EPA-650/2-75-029
January 1975
Environmental Protection Technology Seri
es
••••••iWliiOI'T FOR MtASURIIIIIiilil
x%svX'XvX^v>x-Xv>x%v*v.v/.\vA\viw*wX*~w^*.
-------
-------
EPA-650/2-75-029
DEVELOPMENT OF SCANNING ELECTRON
MICROSCOPY FOR MEASUREMENT
OF AIRBORNE ASBESTOS
CONCENTRATIONS
by
A. Pattnaik and J . D. Mcakin
The Franklin Institute Research Laboratories
Benjamin Franklin Parkway
Philadelphia, Pennsylvania 19103
Contract No. 68-02-1268
ROAP No. 26AAM
Program Element No. 1AA010
EPA Project Officer: Jack Wagman
Chemistry and Physics Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, D.C. 20460
January 1975
-------
EPA REVIEW NOTICE
This report has been reviewed by the National Environmental Research
Center - Research Triangle Park, Office of Research and Development,
EPA, and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development. U.S. Environ-
mental Protection Agency, have'been grouped into scries. These broad
categories were established to facilitate further development and applica-
tion of environmental technology. Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields. These series are:
1. ENVIRONMENTAL HEALTH EFFECTS RESEARCH
2. ENVIRONMENTAL PROTECTION TECHNOLOGY
3. ECOLOGICAL RESEARCH
4. ENVIRONMENTAL MONITORING
5. SOC1OECONOMIC ENVIRONMENTAL STUDIES
6. SCIENTIFIC AND TECHNICAL ASSESSMENT REPORTS
9. MISCELLANEOUS
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to
develop and demonstrate instrumentation, equipment and methodology
to repair or prevent environmental degradation from point and non-
point sources of pollution. This work provides the new or improved
technology required for the control and treatment of pollution sources
to meet environmental quality standards.
This document is available to the public for sale through the National
Technical Information Service, Springfield, Virginia 22161 .
Publication No. EPA-650/2-75-029
11
-------
F-C3779
CONTENTS
Title Page
ABSTRACT vi
1. INTRODUCTION 1-1
2. EXPERIMENTAL PROCEDURE 2-1
2.1 Statement of the Problem 2-1
2.2 Description of Methodology for the Analysis
of Airborne Asbestos 2-2
2.2.1 Sampling 2-2
2.2.2 Specimen Preparation for Scanning Electron
Microscopy and Transmission Electron Microscopy 2-3
2.2.3 Measurement Techniques 2-6
2.2.4 Analysis of Microscopic Data 2-8
2.2.5 Transmission Electron Microscopic Observation . 2-9
3. EXPERIMENTAL RESULTS 3-1
3.1 Scanning Electron Microscopic (SEM) Observation of . . 3_i
Standard Asbestos
3.2 Energy Dispersive (E.I).) X-Ray Fluorescence Analysis
of Standard Asbestos 3-4
3.2.1 Beryllium (Be) - Stud Specimen 3-7
3,2.2 Beryllium Oxide (BeO) - Substrate Specimen . . 3.7
3.3 Transmission Electron Microscopic (TEM) Observation
of Standard Asbestos 3-10
3.4 Analysis of Airborne Asbestos 3-13
4. DISCUSSION 4-1
4.1 E.D. X-Ray Fluorescence Analysis of Standard Asbestos
in a Field Emission SEM 4-1
-------
F-C3779
CONTENTS (cont)
Seat-Ion Title Page
4.2 Analysis of Airborne Asbestos - Overall View 4-3
4.3 Feasibility of an Automated Counting System 4-7
5. CONCLUSIONS 5-1
6. ACKNOWLEDGEMENTS 6-1
7. REFERENCES 7-1
APPENDIX A - BeO Substrate Preparation A-l
APPENDIX B - Computer Program for the Identification
of Asbestos Fibers D-l
11
-------
F-C3779
FIGURES
Kwnber Title Page
1 Schematic Diagram of the Steps Involved in Specimen Prepa-
ration ............. 2—4
2 SEM Micrographs of Johns Manville Chrysotile Asbestos, X150,
(a) Before LT Ashing, (b) After LT Ashing 3-2
3 High Resolution SEM Micrographs of Johns Manville Chrysotile
Asbestos. (a) Before LT Ashing, Structure of the Carbon
Extraction Replica of Millipore Filter, X3.000. (b) After
Partial LT Ashing, Note the same fiber in (a) and (b),
X3,000 (c) After Complete LT Ashing, X3.000 .... 3-3
4 SEM Micrographs of John Manville Chrosotile Fibers Aspirated
on a BeO Substrate, Secondary Electron Mode, (a) X300; (b)
X10,000, White spots correspond to the points where x-ray
analysis was carried out 3-5
5 X-ray Fluorescence Spectra from U.I.C.C. Standard Asbestos
Fibers and from Asbestos Fibers in Duluth Water and Phila-
delphia Air ............ 3-6
6 X-ray Fluorescence Spectra from Johns Manville Chrysotile on
Be-Stud Specimen at 18 KV and 6 x 10~10 amp, Minimum Times
used for Identification ......... 3-8
7 X-ray Fluorescence Spectra from Johns Manville Chrysotile on
BeO Substrate Specimen at 25 KV and 2 x 10~9 amp, Minimum
Times used for Identification 3-9
8 Selected Area Electron Diffraction Patterns from U.I.C.C.
Standard Asbestos Fibers. (a) Chrysotile (b) Anthophyllite
(c) Amosite, (d) Crocidolite 3-11
9 TEM Study on BeO Substrates, (a) Low-Magnification Image
of Copper Grid with BeO Substrates, Canadian Chrysotile
Processed according to Figure 1. (b) SAD Rings Arising from
Mostly Fine-Grained BeO. (c) SAD Pattern from Chrosotile
Fiber 3-12
10. TEM Micrographs of U.I.C.C. Canadian Chrysotile on BeO Sub-
strate Processed According to Figure 1. (a) X3,000, (b)
X9.000; (c) X35.000 3-14
iii
-------
FIGURES (Cont'd.)
Number Title Page
11 SEM Micrographs of Point Source Sample, #A-228, XI,000
((a) and (b) refer to different areas) 3-17
12 SEM Micrographs of Point Source Sample, //A-224; XI,000 . 3-18
13 SEM Micrographs of Near Point Source Sample, //A-223; (a)
XI,000 (b) X3.000 3-19
14 SEM Micrographs of Near Point Source Sample, //A-226, X3,000 3-20
15 TEM Micrographs of Near Point Source Sample, //A-226. (a)
X4.000, (b) 45,000 3-21
16 Cumulative Distribution of Asbestos Fiber Diameters and
Lengths for Point Source and Near Point Source Samples . 3-23
17 Cumulative Distribution of Asbestos Fiber Diameters and
Lengths for a Point Source Sample, #A-227 (A-227-1 and A-227-2
refer to differnt portions of the filter) 3-24
18 SEM Micrographs of Demolition Site Sample, XI,000. . . 3-25
19 SEM Micrographs of the FIRL Ambient Air Sample, Roof Level,
0.165m3 of Air; X100. (a) Before LT Ashing, (b) After LT
Ashing 3-27
20 SEM Micrographs of the FIRL Ambient Air Sample, Roof Level,
0.165m3 of Air; X3.000. (a) Before LT Ashing; (b) After LT
Ashing (Note: Even very small particles are in place after
LT ashing) 3-38
21 SEM Micrographs of Chrysotile Asbestos Observed in One of
the FIRL Ambient Air Sample, Roof Lebel, 10m3 of Air; (a)
Single Asbestos Fibers; X3,000; (b) Fiber Lump, XI,000 . 3-29
22 TEM Micrographs of the FIRL Ambient Air Sample, Roof Level,
10m3 of Air on 0.45u Millipore Filter, BeO Substrate, (a)
Non-asbestos Fibers, X38.000, (b) Chrysotile Fibril, X76,000 3-30
23 SEM Micrograph of Near Point Source Sample, //1000367 Collec-
ted on Millipore Filter with a Nylon Backing. Specimen
Prepared after Oven-ashing, Dispersion by Mild Ultrasonifi-
cation and Re-filtering through a Millipore Filter (Note:
Ash in the background), XI,000 3-33
iv
-------
F-C3779
FIGURES (Cont'd.)
Number Title Page
24 SEM Micrographs of Near Point Source Sample, #1000274, Par-
ticles Originally Collected on Millipore Filter with a Nylon
backing; (a) XI,000, (b) X3,000 3-34
25 Cumulative Distribution of Asbestos Fiber Diameters and
Lengths for the Sample #1000367 3-36
26 X-Ray Fluorescence Spectra from U.I.C.C. Canadian Chryso-
tile on Be-stud Obtained in a Field Emission SEM with E.D.
Analysis at 18Kv and 3 x 10-9 amp 4-2
27 X-ray Fluorescence Spectra from U.I.C.C. Canadian Chryso-
tile on BeO Substrate Obtained in a Field Emission SEM with
E.D. Analysis at 18Kv and 2 x 10~9 amp 4-4
28 Field Emission SEM Micrographs of U.I.C.C. Canadian Chryso-
tile on BeO Substrate, Prepared According to Figure 1, (a)
X6.000, (b) 300A Chrysotile Fibril from which x-ray spectrum,
shown in Figure 27, was obtained in the "scan mode",
X30.000 4-5
TABLES
Number Title Page
1 Analysis of Airborne Asbestos Samples 3-16
2 Details of Samples Analyzed in Table III 3-32
3 Analysis of Airborne Asbestos Samples 3-35
-------
F-C3779
ABSTRACT
The methodology that was developed at the Franklin Institute Research
Laboratories (FIRL) under the EPA Contract No. 68-02-0544, for the
determination of airborne asbestos has been perfected and developed
further. Moreover, the newly perfected technique has been applied to
point source, near point source and ambient air samples.
This report describes the analytical method which employs a
scanning electron microscope equipped with energy dispersive x-ray
analysis for the identification and counting of airborne asbestos. The
specimens, prepared in a unique manner, are suitable for image analysis
and for a possible automated counting system.
Results of the analysis on airborne asbestos are presented, and
limitations and advantages of the present technique are discussed.
VI
-------
F-C3779
1. INTRODUCTION
The problem of asbestos in the environment is a very severe one, arising
from the great increase in usage of asbestos and the increasing frequency of
observation of "asbestos bodies" in autopsy material. The carcinogenic
properties of asbestos are well documented and the "asbestos bodies"
have been verified to contain asbestos fibers in their cores * . Occupa-
tional exposure to asbestos dust is associated with serious risks of lung
scarring and neoplastic disease * . Similar health hazards exist, although
to a lesser extent, with indirect occupational exposure ' , with family
contact, ' and as a result of neighborhood and environmental exposure '
In New York City, chrysotile asbestos fibers have been identified in the lungs
(3)
of persons with no known occupational asbestos exposure
The biological effects of airborne and xgraterborne asbestos have been
>resei
(18)
discussed recently in an international meeting . Moreover, the present
controversies on the safety of asbestos have been narrated by Wagner
The present situation is that exposure to a high dosage of airborne asbestos
is carcinogenic and that the carcinogenic effects of exposure to a constant
and low dosage of airborne asbestos remains to be established. '
The influence of asbestos type, fiber size, and cofactors on biologic
effects are not well documented . Moreover, the health effects of the
size distribution of asbestos fibers are not fully established. The
(19)
investigation by Stanton and Wrench indicates that respirable asbestos
fibers with diameters in the range of 0.5y to 5\i and up to 80y in length
result in the development of cancer in hamsters; fibers either smaller or
larger than this range are harmless. However, single chrysotile fibrils with
o (3 4)
diameters ^ 400A have been observed in human lung tissue * . Similar
(18)
experiments by Wagner indicate that irrespective of the mineralogical
nature, all fibers less than 0.5p in diameter may produce tumors if
(18)
innoculated into the pleural cavity of rats. In addition, evidence indicates
that the physical form of the fibers is an important factor in the development
of experimental mesotheliomas, and that the crocidolite fiber having a particularly
fine needle-like shape is more carcinogenic than other types of asbestos.
1-1
-------
F-C3779
Fiber size is critically important in determining respirability, deposition,
retention and clearance from the pulmonary tract and is probably an important
determinant of the site and nature of biologic action. Timbre11 has
carried out research in this direction and has shown that the large differences
observed in the incidence of asbestos—associated cancer of the pleura in some
geographical locations are likely to be related to the size distribution of
asbestos fibers. Since fiber length affects deposition, longer fibers
(21)
apparently have greater fibrogenic effects . The investigations by
(22)
Webster (reported in ref. 17) and by Cunningham and Pontefract indicate
that very little ingested asbestos penetrates the walls of the stomach and
colon, and that almost all that does penetrate is of the smaller crysotile
type. The above studies show the importance of determining the original
size distribution of airborne asbestos,
Several experimental techniques have been investigated to determine
asbestos in the environment. There are, however, many uncertainties as to
the best method of sampling, identifying, and quantifying asbestos in the
environment and interpreting the data so obtained . The monitoring
problem lies in identifying a very small number of asbestos fibers against
a background of a very large number and variety of other particles in the
same sample.
An x-ray diffraction technique has been carried out on airborne
(22)
asbestos samples . However, only the total amount of asbestos, without
any reference to size distribution, can be determined by this technique.
3 3
In addition, only amounts greater than ^ lOyg/m (10,000 ng/ra ) can be
determined by this method. The amount of asbestos in ambient air
3
is believed to lie in the range of 0 to 5,000 ng/m . New refinements in
specimen preparation technique are underway in some laboratories so that
the detection limit by x-ray diffraction technique is expected to be lower
3 (23)
than ^ I0yg/m
(3 24-29)
Transmission electron microscopy (TEM) has been applied ' ' in
the determination of airborne asbestos. There are many limitations in the
use of TEM techniques for the determination of asbestos in the environment.
(24,25)
Some of the TEM techniques developed only give the total mass of asbestos
1-2
-------
F-C3779
( 26—29"i
and some give mass as well as size distribution . A TEM technique
is not easily amenable to automation. Moreover, fiber identification is
very often based on morphology in this technique, which is applicable
only for chrysotile. Selected area electron diffraction (SAB) can be
(30-33)
used to identify particular type of asbestos ; however, SAD patterns
(34)
can only be obtained from fibers in a particular size range . Hence,
asbestos fibers only in a limited size range can he identified by TEM.
Urban air particulate and airborne asbestos identification and
characterization by combining transmission electron microscopy and x-ray
/oc 'lA')
microanalysis (EMMA) have been attempted recently ' . Recently,
(37)
Maggiore and Rubin have employed a scanning electron microscope (SEM)
with a field emission source to identify asbestos fibers using energy
(37)
dispersive x-ray fluorescence analysis. The above study ' has been
carried out using only standard asbestos samples and its applicability
to airborne asbestos has not been clearly delineated. Moreover, no
attempt was made to develop the system for automated identification
and counting of asbestos fibers.
Nearly all methods published in the literature for sizing and counting
of airborne asbestos do not preserve the original size distribution of
(38")
asbestos fibers . Specimen preparation for SEM or TEM may include the
following steps: filtration, centrifuging, ashing and ultrasonification.
(38)
An unknown quantity of asbestos fibers are lost during centrifuging
Splitting and loss of asbestos fibers may be visualized during ignition in
an oven ashing. Ultrasonification is known to break down chrysolite asbestos
(24 25)
fibers into fibrils ' . The measurement and transfer of small volumes
by a micropipet or syringe constitute a significant fraction of the total
( 38}
error in the enumeration of asbestos fibers
As a step towards solution to this particulate monitoring problem, a
methodology has been developed in our laboratory under contract to
Environmental Protection Agency, EPA (68-02-0544). The technique is described
(39)
in the EPA reportv .
1-3
-------
F-C3779
The present report describes further developments that have been
carried out under a continued grant from EPA. Results obtained from
airborne asbestos are also presented. The current technique identifies
and counts the asbestos fibers, and determines their original fiber
size distribution using a scanning electron microscope with automated
energy dispersive x-ray analysis. The specimen preparation technique
aims at preserving the original size distribution of airborne asbestos and
is such that a completely automated operation using a SEM, x-ray analyzer
and image analyzer is feasible. At present, the developed technique is
used in a 'manual mode' and results so obtained on airborne asbestos are
presented here. The technique gives mass concentration, particle
concentration and original fiber size distribution of airborne asbestos.
1-4
-------
F-C3779
2. EXPERIMENTAL PROCEDURE
2.1 Statement of the Problem
The monitoring problem can be stated as "identifying and counting a
very small number of asbestos fibers against a background of a large number
and variety of other particles". This implies that one is interested in
determining both the amount and the size distribution of asbestos fibers
as they exist in air. Any steps to isolate and concentrate asbestos
fibers with respect to other particles should aim to preserve the original
size distribution of the asbestos fibers. There is a greater chance for
fibrillation during the specimen preparation steps of oven ashing and
ultrasonification. Low temperature oxygen plasma ashing is preferable
to oven ashing. Unknown quantities of fibrils are lost during centrifuging,
and, as stated previously, measurement and transfer of small volumes of
liquid containing dispersed particles by a micropipet or syringe constitute
a significant fraction of the total error in the enumeration of fibers
using electron microscopy.
The present technique consists of specimen preparation steps that
reduce fibrillation to a minimum and aims at determining original asbestos
fiber size distribution. Moreover, the technique aims ultimately at
identifying and counting asbestos in air in an automated manner. Initial
sample collection is by absolute filtration using a membrane filter. The
final specimen for scanning electron microscopy (SEM) or for transmission
electron microscopy (TEM) prepared from the airborne particles should be
on a featureless background so that the particles can be analyzed unequivo-
(39-41)
cally by an image analyzing system . Furthermore, the x-ray emission
from the asbestos fibers should be without either significant background
or any extraneous x-ray emission peaks from other particles or the substrate,
Similarly, selected area electron diffraction (SAD) patterns from a fiber
obtained in the TEM should not be superimposed by SAD patterns from
neighboring particles. Hence, the amount of air filtered should be such
2-1
-------
F-C3779
that little or no superposition of particles takes place. All operating
features of the final system should be compatible with automated image
analysis
2.2 Description of Methodology for the Analysis of Airborne Asbestos
Following considerable experimentation a satisfactory technique to
analyze airborne particles by SEM along with energy dispersive x-ray
fluorescence analysis and image analysis was developed under an EPA
(42)
grant . Further developments have been achieved in the specimen
preparation techniques which are presented here. We believe that this
technique is reliable and could be repeated in any laboratory without
undue expense or difficulty.
2.2.1 Sampling
Airborne particulates are first collected on a Millipore membrane
(MF type: mixed esters of cellulose) filter having a diameter of 47mm
and an average pore size of 0.45u. Though the Millipore filter of
the above specification is preferred, the present technique is applicable
to MF-type Millipore filters of different pore sizes and diameters. The
pore size of the filter is not a critical variable but does influence
the superposition of particles. Using a 0.45y Millipore filter, a large
number of ^ O.lp size particles are collected on the filter. These air-
borne particles are not collected on a 0.8u pore size Millipore filter^42\
Even though the 0.8p pore size is larger than the largest dimension of
/27\
some asbestos fibrils, it has been observed ' that the surface properties
of the membrane filter and the asbestos fibers, as well as the circuitous
path through the filter, result in virtually complete collection of all
asbestos material. However, no systematic study was undertaken to verify
such a proposition in the present investigation.
2-2
-------
F-C3779
3 3
The volume of air filtered lies in the range of 2m to 15m for a
47mm diameter Millipore filter with an average pore size of 0.45u. If
3
the sample volume exceeds ^ 15m , overcrowding of particles occurs. The
rate of filtration is not deemed to be an important variable; however,
longer time of filtration with a slow rate is recommended.
2.2.2 Specimen Preparation for Scanning Electron Microscopy and
Transmission Electron Microscopy
The steps involved in the specimen preparation are shown schematically
in Figure 1. The Millipore filter with collected airborne particles is
mounted on a clean metal disc; mounting is done with rubber cement which
o
is put only at the periphery of the filter. A thin carbon layer (^ 100A)
is evaporated onto the filter in a vacuum evaporator. A thicker carbon
layer is not recommended since the time to eliminate the carbon extrac-
tion replica during low temperature (LT) ashing increases considerably.
A *v« 25mm diameter circular piece of the "composite" film is now cut
out by a razor blade. The composite film is held on an electroplished
single crystal berryllium (Be) stud which is featureless and mounted
with carbon side down, as shown in Figure l(b), in a brass specimen
holder. Single crystal beryllium studs exhibit better featureless
surface than poIyerystalline studs which usually contain oxide inclusions.
If TEM study along with SEM observation is being contemplated, then
two or three BeO substrates on copper grids are inserted between the Be-stud
and the composite film, Figures l(a)(iii) and l(b). The BeO substrate
side of the copper grid faces the carbon side of the composite film.
Details of the preparation of BeO substrates are given in Appendix A.
The specimen holder, Figure l(b), is next put in a covered petri
dish containing acetone which is a solvent for mixed esters of cellulose.
The level of acetone in the petri dish is such that the solvent does
not attack the membrane filter directly, "Figure 1 (b). The specimen
undergoes acetone vapour attack for about half an hour. Initially.
the composite film swells under the solvent attack but settles down on
2-3
-------
—
PARTICLES
CAR80N
P "P 9 _
ix^x y x >i
v
MILLIPORE
FILTER - TYPE
(MF)
r^Blfo®
CARBON EXTRACTION
REPLICA
(o)
(MI)
MILLIPORE FILTER
(iv!
Ni
BRASS SPECIMEN
HOLDER
ACETONE LEVEL
IN PETR1 DISH
.ORGANIC PARTICLES
AND CARBON FILM
ELIMINATED
m-
LOW-TEMPERATURE
PLASMA ASHING
3 mm
SEM
SPECIMEN
TEM
SPECIMEN
(USED IF NECESSARY)
(b)
FEATURELESS
BeO SUBSTRATE
ON COPPER GRID
Figure 1. Schematic Diagram of the Steps Involved in Specimen Preparation
-------
F-C3779
the Be-stud, which can be observed visually. The settled composite
film adheres to the Be-stud in such a way that the following steps can
be undertaken.
The brass specimen holder, Figure l(b), serves to keep the film in
position while swelling occurs. During solvent vapor attack, the
Millipore filter is rendered transparent, Occassionally one or two
bubbles form when the film is settling down on the Be-stud causing breaks
in the continuous film. This does not affect the general procedure
and on an average, 90% of the fj'in LX M.I ins continuous.
After the film has completely settled down on the Be-stud, the
brass specimen holder is placed in a beaker in an inclined position
(^30°) and acetone is poured slowly in order to immerse the film com-
pletely. The solvent dissolution of the filter is carried out for one
to two hours; accelerated dissolution of the filter by stirring the
solvent is not recommended because the fragile carbon extraction replica
is often disturbed or broken up by mechanical forces. Later the acetone
is sucked out of the beaker slowly using a large capacity pipet. After
the specimen is dry the Be-stud is put in a low temperature oxygen
plasma asher for nearly 12 hours (overnight). The plasma stream flows
directly over the stud ashing all the organic particles and the carbon
extraction replica. This results in the collection of all inorganic
particulates on the featureless Be-stud without disturbing the original
particulate distribution. No evidence has been found as to the loss
of small particulates along the plasma stream in order to warrant
indirect plasma flow over the particulates. The rate of flow of oxygen
through the system is ^ 0.3 liters/rain. The oxygen plasma is excited
by a high frequency induction coil. Specimens prepared from standard
U.I.C.C. asbestos fibers exhibit no degradation of the fibers due to
plasma ashing when observed either in the SEM or in the TEM. When the
specimen; cools down to the room temperature, it is ready for EM obser-
vation.
2-5
-------
F-C3779
If TEM copper grids with BeO substrates had been used, they are
removed from the Be-stud carefully with a pair of tweezers. The Be-studs
can be used repeatedly if the particles are removed by repeated ultra-
sonic cleaning in acetone.
In summary, the above technique results in two types of specimens,
namely, one ^25mm size Be-stud and one or two EM grid specimens (BeO
substrate), on which all the airborne particulates lie on a featureless
background. The Be-stud specimen is used in the SEM primarily. The
BeO substrate specimen is either used in the SEM or in the TEM depending
on the type of Investigation being contemplated, i.e., either x-ray
fluorescence analysis or SAD analysis of asbestos fibers respectively.
The Be-stud or the BeO substrate gives rise to no extra x-ray fluorescence
peaks during an energy dispersive (E.d.) analysis either in the SEM
or in the TEM.
2.2.3 Measurement Techniques
Even though a feasibility study of a completely automated counting
(42)
system for airborne asbestos has been carried out , the following
summarizes counting procedure for 'semi-automated1 mode of operation.
Most of the steps are carried out manually except the identification
of asbestos fiber, which is performed by a computer.
Operational modes are described for a scanning electron microscope,
o
JSM-50A, with a guaranteed resolution of M.OOA. The E.D. x-ray analyzer,
Northern Scientific NS-880, has a resulution of 160ev and is highly
computerized with respect to handling x-ray fluorescence data.
The Be-stud specimen is first observed in the 'picture-mode1 at a
condenser lens setting, which gives an absorbed electron current of ^2
x 10 amps. At this setting, the resolution of the microscope is
o _1Q
VLOOA. However, an electron current of >. 2 x 10 is necessary for x-ray
fluorescence analysis giving reasonable count rates. An electron current
of ^ 2 x 10 is an optimum value for which the resolution is adequate
and count rates are reasonable. The accuracy of determining the size
2-6
-------
F-C3779
of fibers decreases to a certain extent at a current of 'v 2 x 10 amps.
due to a loss of resolution. Generally, the fibers appear larger than
their actual sizes. The asbestos fibers appear brighter at this current
of 2 x 10 amp. due to increased secondary electron yifild. Iterating
between the two current levels of 2 x 10 amps, and 2 x 10 amps.
is time consuming since an image shift is observed when the electron
current is changed.
A magnification of XI,000 or X3,000 is suitable for observing and
counting airborne asbestos, particularly for near point-source samples
which usually contain large fibers as well as fibrils. Fibrils can be
detected at a magnification of X3,000 which was the usual magnification
used in the present investigation.
The E.D. x-ray fluorescence analysis for the identification of
asbestos fibers is carried out in the following manner. First, the
number of fibers and fibrils are counted on the viewing screen and
the size (diameter and length) of the fibers is recorded at a magnifi-
cation of X3,000. The magnification is usually increased to a suitable
higher magnification for accurate determination of the size of relatively
small fibers. Each fiber is then brought to the center of the screen
approximately and the magnification increased to as high a value as neces-
sary to carry out x-ray fluorescence analysis. The 'picture mode1 is
changed to a 'small-square-scan1 mode (5mm x 5mm on the screen) and the
small square is centered on the fiber accurately. Iteration between
the 'picture mode' and the 'small-square-scan' mode is done to verify
that the small square lies indeed on the fiber. X-ray counts are
accumulated from 5 sees, to 200 sees, depending on the fiber size; small
fibers take a longer time to yield significant Mg, Si, Ca and Fe peaks*
The type of asbestos fibers (e.g., chrysotile, amosite etc.) is identi-
fied by their characteristic x-ray fluorescence peaks by a computer
program on the NS-880, Appendix B. The same procedure is repeated for
other fibers in the field of view. After all the fibers in a field of
view have been analyzed, the field of view is changed to another location.
2-7
-------
F-C3779
The number of fields of view usually observed for the analysis of
asbestos fibers in a statistical manner ranges from 25 to 100 depending
on the sample. Counting is continued until 60 to 100 asbestos fibers
have been identified 'and their sizes recorded. The fields of view
observed usually form a square-grid pattern that covers as much area
as possible on the 'v 25mm diameter Be-stud. Hence, a better statistical
average is obtained than when scanning a small area out of a 47mm dia-
meter filter paper.
2.2.4 Analysis of Microscopic Data
The total amount (number or mass) of asbestos and its size dis-
tribution are determined in the present technique. Twenty-five to a
hundred areas are scanned at a convenient magnification, preferably
X3.000, for a statistical count using the following two formulae:
Average number of fibers X K
3
Volume of air filtered (m )
= Number of fibers/m (1)
Average mass of fibers X K ^
5— = Mass of fibers/m (2)
Volume of air filtered (m )
where
2
Filter area (mm )
K = — 2
Field area (mm )
The mass of an asbestos fiber is calculated from its length and diameter
/On /o\
knowing its density ' .
Recognizing that no conclusive results are available as to determine
which physical parameters of asbestos fibers control health hazards, the
2-8
-------
F-C3779
size distribution of asbestos fibers is represented by cumulative frequency
distributions of lengths and diameters separately. Moreover, respirable
and non-respirable fibers are analyzed together in the present investigation.
The total number of particles that need to be counted for presenting
a statistically reliable cumulative frequency distribution is debatable and
depends on the technique used to count them . The total number of asbestos
fibers identified and counted in the present investigation lies in the range
of 60 to 100 in order to get a good distribution plot. The total time
necessary to carry out such an analysis in the SEM lies between 2 and 4
hours.
2.2.5 Transmission Electron Microscopic Observation
Usually the. Be-stud specimen was used for identification and counting
of airborne asbestos in the present investigation. However, transmission
electron microscopic (TEM) observation using the BeO substrate specimen was
occasionally warranted in some cases. It has been well documented » »• » »
that partial leaching of Mg from chrysotile asbestos fibers occurs in aqueous
and some other environments. Under such circumstances, the chemical composi-
tion of the asbestos fiber changes, which hinders unique identification of
the type asbestos by x-ray fluorescence analysis. Fortunately, the amount
of Mg-leaching is such that even though the chemical composition changes
partially, the chrystalline structure still remains intact. Under such
situations, asbestos fibers can be identified by selected area electron
diffraction (SAD) patterns obtained in a TEM( ~33'.
However, TEM study of asbestos fibers is very cumbersome and time
consuming, and SAD pattern can only be obtained from asbestos fibers in a
(34)
particular size range. Nevertheless, the use of TEM for identification
of asbestos was occasionally warranted in the present investigation.
Airborne asbestos does not usually exhibit any change in chemical composition
in contrast to waterborne asbestos.
The BeO-substrate specimen can also be used in new generation transmission
electron microscopes in which E.D. x-ray fluorescence analysis can be carried
(47)
out with great advantage in addition to usual SAD analysis.
2-9
-------
F-C3779
Additionally, the BeO-substrate specimen can also be used in a scanning
(37)
electron microscope either in a * transmission mode' or in a 'semi-
transmission mode' (to be described in Section 3.1). Under these two modes
of operation, the x-ray fluorescence spectrum from an asbestos fiber has a
negligible background spectrum, which aids identification considerably
(Section 3,2.2) and smaller fibers and fibrils can be identified with more
confidence than that is achievable in the standard SEM mode.
2-10
-------
F-C3779
3. EXPERIMENTAL RESULTS
3.1 Scanning Electron Microscopic (SEM) Observation of Standard Asbestos
A specimen was prepared by aspirating Johns Manville chrysotile on
a 0.45]j Millipore filter and then following the sequence of steps shown
in Figure 1. The specimen was observed in the JSM-50A SEM before and
after low temperature (LT) oxygen plasma ashing. A representative
area at a relatively low magnification is shown in Figure 2. Two organic
fibers (upper left and right) and the carbon extraction replica of the
Millipore filter have been eliminated. Some of the small fibers which
were not visible in Figure 2(a) appear distinctly in Figure 2(b).
The advantages of LT ashing are better shown in Figure 3. The
structure of the carbon extraction replica with an embedded fiber is
clearly visible in Figure 3(a); the same area is shown in Figure 3(b)
after partial LT ashing. As the carbon extraction replica is eliminated
due to oxidation of carbon, the embedded fibers are freed and lie on the
featureless Be-stud. After complete ashing of the carbon replica, fibers
and fibrils appear distinctly on a featureless background, Figure 3(c).
This LT ashing step has been developed: during the present program
and constitutes an additional step to the technique that was originally
(42)
developed under an earlier EPA contract . Beryllium studs stand LT
ashing without any degradation unlike pyrolytic graphite studs used
(42)
earlier . The LT ashing step has the following advantages:
(i). It is a concentration process. Since all organic parti-
culates are eliminated, the ratio of the asbestos fibers
to all other particles increases.
(ii) The visibility of small fibers and fibrils in the SEM improves
considerably.
(iii) The x-ray fluorescence spectra from small fibers and fibrils
are improved considerably since background spectrum from a
physically rough surface, Figure 3(a), is eliminated. This
will be discussed further in Section 3.2.1
3-1
-------
F-C3779
(a)
(h)
Figure 2. SEM Micrographs of Johns Manville Chrysotile Asbestos, XI50.
(a) Before LT Ashing, (b) After LT Ashing
3-2
-------
F-C3779
(a)
(b)
(c)
Figure 3. High Resolution SEN Micrographs of Johns Manville Chrysotile
Asbestos, (a) Before LT Ashing, Structure of the Carbon
Extraction Replica of Millipore Filter, X3,000. (b) After
Partial LT Ashing, Note the same fiber in (a) and (b), X3,000
(c) After Complete LT Ashing, X3,0f>n
3-3
-------
F-C3779
It is of interest to note here that the SEM micrographs, shown
o
in Figure 3, were taken at a resolution of M.OOA. However, a specimen
containing relatively large asbestos fibers can be observed and analyzed
without the LT ashing step, Figures 2(a) and 3(a) particularly when the
SEM is operated at a higher electron current. With higher electron
current and correspondingly lower resolution, inorganic fibers and
particulates appear brighter with respect to a tolerable background
appearing from the carbon extraction replica. This was the method used
(42)
earlier in this laboratory to analyze airborne asbestos. However,
fibril observation and identification, even though possible, was not
easily carried out (see Figures 7 and 8 in ref. 42).
SEM micrographs of Johns Manville chrysotile aspirated on a BeO
substrate, held on a EM copper grid, are shown in Figure 4. The copper
grid is placed on a graphite block with a cavity, and conductive graphite
paint is used to reduce charging effect. Moreover, unlike silver paint,
graphite paint is not detected by E.D. x-ray detector. The specimen
is observed in the secondary electron mode, Figure 4, and E.D. x-ray
fluorescence analysis is carried out simultaneously. The advantage of
using the BeO substrate instead of the Be-stud under similar conditions
is that the background spectrum produced by bremsstrahlung is reduced
to a minimum during x-ray analysis, Section 3.2.2., since the electron
beam penetrates through the BeO substrate. This mode of operation is
referred to here as 'semi-transmission1 mode in contrast to true trans-
(37)
mission mode , in which a transmission electron detector is used under
the specimen.
3.2. Energy Dispersive (E.D.) X-Ray Fluorescence Analysis of Standard
Asbestos
Asbestos fibers are identified by their characteristic x-ray fluores-
cence spectra, Figure 5, in the present investigation. E.D. x-ray
analysis is carried out using a highly computerized Northern Scientific
NS-880 system interfaced with the JSM-50A SEM. Unknown asbestos spectrum is
identified by matching it with the spectra from standard U.I.C.C. asbestos;
3-4
-------
F-C3779
(a)
I
(b)
Figure 4. SEH Micrographs of Johns Manville Chrysotile Fibers
Aspirated on a BeO Substrate, Secondary Electron Mode.
(a) X300; (b) X1Q.QOO, White spots correspond to the
points where x-ray analysis was carried out.
3-5
-------
F-C3779
O.«TJCT-M
UICC
(!.>•„, v, sn:.)
"^N. U-
t - L^
/scent» rixo i»
elktrlK «JlTt<
{!.:_. -., rj .)
LI. .
Figure 5. X-ray Fluorescence Spectra from U.I.C.C. Standard Asbestos Flbel
and from Asbestos Fibers in Duluth Water and Philadelphia Air.
3-6
-------
F-C3779
this is carried out automatically be a computer program, Appendix B.
E.D. x-ray analysis was carried out either on Johns Manville chryso-
tile or on U.I.C.C. Canadian chrysotile in order to determine the
minimum time required to identify asbestos fibers of all sizes. However,
not all the variables were optimized in the SEM-E.D. x-ray analysis
system in the present study. Chrysotile was preferred to other asbestos
(43)
since 90% of asbestos used in this country is chrysotile and the
o o
smallest fibril (300A - 400A) that can be observed among different
(32)
asbestos belongs to the chrysotile type.
3.2.1 Beryllium (Be)-Stud Specimen
E.D. x-ray fluorescence analysis of Johns Manville chrysotile on a
Be-stud specimen was carried out with the NS-880 system interfaced with the
JSM-50A SEM. The minimum times required to identify chrysotile fibers
at 18kV and at 6 x 10 amp. were determined and the data is presented
in Figure 6. It was observed that an unwanted background spectrum produced
by bremsstrahlung is reduced with decreasing applied voltage. Using the
Be-stud specimen, the smallest chrysotile fiber that can be confidently
identified is O.lSp in diameter. Chrysotile fibers with O.lp diameter
could be identified sometimes but not always. Counting longer than 150
sees, on O.ly diameter fibers, Figure 6, does not improve the spectrum
significantly.
3.2.2 Beryllium Oxide (BeO) - Substrate Specimen
Similar studies were carried out using BeO-substrate specimen; the
SEM was operated in a 'semi-transmission1 mode. X-ray fluorescence
spectra from chrysotile fibers of similar diameters, Section 3.2.1, are
presented in Figure 7. Even though the minimum times required to identify
asbestos fibers of similar diameters studied in Section 3.2.1 are not
reduced, the spectra are much more distinct due to a reduced background
spectrum produced by bremsstrahlung (compare similar spectra on 0.2p
and O.lp diameter fibers). The smallest fiber analyzed in this mode was
3-7
-------
1000
u>
oo
t
Mg
I/)
2
UJ
Mg
1.0 p. ,15 SEC.
lOOOr- Si
40O|-
0.2^,70 SEC.
Fe
O.I ^,150 SEC.
Fe
ENERGY
Figure 6, X-ray Fluorescence Spectra from Johns Manville Chrysotile on Be-Stud Specimen at 18 KV and
6 x 10~'° amp, Minimum Times used for Identification.
o
CO
-------
350i-
(0
si
1.0 At, 15 SEC.
FE
35 Of-
Sl
M«
0.2 ii, 90 SEC.
VWV*^^^
U)
J
SO
O
U
H
CO
z
111
I-
200|-
si
MO
0.5 u, zs sec.
FE
ENERGY
ZOOi-
MG
o.ia, 200 SEC.
FE
Fiqure 7. X-ray Fluorescence Spectra from Johns Manville Chrysotile on BeO Substrate Specimen
at 25 KV and 2 x 1(H* amp, Minimum Times used for Identification
r>
UD
-------
F-C3779
O.ly fiber; however, it may be concluded from Figure 7 that even smaller
than O.ly fibers can be identified, probably O.OSp diameter fiber.
Study of Figures 6 and 7 leads one to conclude that BeO substrate
specimens are superior to Be-stud specimens as far as positive identifi-
cation of small asbestos fibers are concerned. However, other considera-
tions make the Be-stud preferable to the BeO substrate as delineated below.
(i) A 25mm diameter Be-stud specimen is easier to handle.
(ii) The overall area on the Be-stud is about X50 that on the
BeO sample giving much better counting statistics.
(iii) The continuous featureless background of the Be-stud specimen
is more amenable to automated image analysis than the BeO
substrate specimen (compare Figure 2 with Figure 4(a).
3.3 Transmission Electron Microscopic (TEM) Observation of Standard
Asbestos
The BeO substrate specimen can be either used in a SEM in the 'semi-
transmission1 mode or in a TEM for selected area electron diffraction
(SAD) analysis. Chrysotile asbestos fibers in aqueous enviroment may
lose magnesium * ; hence, there identification by E.D. x-ray fluores-
cence analysis in the SEM is not reliable. Under such circumstances,
SAD analysis in the TEM complements SEM-E.D. x-ray analysis. Moreover,
BeO substrate specimens can be used in new-generation TEMs in which
/oc lA}
E.D. x-ray analysis as well as SAD analysis can be carried out. *
Selected area electron diffraction (SAD) patterns of standard
/ ")f\ 10 O/ ^
asbestos fibers are unique ' ' and can be used to identify unknown
asbestos fibers, Figure 8. A BeO substrate specimen prepared according
to Figure 1 is shown in Figure 9(a); the specimen contained U.I.C.C.
Canadian chrysotile. An SAD pattern from the substrate is shown in
Figure 9(b), which indicates that the substrate is crystalline; the
substrate consists mostly of very fine grained BeO with some Be crystal-
lites. However, the single crystal SAD pattern from a chrysotile fiber
is observed distinctly, Figure 9(c), on the background of faint BeO rings.
When an asbestos fiber occupies most of the selected area, the BeO rings
(arising from a small fraction of the selected area) appear faint.
3-10
-------
F-C3779
(a)
(b)
(c)
(d)
Figure 8. Selected Area Electron Diffraction Patterns from U.I.C.C.
Standard Asbestos Fibers, (a) Chrysotile (b) Anthophyllite
(c) Amosite, (d) Crocidolite.
3-11
-------
F-C3779
(a)
(b)
(c)
Figure 9. TEM study on BeO Substrates, (a) Low-Magnification Image
of Copper Grid with BeO Substrates, Canadian Chrysotile
Processed according to Figure 1. (b) SAD Rings Arising
from Mostly Fine-Grained BeO. (c) SAD Pattern from Chrosotile
Fiber.
3-12
-------
F-C3779
The BeO substrate is, more or less, featureless and Canadian chry-
sotile fibers, as observed on such a substrate which has been processed
thorugh the steps shown in Figure 1, are shown in Figure 10 at different
magnifications. At a relatively high magnification, Figure 10(c), the
substrate is not truly featureless but it is believed that it will not
(39 41)
exclude image analysis * . Beryllium (Be) was evaporated onto a
substrate kept at room temperature, Appendix A; this resulted in a film
which is shown in Figure 10. It is expected that a more featureless
BeO substrate can be prepared by depositing Be on a colder substrate.
Additionally, the substrate may turn out to be amorphous which will
result in a fewer diffuse rings in comparison to those shown in Figure
9(b).
3.4 Analysis of Airborne Asbestos
Airborne asbestos samples, comprised of point-source (inside an
asbestos manufacturing factory), near point source (outside an asbestos
factory) and ambient air samples, have been analyzed using the FIRL
technique. Specimens were prepared using the steps shown in Figure 1
whenever possible. Primarily the SEM-E.D. x-ray analysis was used for
airborne asbestos analysis (Sections 3.1 and 3.2); occasionally, the
TEM-SAD technique (Section 3.3) has been applied to some samples.
At the termination of this investigation an important variable,
the magnification used for observing and counting of asbestos fibers,
had not been finalized. A magnification of X3,000 is probably the most
suitable for analyzing airborne asbestos samples which contain relatively
large fibers as well as fibrils, Figure 3(c), although, a magnification
XI,000 was used for some samples. Further investigations are necessary
to clarify the influence of magnification on the statistical signifi-
cance of the counts obtained.
3-13
-------
F-C3779
V
(a)
,..-
(b)
0.5
" . t-
(c)
Figure 10. TEM Micrographs of U.I.C.C. Canadian Chrysotile on BeO
Substrate Processed According to Figure 1. (a) X3.000,
(b) X9,000; (c) X35,000
3-14
-------
F-C3779
The optimum magnification for counting asbestos fibers and deter-
mining the size distribution seems to depend on the type of samples
being analyzed. The magnification chosen should be such that only
a few fibers are seen on the viewing screen. For a sample containing
relatively large fibers, a magnification of XI,000 is more convenient
than X3.000; but the chances of missing fibrils obviously increases.
The equations used for statistical analysis, Equations [1] and [2],
(Section 2.2.4) are independent of magnification. Nevertheless, variable
statistics have been obtained at different magnifications in preliminary
tests. This result will not be presented here, as it meeds more explo-
ration. The following results are based on an assumed insensitivity
to the magnification used.
A number of point source and near point source samples had been
(42)
analyzed in a previous investigation on a preliminary basis, i.e.,
each fiber in them was not identified and the specimen preparation
did not include the LT ashing step. The same samples have been analyzed
in the present investigation systematically following all the steps
shown in Figure 1 for the specimen preparation. The data is presented
in Table 1 and representative SEM micrographs are shwon in Figure 11 to
15.
The point source samples, Figure 11 and 12, contain relatively
large fibers with some fiber clumps. The near point source samples
also contain some fiber clumps and single asbestos fibers, Figures 13
to 15. SEM micrographs and TEM micrographs of the same near point
source sample are shown in Figure 14 and 15 respectively. In manual
operation, the operator can make judicious judgement in counting the
fiber clumps as to their sizes, which is necessary for the calculation
of mass concentrations. Some of the chrysotile fibers appear to coexist
with other particles, Figure 14(b). The E.D. x-ray analysis is carried
out in relatively free portions of the fibers; a similar approach is
applicable to SAD analysis. If the E.D. x-ray analysis is carried out
on the fiber very near to other coexisting particles, then secondary
x-ray fluorescence peaks usually appear from the extraneous particles.
3-15
-------
Sample*
Type-
A-228
A-224
A-227-1
A-227-2
A-226
A-223
PS
PS
PS
PS
NPS
NPS
Demol i tion
Site
TABLE 1
ANALYSIS OF AIRBORNE ASBESTOS SAMPLES
Volume of
Air (m3)
0.135
0.090
0.135
0.135
0.135
0.135
0.96
0.3
Magni fi cat ion
Used
XI ,000
X3,000
X3,000
X3,000
XI, 000
XI ,000
XI, 000
X3,000
Number of
Asbestos
Fibers
Counted
97
24
88
72
60
56
9
6+21 umps
Fiber
Concen.
(*/«&
x 103
3,800
1,900
5,100
7,800
2,200
400
11
2
Fiber
Concen.
( gm/m3)
x!0"9
200,000
530
795,000
478,000
14,300
153,000
4,000
67
Ambient Ai r
(Philadelphia)
"Different designations like A-228, etc, refer to different places inside or outside the
factory and to different times and dates.
**PS: Point Source (Inside factory)
NPS: Near Point Source (20 feet from factory)
Fiber Type
Chrysotile
Mostly Chrysotile
Chrysotile
Mostly chrysotile,
some tremolIte
Equal amounts of
chrysotile and
tremolite
Chrysotile and
tremolite
Chrysotile
10
-------
F-C3779
(a)
(b)
Figure 11. SEM Micrographs of Point Source Sample, #A-228, XI,000
((a) and (b) refer to different areas)
3-17
-------
(a)
(b)
Figure 12. SUM Micronraohs of Point Source Sample, #A-224; XI,000.
3-18
-------
F-C3779
(a)
(b)
Figure 13. SEM Micrographs of Near Point Source Samele, #A-223;
(a) XI,000 (b) X3,000
3-19
-------
F-C3779
(c)
Figure 14. SEM Micrographs of Near Point Source Sample, M-226,
X3.000.
3-20
-------
F-C3779
(a)
(h)
Figure "15. TEM Micrographs of Near Point Source Sample, #A-226.
(a) X4.000, (h) 45,000
3-21
-------
F-C3779
The difference between point source and near point source (20 ft. from
the factory) samples is well exhibited in their size distribution, Figure
16. Both the samples contain asbestos fibers having a similar range
of diameters; the near point source sample contains more fibers with
smaller diameters. However, their length distributions are distinctly
different. The near point source sample rarely contains asbestos fibers
having lengths greater than 30y, Figure 16.
The importance of determining size distribution of asbestos fibers
is also exhibited in Figure 16. The ooint source sample (A-228) and
the near point source sample (A-226) have similar number of fibers but
significantly different mass concentrations, Table 1 and Figure 16. This
arises from the observation that the point source sample has relatively
larger fibers, both in diameter and length. Hence, a complete analysis
should include the number and the mass of asbestos fibers per unit
volume and the fiber size distribution.
In order to study variation in the statistical counting of asbestos
fibers from different parts of a filter, the sample A-227 was sectioned
into four parts. Two separate portions were prepared on two different
beryllium studs and statistical counting was carried out. The data so
obtained are presented in Table 1 and Figure 17 as A-227-1 and A-227-2.
It is concluded that some variations are observed between two portions
of the filter, without carrying out systematic statistical analysis.
Observations of such variations in the same filter have led Bartosiewica
to design an aerodynamic funnel system in order to obtain a more uniform
distribution of particles (air was collected with an open-ended funnel
in the present investigation).
An airborne asbestos sample was obtained from a demolition site in
Philadelphia City and representative SEM micrographs of this sample are
shown in Figure 18. Relatively large chrysotile and tremolite fibers
have been observed in this sample. In addition, the sample contained
large extraneous particles since air was collected at street level. The
particle and mass concentration for this sample are presented in Table I.
3-22
-------
F-C3779
ifl_
{NEAR POINT SOURCE)! 14,300x10
A-228 J -«)
(POINT SOURCE) \200,000XtO
MAG: x 1,000
1 1
1 1 1 1 1 1 1 1
<0.2 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
FIBER DIAMETER IN MICRONS
lOOh
A-226
•A-228
I I I I
J_
5 15 25 35 45 55 65 75 85 95 105 U5
FIBER LENGTHS IN MICRONS
Figure 16. Cumulative Distribution of Asbestos Fiber Diameters and
Lengths for Point Source and Near Point Source Samples.
3-23
-------
100 -
20
A-227-2
—o-
k-227-l
POINT SOURCE SAMPLE
A-227
IMAG: xs.ooo I
-
1 1 1 1 1 1 1
1 1 1 1
F-C3779
<0.2 0.5 1.0 1.5 2.0 2.5 3.0 3-5 40 4.5 50 5,5 6.0
FIBER DIAMETER IN MICRONS
100
80
60
•a
_)
3
5
40
20
r
A-227-2
6)
POINT SOlinct SAMPLE
A-227
xT.ooo~|
Ll 1
I
I
I
I
I
L L
J_
-------
F-C3779
'
MM
lOjLL
• * i
(a)
(b)
Figure is. SEM Micrographs of Demolition Site Sample, XI,
000.
3-25
-------
F-C3779
A few ambient air samples, collected at the FIRL, have been analyzed.
The sample collected at the street level had numerous relatively large
particles, most of which were inorganic in origin. The fraction of
organic particles was negligible since no reduction in particle concentra-
tion was observed after LT ashing.
One ambient air sample, collected at the roof of the four-storied
FIRL building, was systematically analyzed. The specimen was observed
before and after LT ashing step in order to study the effectiveness of
the LT ashing as a concentration step. SEM micrographs of this sample
are shown in Figures 19 and 20. Observation of Figure 19 indicates Chat
most of the particles are inorganic in nature. However, small particles
and fibers, which were embedded in the carbon extraction replica, Figure
20(a), are freed and appear distinctly after LT ashing, Figure 20(b). The
specimen was ashed while the oxygen plasma was directly flowing over the
particles. There is no evidence, of any loss of small particles by the
plasma stream and the particle distribution is not disturbed significantly.
A certain amount of difficulty was faced when carrying out statistical
counting of asbestos fibers in the ambient air sample. A random viewing
of the sample, following a square grid pattern as was done for point source
and near point source samples, showed no asbestos fibers. Later, the 25mm
beryllium stud specimen was scanned in a more rigorous manner, i.e., viewing
every area without leaving out any of them. Following such steps, many
chrysotile asbestos fibers were identified including some fiber clumps,
Figure 21. A statistical count could be carried out and the result is
presented in Table I. Chrysotile fiber clumps in ambient air samples
(28)
have also been observed by Holt and Young . TEM micrographs of the same
sample are shown in Figure 22. A few chrysotile fibrils were observed after
thorough viewing of the BeO substrate specimen.
A few point source and near point source airborne asbestos samples
had been obtained from the project officer. The airborne particulates
had been collected on MF type Millipore filters with a nylon backing;
3-26
-------
F-C3779
(a)
(b)
Figure 19. SEM Microaranhs of the
0.165m3 of Air; X100.
LT Ashing.
FIRL Ambient Air Sample, Roof Level,
(a) Before LT Ashing, (b) After
3-27
-------
F-C3779
(a)
(b)
Figure 20- SEM Micrographs of the FIRL Ambient Air Sample, Roof Level,
0.165m3 of Air; X3,000. (a) Before LT Ashing; (b) After
LT Ashing (Note: Even very small particles are in place
after LT ashing).
3-28
-------
F-C3779
(a)
(b)
Figure 21- SEM Micrographs of Chrysotile Asbestos Observed in One of
the FIRL Ambient Air Sample, Roof Level, 10m3 of Air;
(a) Single Asbestos Fibers; X3,000; (b) Fiber Lump, XI,000.
3-29
-------
. -' „
F-C3779
,
0.2 fi
(a)
(b)
Figure 22- TEM Micrographs of the FIRL Ambient Air Sample, Roof Level,
lOm^ of Air on 0.45p Millipore Filter, BeO Substrate.
(a) Non-asbestos Fibers, X38,000, (b) Chrvsotile Fibril,
X76,000.
3-30
-------
F-C3779
this type of filter is not well suited for preparing specimens following
the steps shown in Figure 1. Hence, certain modifications had to be
incorporated in order to handle these specimens.
A fraction of the specimen, #1000367 (Table II), was oven ashed at
450°C for four to eight hours. The resultant ash was dispersed in 250cc
of distilled water in an ultrasonic bath for 10 minutes and filtered through
a Q.22\i MF Millipore filter. Minimum time to disperse the ash ultrasonically
was used since extensive ultrasonification is commonly used to break down
chrysotile fibers into fibrils ' . It is believed that a ten minute
ultrasonification only disperses the ash without breaking down fibers into
smaller fragments; however, breaking down fiber clumps into single asbestos
fibers is not ruled out.
All such specimens, Table II, were oven ashed at 450°C followed by dis-
persion and filtration. The MF Millipore filter is then dried and processed
through the specimen preparation steps shown in Figure 1. Only a fraction of
the available filter, for specimen #1000367, was ashed and the resulting
specimen is shown in Figure 23. Chrysotile asbestos fibers and other
inorganic particles (bright areas) are seen in a dull background of the
resulting ash which arises primarily from the nylon grid. Identification
of asbestos fibers could be carried out on specimens similar to that
shown in Figures 23 and 24 and the statistical results are presented in
Table III. The ash generally gives rise to an extra sulphur x-ray fluorescence
peak.
The size of the nylon-backed Millipore filter that is ashed plays an
important role in the final specimen preparation. A larger filter gives
rise to a greater amount of ash. If the ash content is high, small fibers
and fibrils are embedded under the ash. Such was the case with specimens
#1000376 and #1000378, Table III, and the resulting final specimen was
not very conducive to statistical analysis.
Usually no asbestos fiber clumps were observed in specimens listed
in Table III. The fiber clumps are probably broken down to single fibers
by the ten-minute ultrasonification. The size distribution of asbestos
fibers observed for the specimen //1000367 is shown in Figure 25.
The significance of these results will be discussed further in
Section 4.2 as to the reliability, advantages and limitations of the
FIRL technique.
3-31
-------
F-C3779
TABLE:
Details of Samples Analyzed in Table III
Number
1000367
100027**
1000210
1000376
1000378
Collection Stte
#7, Nfcolet Settling Pond, Pennsylvania,
Date: 10/17/73
#3D Vermont, Date: 9/29/73
#8A, Vermont, Date 9/28/73
#10 Herald Construction Yard, Pennsylvania,
Date: 10/16/73
#9 Hopewood Residence
212 Chestnut, Pennsylvania,
Date: 10/16/73
3-32
-------
F-C3779
Fiqure 23. SF.M Micrograph of Near Point Source Sample, #1000367
Collected on Mi Hi pore Filter with a Nylon Backing.
Specimen Prepared after Oven-ashing, Dispersion by Mild
Ultrasonifi cation and Re-filtering through a Hi Hi pore
Filter (Note: Ash in the background), XI ,000
3-33
-------
F-C3779
(a)
(b)
Figure 24. SEM Micrographs of Near Point Source Sample, #1000274,
Particles Originally Collected on Mi Hi pore Filter with a
Nylon backing; (a) XI ,000, (b) X3.000
3-34
-------
TABLE 1 I!
Analysis of Airborne Asbestos Samples*
u>
1
u>
Ln
Sample** Volume of Magnification Number of
Ai r (m3) Asbestos
Fibers
Counted
1000367 6 X3,000 *»6
100027*1 25 XI, 000 7^
1000210 22 X3.000 68
1000376**" kk xi ,000 11
1000378 172 X3.000 9
Fiber Cone. Fiber Cone. Fiber Type
(#/m3) x 103 (gm/m3)x 10"9
2.5 3,^00 Mostly chrysoti le
123 3,000 Chrysotile; a few
croc idol ite
1900 2,100 Chrysotile
7.3 3,500 Chrysotile
13.2 15 Chrysotile
"Asbestos fibers had been collected on Millipore filters with a Nylon backing -
Such filters are not well suited for the FIRL technique. Hence, the data
should be viewed accordingly.
**For details, see Table II
---For 1000376 and 1000378, data is not reliable
o
to
-------
F-C3779
100-
I I
I I I
0.2 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
FIBER DIAMETER IN MICRONS
100
80
r o
60
i-
40
20
I I I I I III I I I I
5 15 25 35 45 55 65 75 65 95 105 115
FIBER LENGTHS IN MICRONS
Firjure 25.
Cumulative Distribution of Asbestos Fiber Diameters and
Lengths for the Samnle #1000367
3-36
-------
F-C3779
4. DISCUSSION
4.1 E.D. X-Ray Fluorescence Analysis of Standard Asbestos
Using a Field Emission SEM
The present investigation has been carried out using a JSM-50A which
is a conventional SEM with a tungsten hair-pin filament. Recently, SEM's
have been commercially available with a field emission source. An x-ray
fluorescence spectrum from a small fiber depends on beam voltage and
current density, i.e., current per unit area. The field emission source
gives a higher current density than the thermionic emission source when
the electron beam size is smaller than 1000A \ Hence, one would expect
to obtain a better asbestos x-ray spectrum using a field emission SEM than
a conventional one, particularly for asbestos fibrils which have diameters
of % 3001. A study has been carried out on amosite fibers and fibrils in
(37)
the true transmission mode of operation using a field emission SEM
It was of interest to know what improvements are obtained by using a
field emission SEM on the specimens prepared by the FIRL technique. A
Be-stud specimen and a BeO substrate specimen were prepared from U.I.C.C.
Canadian chrysotile following the FIRL specimen preparation technique,
*
Figure 1. Both the specimens were observed in a CWICSCAN field emission
SEM; the Be-stud was observed in the normal mode and the BeO substrate in
the fsemi-transmission* mode.
X-ray fluorescence spectra obtained from Canadian chrysotile fibers
on a Be-stud specimen are shown in Figure 26. A comparison between
Figures 26 and 6 leads one to conclude that the field emission SEM is
superior to conventional SEM in the identification of small asbestos
fibers. The smallest fiber that was analyzed for the Be-stud specimen
was O.lp in diameter, Figure 26; however, even smaller than O.lp, say,
0.05)j, might be identified using the field emission SEM. The x-ray
spectrum from the O.lp fiber, Figure 26, is better delineated than the
*Trade name for field emission SEM manufactured by Coates and Walter,
Sunnyvale, California
4-1
-------
t
O
O
MG
IAX, 26 SEC.
FE
MG
0.2 XA., 51 SEC.
ro
CO
2
UJ
MG
0.4XX, 19 SEC
FE
MG
ENERGY
Figure 26. X-Ray Fluorescence Spectra from UM.C.C. Canadian Chrysotile on Be-stud
Obtained in a Field Emission SEM with E.D. Analysis at 18Kv and 3 x 10-9 amo
o
CO
vo
-------
F-C3779
similar spectrum in Figure 6. Sulphur appears as an impurity peak from
the ash which is a result of LT ashing.
X-ray fluorescence spectra from the BeO suhstrate specimen are
presented in Figure 27, Extra copper peaks appear from the supporting
copper grid. A comparison of Figures 27 and 7 again leads one to conclude
that the field emission source SEM is superior to conventional SEM for
asbestos fibril identification. In fact, a chrysotile fibril of 0.03y
(300A*) was analyzed in 83 sees., Figure 27. The times spent in identifying
the asbestos fibers in Figures 26 and 27 are close to the minimum possible
for reliable spectrum generation.
Field emission SEM micrographs of the 3001 chrysotile fiber that
was analyzed, Figure 27, are shown in Figure 28. The spectrum for the
Q
300A fibril shown in Figure 27 was obtained in a 'picture-mode' since the
'reduced picture (or scan)-modef of operation was not available in the
system that was used for the present investigation. However, even better
x-ray spectra than shown in Figure 27 can be expected if 'reduced-scan-
mode' is used as was the case for conventional SEM. observation, Figures 6
and 7. Reduced-scan-mode of operation reduces the background spectrum
considerably.
In summary, the field emission SEM is superior to conventional SEM
as far as identification of asbestos fibrils are concerned. The detection
limit of chrysotile asbestos fibers by the two types of SEMs is as follows:
Specimen Conventional Field Emission
SEMSEN
Be-stud >0.1p <0.1|j
BeO-substrate £0-1*J £0
4.2 Analysis of Airborne Asbestos - Overall View
Hany electron microscopic techniques have been developed for the
(24-29)
analysis of airborne asbestos in addition to the FIRL technique.
Some laboratories prefer transmission electron microscopic techniques,
in which the asbestos fibers are identified by their SAD patterns and
some laboratories prefer scanning electron microscopic techniques, in which
the asbestos fibers are identified by their x-ray fluorescence peaks.
4-3
-------
t
I-
0.6U, 28 SEC.
O
o
LU
O.I U, 25 SEC.
FE
CU
ENERGY
o
u>
Figure 27. X-ray Fluorescence Spectra from U.I.C.C. Canadian Chrysotile on Be^ Substrate Obtained
in a Field Emission SEM with E.D. Analysis at 18Kv and 2 x 10-9 amp.
-------
F-C3779
(a)
(b)
Figure 23. Field Emission SEN Micrographs of U.I.C.C. Canadian Chrysotile
on BeO Substrate, Prepared According to Fiqure 1.
(a) X6.000, (b) 300A Chrysotile Fibril from which x-rav
spectrum, shown in Figure 27, was obtained in the "scan mode" ,
X30,000.
4-5
-------
F-C3779
A special advantage in the use of TEM for chrysotile asbestos is that
such fibers can also be identified by their tube-like unique morphology.
The SEM - E.D. x-ray analysis is faster than the TEM-SAD analysis
but each technique has advantages and disadvantages which are discussed
below specifically for airborne asbestos analysis.
SEM-E.D. X-RAY ANALYSIS
(1) The resolution of the SEM is M.OOA compared to V5A for TEM.
Hence, there is a greater chance to miss very small asbestos fibrils,
Figure 22(b), when these are situated among numerous other
particles.
(2) The difficulties involved in identifying asbestos fibers by
E.D. x-ray analysis increase with decreasing fiber size. In
conventional SEM's asbestos fibers up to V).ly in diameter can
be identified using thin substrate specimens. However, even
the smallest chrysotile fibril can be identified if a field
emission SEM is used.
TEM-SAD ANALYSIS
The major disadvantage of this technique is that identifiable SAD
patterns can only be obtained from asbestos fibers having a limited range
of diameters. Very small fibers and large fibers do not give rise to usable
SAD patterns.
Apart from the experimental difficulties involved, there are many
other factors which have to be considered in the enumeration of airborne
/ r\ / *) C \
asbestos. Many of the electron microscopic techniques *" destroy
the original asbestos fiber size distribution and give only a mass
concentration of asbestos fibers. Such techniques are probably necessary
for ambient air samples in which the concentration of asbestos fibers is
rather low. Techniques employing steps to preserve the original fiber
(28)
size distribution , including the present technique (Section 3.4),
have not proved very successful for ambient air samples.
4-6
-------
F-C3779
However, reliable statistical counting of asbestos fibers can be
performed on point source and near point source samples, while preserving
the original fiber size distribution, Section 3.4. The present technique
has proved very successful in this regard except for one problem which
needs further discussion. Point source and near point source samples
very often contain fiber clumps, Figures ll(a), I4(a) and (b), and 15.
It is difficult to specify the size of a clump sdimply and to estimate the
number of fibers in a clump. It therefore seems desirable to report a clump
of asbestos fibers as a clump but estimates of mass concentration will be
inaccurate if a sample has a significant number of clumps.
One way to overcome the problem arising from fiber clumps is to
specify:
3 3
(1) fiber concentration/m and clump concentration/in
(2) fiber size distribution (free fibers only)
without an estimate of mass concentration. It is unlikely that fiber
clumps can be inhaled and can reach the lungs. A different approach should
be taken if automated counting is to be developed - this is discussed in
Section 4.3.
4.3 Feasibility of an Automated Counting System
The present technique has been developed for a possible automated
counting system for airborne asbestos. The specimen preparation techniques
are such that asbestos fibers are observed on a featureless background
(39-41)
which is essential for image analysis " . The feasibility of an
automated system consisting of an SEM, E.D. x-ray analyzer and an IMANCO
(42)
image analyzer (Quantimet 720 system) has been established even though
many peripheral problems remain to be investigated.
The present investigation included analyzing point or near point
source airborne asbestos samples. Reliable statistical data have been
obtained for such samples, Section 3.4. A previously unexpected problem
has arisen as far as automated image analysis is concerned and this needs
further discussion.
4-7
-------
F-C3779
Automated image analysis will be unreliable for samples containing
fiber clumps, Figures 11(a), 14, 15 and 21 , and fibers in close proximity
with other particles or fibers clumped with other particles, Figures 11(a),
14(b), 15(a) and 21(b). Hence, if automated image analysis is being contem-
plated, then some specimen preparation steps should be investigated to free
the asbestos from either asbestos fiber clumps or heterogeneous clumps
without disturbing the original fiber size distribution (or, disturbing the
size distribution to a minimum).
FUTURE WORK
The following steps should be investigated for freeing the asbestos
fibers from clumps if automated image analysis is to be developed.
(1) The Millipore filter with collected airborne asbestos is
ashed in a LT asher. The LT ashing is preferable to oven-ashing,
(2) The ash is dispersed in filtered distilled water in an ultrasonic
bath for 10 mins. Experience has shown that a 10-minute ultra-
sonification is only necessary to disperse the particles uniformly.
It is expected that fibrillation is minimal during the ten-minute
*
ultrasonification.
(3) Refilter the ash through a MF-type Millipore filter and then
follow the steps shown in Figure 1,
*Ultrasonification is carried out for 4-12 hours to break all chrysotile
fibers into fibrils(24,25)
4-8
-------
F-C3779
5. CONCLUSIONS
The following conclusions have been drawn from the present investi-
gation.
1. Further developments have been carried out in the specimen
preparation technique that was developed earlier under an
EPA research contract. /. J
2m Specimens for the transmission electron microscope (TEM) are
prepared simultaneously along with the specimens for the
scanning electron microscope (SEM).
3. Both types of specimens exhibit featureless background for
convenient observation of asbestos fibers and fibrils, and
are suitable for automated image analysis.
4. The ^25mm diameter Be-stud specimen is preferred for a truer
statistical analysis in the SEM than the 3mm copper grid
specimen meant for use in the TEM.
5. The specimen preparation technique preserves the
fiber size distribution with minimal loss of large particulates.
6. Asbestos fibers are identified in an automated manner by the
programmed NS-880 E.D. x-ray analyzer interfaced with the SEM.
Hence, less skilled operators are required for carrying out a
statistical analysis.
7. The smallest chrysotile asbestos fiber that is confidently
identified, using the Be-stud specimen, is ^0.15)J in diameter
while <, O.lp asbestos fibers can be identified using the BeO
substrate specimen. The above limit of detection refers to a
conventional SEM.
8. The field emission SEM is superior to conventional SEM as far
as identification of small asbestos fibers are concerned. It is
possible to identify the smallest chrysotile fibril (^300A) in
the field emission SEM using the BeO substrate specimen.
9. Observation of small asbestos fibers in the presence of a large
number of other particulates is made easier by the use of TEM
in comparison to SEM.
10. Identification of asbestos fibers needs to be carried out in the
TEM if a change in the chemical composition of the fibers has
taken place. This is applicable more to waterborne asbestos and
to asbestos in foods, drugs and tissues than to airborne asbestos,
A chemical change for the airborne asbestos is very unlikely,
5-1
-------
F-C3779
11. The present technique is applicable to point source and near point
source airborne asbestos samples satisfactorily and less satisfactorily
to ambient air samples.
12. Asbestos fiber clumps have been observed in point source, near point
source and even in ambient air samples. In a manual mode of operation,
this problem can be dealt with. However, problems are foreseen in a
completely automated counting system. The automated image analyzing
system may not be able to identify and count an asbestos fiber clump
as a 'clump'.
13. It is recognized that a step involving mild ultrasonification is
necessary to free asbestos fibers from adhering non-asbestos participates
and possibly to break down a fiber clump into asbestos fibers (without
fibrillation). This step will be necessary if automated image analysis
is being contemplated.
5-2
-------
F-C3779
6. ACKNOWLEDGEMENTS
The authors gratefully acknowledge Mr. Stanley Luszcz's effort in
developing a computer program for automatic identification of asbestos
fibers, and in the preparation of Beryllium oxide films. Valuable efforts
rendered by Miss L. Marchant, Mr. Louis Cinquina and Mr. Jerome Liss are
also acknowledged,
Acknowledgements are also due to Dr. R. J. Vadimsky of Bell Telephone
Research Laboratories, Murray Hill, Li.J. lui: conducting the work related to
the Field Emission Scanning Electron Microscope and to Dr. J. Wagman of
the Environmental Protection Agency, Research Triangle Park, for his
continued interest and encouragement.
Submitted by:
Amitav Pattnaik
Senior Research Metallurgist
Approved by:
K. E. Dorschu
Manager
Metals and Ceramics Laboratory
6-1
-------
F-C3779
7. REFERENCES
1. Whipple, H. E., Biological Effects of Asbestos; Annals of the New
York Acad. of Sciences, 132, 1-766 (1965)
2. Tabershaw, I. R., Asbestos as an Environmental Hazard, J. Occup.Med.,
16_, 32 (1968)
3. Langer, A. M., Selikoff, I. J., and Sastre, A., Chrysotile Asbestos
in the Lungs of Persons in the New York City; Arch. Environ. Health,
22, 348 (1971)
4. Langer, A. M. et alt Identification of Asbestos in Human Tissues;
J. Occup. Med., March 1973
5. Gilson, J. C., Health Hazards of Asbestos; Composites, ^, 59 (1972
6. Sherrill, R., Asbestos, the Saver of Lives has a Deadly Side, New
York Times Magazine, Jan. 21, p. 12, (1973)
7. Asbestos - The Need for and Feasibility of Air Pollution. Controls:
Committee on Biological Effects of Atmospheric Pollutants, Publ. by
National Academy of Sciences, (1971)
8. Langer, A. M., Rubin, I. B. and Selifoff, I. J., Chemical Characteri-
zation of Asbestos Body by Electron Microprobe Analysis; J. Histochem.
and Cytochem., 20, 723 (1972)
9. Langer, A. M. et al, Chemical Characterization of Uncoated Asbestos
Fibers from the Lungs of Asbestos Workers by Electron l-ticroprobe
Analysis; ibid, 2J>, 735 (1972)
10. Selikoff, I. J. et al, Asbestosis and Neoplasia; Amer. J. Med., j42_,
487 (1967)
11. Enticknap, I. B. and Smither, W. J., Peritorial Tumors in Asbestosis;
Brit. J. Indust. Med., _21, 20 (1964)
12. Selikoff, I. J., Hammond, E. C., and Chung, J., Asbestos Exposure,
Smoking and Neoplasia> JAAIA, 204, 20 (1968)
13. Harris, P. G., Asbestos Hazards in Naval Shipyards; Ann. Occup.
11, 135 (1968)
14. McEwen, J. et al, Mesothelioma in Scotland, Brit. Med. Jour., 4^ 575
(1970)
15. Newhouse, M. and Thomson, H., Mesotheliuma of Pleura and Peritoneum
Following Exposure to Asbestos in the London Area; Brit. J. Indust.
Med., 22, 261 (1965)
7-1
-------
F-C3779
16. Lieben, I. and Plstawka, H., Mesothelioma and Asbestos Exposure;
Arch. Environ. Health, 14, 599 (1967)
17. Asbestos Health Quest-ion Perplexes Experts: Chemical Engine. News., 51.
#50, 18 (1973)
18. Wagner, C.; Disputes on the Safety of Asbestos; New Scientist, 61,
#888, 606 (1974) —
19. Stanton, M. F. and Wrench, C., Mechanism of Mesothelioma Induction
with Asbestos and Fiber Glass; J. Nat. Cancer Inst., ^8, 797 (1972)
20. Timbrell, V., Inhalation and Biological Effects of Asbestos* in
"Assessment of Airborne Particles", Ed. Mercer et al*Charles C.
Thomas Publ., U.S.A., p. 429 (1972)
21. Vorwald, A. J., Durkan, T. M. and Pratt, P. C., Experimental Studies
of Asbestosis; A.M.A. Arch. Ind. Hug. Occup. Med., _3, 1 (1951)
22. Richards, A. L., Estimation of Trace Amounts of Chrysotile Asbestos
by X-ray Diffraction; Anal. Chem., 44, 1872 (1972)
23. Wagman, J., Environmental Protection Agency, Private Communication
24. Henry, W. H. et al.^ Development of a Rapid Survey Method of Sampling
and Analysis of Asbestos in Ambient Air* Final Report, Battelle
Columbus Labs., Contract No. CPA-69-110, Feb. 1972
25. Richards, A. L., Estimation of Submicron Quantities of Chrysotile
Asbestos by Electron Microscopy; Anal. Chem., 45, 809 (1973)
26. Staff, Collodion Film Method for the Determination of Asbestos in
Ambient Atmosphere; Air and Industrial Hygiene Laboratory, California
State Department of Public Health, private communication (1972)
27. Selikoff, J. J., Nicholson, W. J., and Langer, A. M., Asbestos Air
Pollution; Arch. Environ. Health, _25_, 1 (1972)
28. Holt, P. F. and Young, D. K., Asbestos Fibers in the Air of Towns,
Atmosph. Environ., 2» 481 (1973)
29. Chatfield, E. J., Quantitative Analysis of Asbestos Minerals in Air
and Water, 32nd. Ann. Proc. Electron Microscopy Soc. Amer., St. Louis,
Missouri, Ed. Arceneaux, C. J., p. 528 (1974)
30. Skikne, M. I., Talbot, J. H., and Rendall, R. E. G., Electron Diffraction
Patterns of U.I.C. Asbestos Samples* Environ, Res., -_4, 14}. (1971)
7-2
-------
F-C3779
31. Data Sheet of Physical and Chemical Properties of U.I.C.C. Standard
Reference Samples, N.R.I.O.D., Johannesburg, South Africa, (1972)
32. Timbrell, V., Characterization of the U.I.C.C. Standard Reference
Samples of Asbestos; in Pneumoconiosis, Proc. Int. Conf., Johannesburg,
Ed. H. A. Shapiro, Oxford Univ. Press, p. 28 (1970)
33. Clark, R. L. and Rudd, C. 0., Transmission Electron Microscopy Stand-
ards for Asbestos, Micron, _5, #1, 83 (1974)
34. Ferrell, R. E., Paulson, G. G., and Walker, C. W., Pollutant Identi-
fication by Selected Area Electron Diffraction, I-Method, II - The
Limitations, ref. 29, p. 532, p. 534 (1974)
35. Yakowitz, H., Jacobs, M. H., and Honneyball, P. D., Analysis of Urban
Particulate by Means of Combined Electron Microscopy and X-ray Micro-
analysis; Micron, _3» 498 (1972)
36. McCrone, W. C. and Stewart, I. M., Asbestos, American Laboratory, j^,
//4, 13 (1974)
37. Maggiore, E. J. and Rubin, I. B., Optimization of a SEM X-ray Spectro-
meter System for the Identification and Characterisation of Ultra-
microscopic Particles, IITRI Conf. on SEM, 1973, Ed. Johari Ora and
Corvin, I., p. 129 (1973)
38. Kramer, J. R., Mudroh, 0., and Tihor, S., Asbestos in the Environment,
Department of Geology, McMaster University, Ontario, Canada, Report
submitted to Research Advisory Board, International Joint Commission
and Environment Canada, June (1974)
39. Gibbard, D. W., Smith, D. J., and Wells, A., Area Sizing and Pattern
Recognition on the Quantimet 720; The Microscope, 2jD, 39 (1972)
40. Fisher, C., The New Quantimet 720; The Microscope, 19, 1 (1971)
41. Jesse, A., Quantitative Image Analysis in Microscopy - A Review;
19., 21 (1971)
42. Pattnaik, A. and Meakin, J. D., Development of an Instrumental
Monitoring; Method for Measurement of Asbestos Concentrations in,
or Near, Sources, Final Report, Prepared for U.S. Environmental
Protection Agency, EPA-650/2-73-016, June 1973
43. Speil, S. and Leineweber, J. A., Asbestos Minerals in Modern Industry;
Environ. Res., 2_, 166 (1969)
44. Corn, M., Statistical Reliability of Particle Size Distributions
Determined by Microscope Techniques; Amer. Indust. Hug. Assoc. J.,
2£, 8 (1965)
45. Langer, A. M. and Pooley, F. D., Identification of Single Asbestos
Fibers in Human Tissues, Proc. Intl. Agency for Res. on Cancer;
Biol. Effects of Asbestos, Lyon, C. Wagner, Ed., Paper 19
7-3
-------
F-C3779
46. Rahman, 0. Viswanathan, P. N., and Tandon, S. K., Influence of
Citrate Ions on the Dissolution of Silica from Asbestos^ Med.
Lavoro., 64, 245 (1973)
47. Russ, J. C., X-ray Spectroscopy on the Electron Microscope,
X-ray Spectrometry, 2.9 11 (1973)
48. Bartosiewicz, L. , Improved Techniques of Identification and Determina-
tion of Airborne Asbestos, Amer. Indust. Hygiene Assoc. J., 34, #6,
252 (1973)
49. Swann, D. J. and Kynaston, D,, The Development of a Field Emission
Scanning Electron Microscope, SEM/1973, IITRI Conf., p. 57 (1973)
50. Holmes, S., The Measurement of Asbestos Dust, Staub-Relnhalt Luft,
^3, 64 (1973)
7-4
-------
F-C3779
Appendix
APPENDIX A
BeO Suhstrate Preparation
FRANKLIN INSTITUTE RESEARCH LABORATORIES
» > M j " M i H r R * n K i : N p i ri * 4V > Y • p H i. « r i > r M i A ^ •: «j N r. • . \ A \ . A i « < o i
-------
F-C3779
APPENDIX A
BeO Substrate Preparation
The optimum substrate for selected area diffraction and microscopy
would be amorphous and featureless. For x-ray fluorescence analysis the
substrate should contribute no characteristic peaks in the energy range
of interest. Moreover, the substrate should be stable under low temperature
oxygen plasma ashing which is a critical step in the preparation of samples.
Initially S10 substrates were prepared and were successfully used in
the specimen preparation. The only disadvantage of SiO substrate is that
Si from the substrate hinders identification of asbestos fibers which are
uniquely identified by their Mg, Si, Ca and Fe x-ray fluorescence peaks.
BeO is an ideal candidate since neither Be nor 0 is detected by E. D.
x-ray analysis. BeO substrates have been successfully prepared in this
laboratory after some initial set-backs. BeO is brittle and initially the
substrates exhibited cracking as the copper grid curled during oxidation
in the LT asher. The normal TEM copper grids have a ring at the periphery
and this seemed to cause the distortion. The successful technique uses
grids punched out of a sheet of copper mesh.
SUBSTRATE PREPARATION STEPS:
1. 3mm diameter copper grids are punched out of 300 mesh sheet.
2. The blank grids are preoxidized overnight at 175°C in air.
3. The copper grids are ultrasonically cleaned in acetone to
remove non-adhering oxide.
4. A cleaned microscope slide is coated with a thin layer of
dehydrated Victawet*.
5. A weighed amount of Be is evaporated onto the glass slide.
The Be vapor oxidizes to BeO in the evaporator and deposits
as BeO film.
6. The glass slide is scratched with a razor blade to form a
square grid pattern on the BeO film.
*Victawet - Available from Ladd Research Industries, Vermont, U.S.A.
A-l
-------
F-C3779
7. The BeO is floated onto the surface of deionized water
containing a few drops of detergent. Small squares of BeO
film are picked up on preoxidized copper grids and dried.
BeO substrates prepared in the above manner survive LT ashing and
the stresses that are encountered during different steps of the specimen
preparation, Figure 1.
The properties of the BeO substrates have already been described,
Section 3.3. It is a crystalline film but is featureless for present purposes,
However, it is believed that the quality of the BeO substrate could
be improved by depositing Be on a colder substrate. Amorphous BeO substrates
should be possible. The present BeO substrates were adequate for this
investigation.
TOXICITY OF BeO
It is well known that beryllium (Be) and its vapor are toxic. In the
solid state, Be studs can be handled without much danger from its toxicity;
however, care should be taken when Be vapor is encountered, which is the
case In the preparation of BeO substrate. The following precautions are
taken at the FIRL whenever BeO substrates are prepared.
(1) While evaporating Be in an evaporator, a somewhat temporary
enclosure, made of thick aluminum foil, is made around the
tungsten filament and the glass substrate. The specially tailored
enclosure still allows viewing of the filament for controlled
evaporation.
(2) A glass beaker of suitable size is put upside down above the
aluminum foil enclosure. Hence escape of Be vapor to the
outer chamber is minimal.
(3) After completion of evaporation, the aluminum foil is discarded
safely and the glass beaker is thoroughly washed under an
exhaust hood.
A-2
-------
F-C3779
(4) The evaporator is also cleaned thoroughly using acetone and
tissue papers after evaporation. Then the tissue papers are
discarded safely.
The FIRL has had extensive experience in the preparation and handling
of beryllium. Routine analysis of beryllium in the air of the beryllium
laboratories of the FIRL is carried out according to environmental standards.
A-3
-------
F-C3779
Appendix
APPENDIX B
Computer Program for the Identification
of Asbestos Fibers
THE FRANKLIN INSTITUTE RESEARCH LABORATORIES
1 >• f Mf'JJAWIN fRANKLIN PARivA'AY
f>Nll»nFlP«l» =CNNSY1VAN.A I <) I (I "•
-------
F-C3779
APPENDIX B
Computer Program for the Identification
of Asbestos Fibers
The JSM-50A scanning electron microscope at the FIRL is equipped
with a computerized NS-880 E.D. x-ray analyzer. The associated mini-
computer is programmed for various applications and x-ray fluorescence
spectra can be stored on magnetic tapes. Recently, a system called
*
Flextran , has been developed for the NS-880 system; with Flextran, the
computer can be programmed using Flextran language. New programs
can be readily written by the user for specific applications. The
following program has been developed for the identification of asbestos
fibers.
Common asbestos fibers can be identified by the three-element
ratio analysis * . The elements of interest are Mg, Si and Fe ai
their ratios are calculated according to the following formula:
..
(B-l)
I V'l'
where, i refers to Mg, Si, and Fe
A : Element ratio
I. : Intensity of x-ray fluorescence peak from element
i in an asbestos fiber
I : Intensity of x-ray fluorescence peak from element i from a
standard sample consisting of pure element i.
The A ratios are plotted on a ternary-type diagram, Figure B-l.
The A. ratios are not unique for a particular type of asbestos and the
spread observed depending on fiber size and composition for the U.I.C.C.
asbestos standards is shown in Figure Bl. This diagram can be used for
identifying unknown asbestos fibers. Similar approach was used by
fo\
Langer et al. while using an electron microprobe with a wavelength
*Tracor-Northern, Inc., Middletown, Wisconsin 53562
B-l
-------
F-C3779
dispersive spectrometer. Their ternary-type diagram (Figure 5 of ref. 8)
is similar but not identical to Figure B-l since the ratio A depends on
many factors, particularly the sensitivity of the detection system.
Presently the FIRL is working on a FDA contract to identify asbestos
fibers in talc, food and drugs. Experience has led to the development
of a new program which is based on a five-element ratio using the following
equations:
(B-2)
I V'i'
where 1 refers to Na, Mg, Si, Ca and Fe. Expanding Equation B-2 for Mg
gives :
A
"
I /I S + I /L S + I /I S + I /I S + I
Na7 Na M'^l ^i' Si Ca7 Ca F
(B-3
etc.
It was observed earlier, Figure B-l, that these A numbers are not
unique and they have a spread corresponding to different regions defining
specific type of asbestos. Hence, A + AA (AA referring to the spread in
numbers) values for different types of asbestos and talc are stored in the
NS-880 computer memory. Similar A numbers are determined from an unknown
spectrum and matched with the numbers for standards in the following manner.
The following numbers are calculated:
~ Mg known ^lg unknown Mg
(B-4)
(ASi * ^Si5 known ~ (ASi}unknown " ' BSi
etc.,
1 i
B : Absolute value of + B
B-2
-------
V
F-C3779
-CHRYSOTILE
AMOSITE CROCIDOLITE
Figure B-l. Ternary Diagram of flg-Si-Fe Ka Emission Ratios of
Standard U.I.C.C. Asbestos
B-3
-------
F-C3779
where, i refers to chrysotile, amosite and talc etc.
The C.'s are calculated for chrysotile, amosite etc., and the
unknown fiber is identified to be that standard fiber which gives the
least value for C.
Before the computer calculates the C values using Equation B-4, the
fluorescence spectrum from the unknown fiber is stripped of the background
spectrum then each peak is normalized with respect to standards.
A pseudo-ternary diagram illustrating the regions for different
asbestos fibers is shown in Figure B-2. The 5-element space is shown
projected onto the 3-element plane. The amount of sodium (Na) differentiates
between the overlapping regions of amosite and crocidolite and hence there
is no ambiguity.
Normalized A ratios with limits for different types of asbestos are
presented in Table B-l. Standard asbestos fibers have been used to obtain
the A ratios presented in Table B-l; the limits have been obtained carrying
out x-ray analysis on a large number of fibers with different diameters
and locations. A close scrutiny of the Table B-l indicates that none of
the asbestos variety has all the 5 ratios similar allowing unique identi-
fication. The uniqueness of identification is evident from Table B-l and
illustrated in Figure B-2 recalling that 5-element space is projected on the
3-element surface.
A schematic diagram of the computer program and logic is shown in
Figure B-3. The program can be operated following two different routes:
(1) The fluorescent x-rays are accumulated as long as the operator
decides and then, the computer is asked to identify the type of
asbestos spectrum.
(2) The computer accumulates data for every 10 sees, and checks if
the peaks are significant with respect to background and then
tries to match the spectrum in order to identify.
B-4
-------
F-C3779
Fe
or
Co
-TALC
Si
Figure B-2.
Pseudo-Ternary Diagram of Na-Mq-Si-Ca-Fe Ka Emission
Ratios of IM.C.C. Standard Asbestos, Trenolite, and Talc.
B-5
-------
F-C3779
TABLE B-l
NORMALIZED A-RATI OS FOR DIFFERENT ASBESTOS FIBERS
Limits of Peak Ratios
Amos. CrocId, Chrys. Anth. Trem. Talc
Na <1 7 - 22 <1 <1 <1
-------
F-C3779
FOR STATISTICAL
Fiqure B-3. Schematic Diagram of the Computer Program for the Identi-
fication of Asbestos
-------
F-C3779
Route #2 takes the minimum time required to identify the asbestos
fiber. During the execution, of the program following either route #1
or route #2, the spectrum is displayed on the oscilloscope screen. The
operator stops the automatic counting process if the spectrum contains
elements that certainly do not belong to any asbestos type.
Examples of fiber identification by the route //I is shown in
Figure B-4(a) and by the route #2 in Figure B-4(b). The analysis was
carried out on an asbestos fiber of 0.6y in diameter. The spectrum was
accumulated for 100 sees, in Figure B-4(a) whereas the spectrum in
Figure B-4(b) was accumulated only for 10 sees.
B-8
-------
F-C3779
bhlb
488KEV,
10. 22E
(a)
r i ptK
BLflNK K£V TO RESTfiRT
*
PHR12 RDD . 820KEV/CH
-<-"= g, 480KEV,
(b)
Figure B-4. Oscilloscope Photographs of Identified Chrysotile Spectra
(a) Data collected for 100 sec. and then identified.
(b) Minimum time required (10 sees.) for the identification
of the same fiber. (Total counts are different for (a)
and (b)).
B-9
-------
./
EPA 650/2-75-029
TECHNICAL REPORT DATA
••ast r>:ad iHUrucrions on die retcrse before completing)
2. 3. RECI
Development of Scanning Electron Microscopy for Jam.
7. AUTHORIS)
Amitav Pattnaik and John D.
Meakin
The Franklin Institute Research Laboratories 1AAC
Benjamin Franklin Parkway
Phi ladelphia , Pennsylvania
11. COM
19103
68-C
12. SPONSORING AGENCY NAME AND ADDRESS 13. TYP
Chemistry and Physics Laboratory Fin
National Environmental Research Center, EPA '*• SPO
Research Triangle Park, N.C. 27711
IE NT'S ACCESSION NO.
RT DATE
ary 1975, date of prep.
10 (26AAN)
2-1268
il. November 1973-Dec J.974
The methodology that was developed at the Franklin Institute Research Laboratories
(FIRL) under the EPA Contract No. 68-02-0544, for the determination of airborne
asbestos has been perfected and developed further. Moreover, the newly perfected
technique has been applied to point source, near point source and ambient air samples.
This report describes the analytical method which employs a scanning electron
microscope equipped with energy dispersive x-ray analysis for the identification
and counting of airborne asbestos. The specimens, prepared in a unique manner, are
suitable for image analysis and for a possible automated counting system.
Results of the analysis on airborne asbestos are presented, and limitations and
advantages of the present techniques are discussed.
a. DESCRIPTORS
b.lDENTlFiens/OPEN EIMOE
Asbestos
Chrysotile
Scanning electron microscopy
X-ray fluorescence
Release unlimited
19. SECURITY CLASS (This
unclassified
unclassified
O TERMS c. COSATI Field/Group
Report/ 31. NO. Of PACES'
84
ragt) 22. PRICE
-------
1. REPORT NUMBER
Inserl the EPA report number as it appears on the cover of the publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
4. TITLE AND SUBTITLE
Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, if used, in smaller
type or otherwise subordinate it to main title. When a report is prepared in more than one volume, repeat the primary title add volume
number and include subtitle for the specific title.
5. REPORT DATE
Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g.. date of issue, date of
approval, date of preparation, etc.}.
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOR(S)
Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.). List author's affiliation if it differs from the performing organi-
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hirearchy.
10. PROGRAM ELEMENT NUMBER
Use the program element number under which the report was prepared. Subordinate numbers may be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant number under which report was prepared.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGENCY CODE
Leave blank.
15. SUPPLEMENTARY NOTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation with. Translation of, Presented at conference of,
To be published in, Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (200 words or less) factual summary of the most significant information contained in the report. If the report contains a
significant bibliography or literature survey, mention it here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
concept of the research and are sufficiently specific and precise to be used as index entries for cataloging.
(b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc. Use open-
ended terms written in descriptor form for those subjects for which no descriptor exists.
(c) COSATI FIELD GROUP - Field and group assignments are to be taken from the 1965 COSAT1 Subject Category List. Since the ma-
jority of documents are multidisciplinary in nature, the Primary Field/Group assignment(s) will be specific discipline, area of human
endeavor, or type of physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
the primary posting(s).
18. DISTRIBUTION STATEMENT
Denote releasability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
the public, with address and price.
19. & 20. SECURITY CLASSIFICATION
DO NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any.
22. PRICE
Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Form 2220-1 (9-73) (Reverse)
------- |