0.43 .1: 0,32
riual Hrid iipenin.«l
of Grid Or- en in. -.IG
sun i rirst i on
Average
Length
(Micron)
0.00 .1: 0.00
S.20 .1 4,50
3.V.9 1 3.24
= 0.000072
- 20
= '..;0000
Aver-aae
Length
To Width
Ft a t.i.o
0.00 :t
15.^5 1
7.81 .i
10.47 i
Go Cm
0.00
1.3.02
5.03
9.27
-------
. " --- , •'•--• • •r. .
;^,:;^;;: f- ;£ APPENDIX B (Continued)
1 " ' !•*' r-~ * * •* '' * '< T-
.TIT RESEARCH INSTITUTE STRUCTURE ANALYSIS DATA
INDIVIDUAL OBJECT DATA TABLED! (F=F.I:BERr B=BUNDLE, C=CLUSTER» M=MATRIX)
TABLE PREPARATION DATE: 21-A:PR-B1 .j?-'
SAMPLE CODE:
vo
00
Size v•-•- -. ^""-Tj71 :'• •.'• •' ~ri r
OFT, ObJ Str Depth Width Lerjath .."Rat'io Chrasptile Amphibole Ambid Asbe Patt X-Rs«.
1A 30 F 0.000 0.
17 31 F 0.000 0
19 32 F 0.000 6.
5.591
,!• f «I^IM I ^v v ^f ^f v v, v A v* r .-t^'viv*-- f f-v t m -- P«B v- -> w
19 33 F 0.000 .O.OA2- .i:.'&'? -: 27V6 •.' j:''".':.-'.« . .
19 34 F 0.000 0.187-.;--.'^5|^. ,:; 10^V3 .'' ;-••.'•!:• '-' :i 0.1/0.
Total Ma'sis .'/•jP.fcc6dpam!i)= ~-':; O.OOO'.^ 27.680
Total Coun-t.-i'.^^.; .;:• .w,|:.= f-.:.:-9!\ :'" 12f
..
X
. X
X
22.
0.
/;Vt,- f:'"1 ; :
'--•: '.'« : f .
-
•*?^-.??'---jr'. .-••-I-
"/v^---^.-'-; #• ' .'•
-F=:rt;.~ar = =.!rf!==.:==!-'r!=
. -. •;-;;.<>.;? :
-------
APPENDIX C
COMPUTER PRINTOUT OF LEVEL II ANALYSIS (EXAMPLE)
TIT RESEARCH INSTITUTE STRUCTURE ANALYSIS DATA
INDIVIDUAL OBJECT DATA TABLE (F=FIRER, B=E«UNDLF_. r:=t;Ll)STfTR, M = MATRIX)
TABLE PREPARATION PATE! 21-APR-81
SAMPLE CODE!
Size (Micron)
Mass < Pi cosrsni)
lira
(Jpn
1
1
3
"S 3
\O
4
5
5
6
/
7
7
7
8
8
9
10
10
ObJ
1
2
3
4
5
A
7
tt
V
10
1 1
12
1 3
14
15
16
17
Str
F
F
f"
F
F
F
F
F
F
F"
F'
F
F
F
F
F
K
Depth
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
,000
.000
,000
.000
.000
.000
.000
.000
. 000
. 000
.000
.000
.000
.000
.000
.000
.000
Uicith t.ensth
0,
0,
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
Total
125
562
625
375
250
062
187
125
125
937
625
312
375
187
437
250
250
Mass
1 . 63
1.87
5.31
4.06
1 .44
0.50
3.00
1.87
0.69
5.00
5 . 00
2.01
2 . 6 9
1 . 56
1 . 63
1,56
2.JH
— - - not
Ratio Chrysotile Amphibole Anibia Asbe
13.0
3.3
0.5
10.8
5.8
8.0
16.0
1 5 . 0
S . 5
5. IS.
8.0
9.0
7.2
B..?
3.7
A.:.1
9 . t:;
(f-'icosl rstni)r-
lolaJ L'o'jnt •=
0.0598
1 . 3977
4.8892
1.3460
.
0.2117
• • t X
0.2485
• * • X
• t . • . .
X
• # « A
• • • '
• • • X
0.1294
• • « X
X
• • * *
0.0000 0.:.'82.?
0. 7, 0. .10.
no
rat*. X-Ray
MG(9)
r;A(7>
'I CGUI
MG(35)
CA( 35)
MG(19)
CA(22)
. MG(66)
FE(2B)
•
. MG(57)
CA<43>
•
•
•
t
•
•
MG
CA<18)
•
•
»
0.
s;i ( 19)
FE ( '/ )
Sl(llO)
F-T< IV)
r>I(59)
Ft: (21 >
!JI(152)
SI (170)
rr.( LA)
KK5U
r i: ( 6 >
-------
A^APPENDIX c (continued)
.TIT RESEARCH INSTITUTE STRUCTURE ANALYSIS DATA
SINGLE SAMPLE SUMMARY TABLES'^Li ' ;.:.". f
SAMPLE CODE: '?^i •'-• i ™BLE PREPARATION DATE: 2i-APR-ei
SSSSSESSSSSSSSISSISSSSSSSSrSSSSSSSSS SSSZSSSS=S=SS=X = S=SSS = SSSSSSSS SSSSSSSZSS^ 33 SSSiSS = = == = = SSSSIS! a =;=:=: = = =::
Aerosol Object Count And Calculated Object Mass Characteristics
Object
Structure Tape
Mass . Average
Actual'.'?: Concert. Concern. Average Average Length
Object.--!^(Number (Pico^ram Width Length To Width
County •('
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing/
REPORT NO.
3.
3. RECIPIENT'S ACCESSION NO.
TITLE AND SUBTITLE
Methodology for the Measurement of Airborne Asbestos
by Electron Microscopy
5. REPORT DATE
July 1984
i. PERFORMING ORGANIZATION CODE
AUTHOFHS)
George Yamate, Satish C. Agarwal, Robert D. Gibbons
B. PERFORMING ORGANIZATION REPORT No
C06470
PERFORMING ORGANIZATION NAME AND ADDRESS
IIT Research Institute
10 West 35th Streeet
Chicago, Illinois 60616
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
2. SPONSORING AGENCY NAME AND ADDRESS
Environmental Monitoring Systems Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
5. SUPPLEMENTARY NOTES
6. ABSTRACT
The provisional electron microscope methodology for measuring the concentration
of airborne asbestos fibers was refined. The methodology is divided into separate
protocols. The step-by-step procuedures for each protocol are nearly identical, so
that cumulative data can be obtained and uncertainties, especially in asbestos iden-
tification, can be clarified. The operational steps encompass (1) type of sample,
(2) collection and transport, (3) sample preparation, (4) examination under the
transmission electron microscope (TEM) and data collection, (5) data reduction and
reporting of results, and (6) quality control-quality assurance.
The TEM analytical protocol is subdidvided into three levels of analysis:
Level I, for screening many samples; Level II, for regulatory action; and Level III,
for confirmatory analysis of controversial samples. Because identification of
asbestos structures is critical, the level of analysis is directly related to the
information sought:
Level I—morphology and visual selected area electron diffraction (SAED)
pattern recognition.
Level II—morphology; visual SAED; and elemental analysis.
Level III—morpholgy; visual SAED, a selected number of SAED micrographs
of zone-axis patterns; and elemental analysis.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
>. IDENTIFIERS 'OPEN ENDED TERMS
COSATI held Croup
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19 SECURITY CLASS (Tins Report;
UNCLASSIFIED
21 NO OF FACES
112
20 SECURITY CLASS iTInspagci
UNCLASSIFIED
22. PRICE
Er-A Form 2220-1
------- |