HATFIELD
TECHNICAL
CONSULTING
, LIMITED
2071 Dickson Road
Mississauga. Ontario
CANADA L5B1Y8
Telephone: (905)896-7611
Fax: (905)896-1930
DRAFT
Analytical Method
for Determination of Asbestos in
Vermiculite and Vermiculite-Containing Products
Prepared for:
Mr. Wayne R. Toland
U.S. Environmental Protection Agency
1 Congress Street. Suite 1100
Boston, MA 02114
Dr. Eric J. Chatfield
President
Chatfield Technical Consulting Limited
DISCLAIMER
This analytical method has not been submitted for formal peer
review nor has it been tested in a competent lab to determine its
accuracy and precision. As a result, EPA does not endorse,
recommend, or otherwise encourage its use in any way until that
process is completed. This method, which is presented here in
draft form only, is being provided at your request and any
decision to use it for analytical purposes rests entirely at the
readers/users discretion.
-------
TABLE OF CONTENTS
INTRODUCTION AND BACKGROUND 1
1.1 General 1
1.2 Required Characteristics for an Analytical Method for Determination of
Asbestos in Vermiculite 1
1.3 Analytical Considerations Specific toVernilculite from Libby, Montana
2
1.4 Analytical Considerations for Vermiculite Sources Other Than Libby
3
2 PRINCIPLE OF METHOD
2.1 Background
2.2 Types of Measurement
3 SCOPE AND FIELD OF APPLICATION
3.1 Substance determined
3.1.1 Weight Percent Asbestos
3.1.2 Concentration of Respirable Fibers
3.2 Type of Sample
3.3 Range
3.4 Limit of Detection
4 DEFINITIONS
5 ABBREVIATIONS
6 EQUIPMENT AND APPARATUS
6.1 General
6.2 Samplepreparation
6.3 Rapid Screening Method
6.4 Measurement of Respirable Fibers by TEM
3
3
4
5
5
5
6
. 6
6
6
8
12
12
12
. 13
13
13
7 REAGENTS 14
8 SELECTION AND PRE-TREATMENT OF SUB-SAMPLE FOR ANALYSIS
8.1 TypesofSample
8.2 Obtaining a Representative Sub-Sample for Analysis
8.3 Pre-Treatment of Sub-Samples
8.3.1 Background
8.3.2 Exfoliated Vermiculite
8.3.3PottingSoil
. .
14
• 14
• 15
.16
16
.16
.16
-------
8.3.4 Lawn Fertilizer . 18
8.3.5 Beneficiated Crude Vermiculite Ore 18
8.3.6 Soil Samples Containing Crude Vermiculite Ore 21
8.3.7 Soi’ Samples Containing Exfoliated Vermiculite 21
8.3.8 Samples of insulation incorporating materials in addition to
vermiculite 21
9 PROCEDURE FOR ANALYSIS 23
9.1 Introduction 23
9.2 Rapid Screening Analysis to Determine the Weight Percent of
Amphibole Asbestos 24
9.2.1 General 24
9.2.2 Separation of Vermiculite from other Components by Flotation
on Water 24
9.2.3 Optional Preparation of TEM Specimens From the Aqueous
Suspension of Vermiculite 25
9.2.4 Density Adjustmentof HeavyLiquid . 25
9.2.5 Separation of Amphibole Fragments by Centrifugation in a
HeavyLiquid 30
9.2.6 Stereo-Binocular Microscope Examination of the Centrifugate
37
9.3.1 Introduction 41
9.3.2 Separation of Coarse Vermiculite 42
9.3.3 Separation of Respirable Fibers by Displacement Sedimentation
42
9.3.4 Preparation of TEM Specimens From Displaced Suspension
43
9.3.5 Examination of TEM Specimens 49
9.4 Combined Procedure 49
10 DATA REPORTING 49
10.1 Rapid Screening Analysis to Determine Weight Percent of Amphibole
Asbestos
10.2 Concentration of Respirable Fibers 50
11 ACCURACY AND PRECISION 51
.12 -REFERENCES •.• •. . 51
APPENDIX A. RECOVERY OF HEAVY UQIJIDS FOR RE-USE . 53
A 1 Background .- 53
A.2 Purification of Heavy Liquids by Distillation 53
-------
1 INTRODUCTION AND BACKGROUND
1.1 General
As is the case for most minerals, deposits of vermiculite contain other
mineral phases, and most of these other mineral phases are removed during
processing. This process by which vermiculite is concentrated from the crude ore
is referred to as beneficiatiori. During beneficiation of crude vermiculite ore, the
vermiculite is also segregated into different size fractions for different applications.
Larger sizes of vermiculite flakes command a higher price.
Most deposits of vermiculite that are the sources of current production
contain low concentrations of amphibole minerals, and some of these fragments of
amphibole may have chemical compositions that are within the ranges of the
regulated amphibole asbestos species. However, these amphibole fragments are
usually present in a non-asbestiform crystal habit. Rarely, a small proportion of the
amphibole may be present in a fibrous crystalline habit, conforming to the
conventional definition of asbestos. During beneficiation, much of any
non-asbestiform amphibole or amphibole asbestos present. in the original ore is
separated, but some of the material passes through the benefuciation process and
appears in the beneficiated vermiculite.
Chrysotile has not been detected in any of the vermiculite marketed in North
America. Moreover, chrysotile would be unlikely to survive the exfoliation process,
since the temperatures used for exfoliation exceed those at which chrysotile
degrades. Some vermiculite, particularly that from the Republic of South Africa,
contains scrolls of vermiculite which are short and exhibit morphological features
similar to those of chrysothe. However, these scrolls can be discriminated from
chrysotileby transmission electron microscopy (Chatfield and Lewis, 1980)
1.2 Required Characteristics for an Analytical Method for Determination of
Asbestos in Vermiculite
Vermiculite is normally used as purchased. and there are no uses for which
it is necessary to grind or pulverize vermiculite to a powder. Accordingly, any risks
presented by use of vermiculite are those presented by the material as used, and
•the analytical method should measure the required parameter without crushing of
the vermiculite. If vermiculite is pulverized or crushed prior to analysis, the results
of the analysis will probably mis-represent any risks associated with its use. In
particular, crushing of any associatednon-asbestiform amphibole fragments will
generate numerous cleavage fragments which will complicate the interpretation of
1
-------
the results. This is because these cleavage fragments were not present as such in
the material as used, and once they are generated the analytical methods cannot
always discriminate between these cleavage fragments and fibers of true asbestos.
Assuming that amphibole fragments are present in the beneficiated
vermiculite, the amount of amphibole present in the final exfoliated product
depends on the practices of the exfoliation facility. During exfoliation, the
vermiculite expands to 5 - 15 times its original volume, and these very light
fragments are separated by air entrainment. The other minerals present in the
original beneficiated vermiculite are not useful, and represent material (usually
referred to as “rock”) that must be disposed of by the exfoliation fécility. Some
facilities return the “rock” to the vermiculite after the exfoliation process, and it is
therefore incorporated into the final product. Other facilities dispose of the “rock”
as a waste material. The importance of this to the analyst is that non-vermiculite
fragments may be common in some samples but relatively rare in others.
Transmission electron microscopy (TEM) is flQL an appropriate method for
determination of the weight percent amphibole asbestos in vermiculite, because
the size range of fiber bundles of amphibole asbestos that may be present in
vermiculite extends up to approximately the dimensions of the vermiculite flakes,
and the majority of the weight of amphibole asbestos is represented by these
larger fiber bundles that are very much larger than can be examined by TEM. Any
attempt to measure the weight concentration by TEM will usually yield a value that
significantly under- estimates the actual concentration.
TEM is an appropriate method for determination of the numerical
concentration of respirable fibers, and is, in fact, the only method by which an
accurate determination can be made. An analytical method, therefore, must
incorporate a procedure by which respirable fibers can be separated from the bulk
material, without generating additional respirable fibers by crushing or grinding of,
the material.
It is most important to recognize that reliable and reproducible results cannot
be obtained by analysis of small samples. Any amphibole particles present in
vermiculite are usually much fewer in number than the flakes of vermiculite, and if
only a small sample size is analyzed the number of amphibole particles included in
the sample will be small and often unrepresentative.
I - Analytical Considerations Specific to Vermiculite from Ubby. Montana
Prior to 1990, a large proportion of the U.S. consumption of vermiculite
2
-------
originated from the mine at Libby, Montana. The vermiculite originating from this
mine was quite unique in that substantial concentrations of amphibole asbestos
were present, and in this respect it was unlike vermiculite from any other source.
During the life of the Libby mine, attempts were made to reduce the level of
amphibole asbestos in the beneficiated vermiculite, but it was never possible
remove it completely. Depending on the date of production, beneficiated
vermiculite from Libby may have contained several percent of amphibole asbestos,
down to a fraction of a percent shortly before the mine was closed in 1990.
From an analytical perspective, it is important to recognize that, with
relatively simple, but appropriate, analytical procedures specified in this method,
the amphibole asbestos in vermiculite from the Libby mine can be readily
recognized and the weight percent of amphibole asbestos reliably determined in the
range of less than approximately 0.01 % to several percent by weight. This
measurement can be made using conventional chemical laboratory equipment, a
stereo-binocular microscope and a polarized light microscope. Samples of products
containing vermiculite from the Libby mine will generally yield sufficient amphibole
asbestos to determine the approximate weight concentration by weighing.
1.4 Analytical Considerations for Vermiculite Sources Other Than Ubby
Analysis of vermiculite from current production is generally a matter of
establishing a sufficiently low limit of detection, and distinguishing between
asbestiform and non- asbestiform amphibole fragments.
2 PRINCIPLE OF METHOD
2.1 Background
If handling of a particular vermiculite sample and its associated mineral
fragments presents any risk, the risk is represented by the vermiculite as it is
normally used. Therefore, in this analytical method, the vermiculite is analyzed in
the condition that it is normally used, and it is not crushed. Crushing of
vermiculite samples prior to analysis misrepresents any hazards and introduces
ambiguities of interpretation as follows:
(a) if asbestiform minerals are present in vermiculite, the fibers and fiber
bundles usually have a large size spectrum ranging from a few
micrometers in length up to the size of the vermiculite flakes
themselves. The risk, therefore, results from only those fibers and
3
-------
fiber bundles that have diameters lower than the upper limit of
respirability;
(b) if fragments of non .asbestiform amphibole are present in vermiculite,
they are usually large and very few of them have diameters within the
respirable size range. However, if the material is crushed, any
non-asbestiform amphibole fragments present will cleave along crystal
planes to generate large numbers of countable fibers that are not
representative of the material as it is used. Moreover, after such
crushing, the only way to determine whether the amphibole fibers
measured originated from amphibole asbestos or from cleavage of
non- asbestiform amphibole fragments is to measure the lengths and
widths of several hundred fibers and then to examine the aspect ratio
distribution. In this case, the fibers being measured would not have
been present in the original material, and reporting of these fibers as
asbestos would be misleading.
2.2 Types of Measurement
Two types of measurement are specified in this method.
(a) A rapid screening procedure for determination of the weight percent
amphibole or the weight percent amphibole asbestos is specified. For
this measurement, a known weight of the sample containing
exfoliated vermiculite is first suspended in water. Most of the
vermiculite floats to the top of the suspension, and this vermiculite is
removed and discarded. After allowing time for most of the
suspended material to settle, the water is decanted, and the sediment
is dried and weighed. The dried sediment, or a known weight of it, is
placed into a centrifuge tube and suspended in a heavy liquid of
density 2.75. After centrifuging, the centrifugate is separated and
weighed. The centrifugate is examined under a stereo-binocular
microscope. If there is more than approximately 0.01 % of amphibole
asbestos in the original sample, the fiber bundles are readily
recognized during the stereo-microscope examination, and it is
possible to hand-pick these fiber bundles from the centrifugate and
weigh them. Representative amphibole particles are identified by
PLM, with confirmation by either SEM .EDXA or TEMEDXA if
necessary.
(b) A TEM procedure is specified for determination of the number of
4
-------
respirable fibers per gram of sample, or per gram of respirable -
particles. This procedure can be applied to samples containing
exfoliated vermiculite, samples containing unexfoliated vermiculite, or
crude vermiculite ore. The sample is first suspended in room
temperature distilled water, and any floating material is removed.
Colder water is then introduced at the base of the container such that
the vertical rate of filling is equivalent to the falling speed of the
maximum- sized respirable particle. The displaced suspension is
collected, and the admission of cold water is terminated when all of -
the original suspension has been displaced. The displaced and
collected suspension contains all of the respirable particles. Aliquots
of the suspension are filtered through membrane filters, and TEM
specimens are prepared from the filters. The TEM specimens are
examined, and fibers are identified and their dimensions are recorded.
The balance of the suspension is filtered on to a pre-weighed
membrane filter. The filter is dried and weighed to obtain the weight
of respirable particles.
If the rapid screening determination of weight percent amphibole and the
concentration of respirable fibers are both required, it is possible to combine the
two procedures during processing of the sample.
In order to obtain an acceptable detection limit for determination of the
weight percent asbestos in vermiculite, the method relies on density separation
methods to remove as much of the vermiculite as possible prior to microscopical
analysis. Organic materials, if present, are removed by treatment of the sample in
a muffle furnace. If the sample contains unexfoliated vermiculite, the sample is
exfoliated either thermally or chemically in order to allow the maximum degree of
separation of the vermiculite by density separation methods.
3 SCOPE AND FIELD OF APPLICATION
3.1 Substance dete.mui d
3.1.1 Weight Percent Asbestos
The rapid screening method specifies a procedure to determine the weight
percent of amphibole asbestos.
5
-------
3.1.2 Concentration of Respirable Fibers
The method specifies a TEM procedure to determine the concentration of
respirable asbestos fibers in vermiculite or vermiculite containing samples. The
concentration of respirable asbestos fibers is expressed as the numerical
concentration per gram of sample, and as the numerical concentration per gram of
potentially respirable particulate material. The lengths, widths and aspect ratios of
the asbestos fibers and bundles are measured. The method allows determination
of the type(s) of asbestos fibers present. As for all routine TEM analytical
methods, this TEM method cannot discriminate between an individual, fiber of the
asbestos and non-asbestos analogues of the same amphibole mineral.
3.2 Type of Sample
The method is defined for solid samples containing vermiculite, including
beneficiated vermiculite, loose fill attic insulation, horticultural vermiculite, potting
soil, slow release horticultural fertilizers and fireproofing materials.
3.3 Range
The range of amphibole asbestos weight concentration that can be
measured is approximately 0.01% to 100%.
The minimum respirable fiber concentration that can be measured is
dependent on the volume of the suspension that can be filtered while still yielding
filters that are appropriately- loaded for preparation of TEM specimens. The
minimum for the respirable fiber concentration can be lowered by examination of a
larger area of the TEM specimens. There is no maximum, since the analytical
parameters can always be adjusted to accommodate high fiber concentrations. For
a sub-sample of 50 grams, it is usually possible to filter approximately 0.3 mL of
the suspension without overloading of the filter, and examination of 19 grid
openings of the TEM specimens yields an analytical sensitivity of approximately
100,000 fibers/g of original sample. However, the actual analytical sensitivity that
can be achieved is dependent on the nature of the sample.
3.4 Limit of Detection
For the rapid screening method, the limit of detection for amphibole asbestos
is less than approximately 0.01 % by weight.
6
-------
Theoretically 1 for determination of the concentration of respirable fibers, the
limit of detection can be lowered indefinitely by increasing the volume of liquid
filtered during specimen preparations and by increasing the area of the TEM
specimens examined in the electron microscope. In practice, for a particular area
of TEM specimens examined, the lowest achievable limit of detection is controlled
by the total amount of particulate material in the respirable size range. There is an
upper limit to the volume of the final suspension that can be filtered, if TEM
specimens of appropriate particulate loading are to be obtained. Lower limits of
detection can be achieved by increasing the area of the TEM specimens that is
examined. In order to achieve lower limits of detection for fibers and bundles
longer than 5 pm, and for PCM equivalent fibers, lower magnifications are specified
which permit more rapid examination of larger areas of the TEM specimens when
the examination is limited.to these dimensions of fiber.
7
-------
DRAFT
4 DEFINITIONS
Acicular: The shape shown by an extremely slender crystal with cross-sectional
dimensions which are small relative to its length, i.e. needle-like.
Amphibole: A group of rock.forming ferromagnesium silicate minerals, closely
related in crystal form and composition, and having the nominal formula:
Ao 1 B 2 C 5 T 8 0 22 (OH, F,Cl) 2
where: A = K, Na;
B = Fe 2 , Mn, Mg, Ca, Na;
C = Al. Cr, Ti, Fe 3 ’, Mg, Fe 2 ’;
T = Si, Al, Cr, Fe 3 , Ti.
In some varieties of amphibole, these elements can be partially substituted by Li,
Pb, or Zn. Amphibole is characterized by a cross-linked double chain of Si.O
tetrahedra with a silicon:oxygen ratio of 4:11, by columnar or fibrous prismatic
crystals and by good prismatic cleavage in two directions parallel to the crystal
faces and intersecting at angles of about 56° and 124°.
Amphibole asbestos: Amphibole in an asbestiform habit.
Analytical filter: A filter through which an aqueous suspension of particles is
passed. and from which TEM specimen grids are prepared.
Analytical sensitivity: The calculated asbestos structure concentration in asbestos
structures/g, equivalent to counting of one asbestos structure in the analysis.
Asbestiform: A specific type of mineral fibrosity in which the fibers and fibrils
possess high tensile strength and flexibility.
Asbestos: A term applied to a group of silicate minerals belonging to the
serpentine and amphibole groups which have crystallized in the asbestiform habit,
causing them to be easily separated into long, thin, flexible, strong fibers when
crushed or processed. The Chemical Abstracts Service Registry Numbers of the
most common asbestos varieties are: chrysotile (12001-29-5), crocidolite
(1 2001-28-4), grunerite asbestos (Amosite) (121 72-73-5), anthophyllite asbestos
(77536-67-5), tremolite asbestos (77536-68.6) and actinolite asbestos
8
-------
(77536-66-4). Less common asbestos varieties include the amphibole minerals
richterit and winchite in asbestiform habits.
Asbestos structure: A term applied to an individual fiber, or any connected or
overlapping grouping of asbestos fibers or bundles, with or without other particles.
Aspect ratio: The ratio of length tO width of a particle.
BeneficiatiOn: The process in which vermiculite is concentrated from the crude ore
and separated into different size fractions.
Blank: A structure count made on TEM specimens prepared from an unused filter,
to determine the background measurement.
Camera length: The equivalent projection length between the specimen and its
electron diffraction pattern, in the absence of lens action.
Chrysotile: A fibrous mineral of the serpentine group which has the nominal
composition:
Mg 3 S i 2 O 5 (OH) 4
Most natural chrysotile deviates little from this nominal composition. In some
varieties of chrysotile 1 minor substitution of silicon by Al 3 ’ may occur. Minor
substitution of magnesium by Al 3 ’, Fe 2 ’, Fe 3 ’, Ni 2 ’, Mn 2 ’ and Co 2 ’ may also be
present. Chrysotile is the most prevalent type of asbestos.
Cleavage: The breaking of a mineral along one of its crystallographic directions.
Cleavage fragment: A fragment of a crystal that is bounded by cleavage faces.
Cluster: An structure in which two or more fibers, or fiber bundles, are randomly
oriented in a connected grouping.
Density separation: A procedure in which particles of different densities are
separated by suspension in a liquid of selected density.
d.spacing: The distance between identical adjacent and parallel planes of atoms in
a crystal.
Electron diffraction: A technique in electron microscopy by which the crystal
structure of a specimen is examined.
9
-------
Electron scattering power: The extent to which a thin layer of substance scatters
electrons from their original directions.
Energy dispersive X-ray analysis: Measurement of the energies and intensities of
X-rays by use of a solid state detector and multi channel analyzer system.
Eucentric: The condition when the area of interest of an object is placed on a
tilting axis at the intersection of the electron beam with that axis and is in the
plane of focus.
Exfoliation: A process in which vermiculite flakes are expanded by sudden heating
or by chemical action.
Falling speed: The speed at which a particle falls through a fluid when the
downward force due to the particle mass is in equilibrium with the resistive force
due to viscosity of the fluid.
Fibril: A single fiber of asbestos, which cannot be further separated longitudinally
into smaller components without losing its fibrous properties or appearances.
Fiber: An elongated particle which has parallel or stepped sides. A fiber is defined
as having an aspect ratio equal to or greater than 5:1 and a minimum length of
0,5 pm.
Fiber bundle: A structure composed of parallel, smaller diameter fibers attached
along their lengths. A fiber bundle may exhibit diverging fibers at one or both
ends.
Fibrous structure: A fiber, or connected grouping of fibers, with or without other
particles.
Funnel blank: A structure count made on TEM specimens prepared by the
direct-transfer method from a filter used for filtration of a sample of distilled water.
Habit: The characteristic crystal growth form or combination of these forms of a
mineral, including characteristic irregularities.
Umit of detection: The calculated airborne asbestos structure concentration in
structures/g, equivalent to counting of 2,99 asbestos structures in the analysis.
Matrix: A structure in which one or more fibers, or fiber bundles, touch, are
10
-------
attached to, or partially concealed by, a single particle or connected group of
non- fibrous particles.
Miller index: A set of either three or four integer numbers used to specify the
orientation of a crystallographic plane in relation to the crystal axes.
PCM equivalent fiber: A fiber of aspect ratio greater than or equal to 3:1, longer
than 5 pm, and which has a diameter between 0,2 pm and 3,0 pm.
Replication: A procedure in electron microscopy specimen preparation in which a
thin copy, or replica, of a surface is made.
Respirable fiber: A fiber of aspect ratio greater than or equal to 3:1, longer than
5 pm, and which has a diameter equal to or lower than 3,0 pm.
Selected area electron diffraction: A technique in electron microscopy in which the
crystal structure of a small area of a sample is examined.
Serpentine: A group of common rock-forming minerals having the nominal
formula:
Mg 3 Si 2 O 5 (OH) 4
Spherical Equivalent Diameter: The diameter of a sphere of unit density that has
the same falling speed in air as the particle under consideration.
Structure: A single fiber, fiber bundle, cluster or matrix
Twinning: The occurrence of crystals of the same species joined together at a
particular mutual orientation, and such that the relative orientations are related by
a definite law.
Unopened fiber: A large diameter asbestos fiber bundle which has not been
separated into its constituent fibrils or fibers.
Zone-axis: The line or crystallographic direction through the center of a crystal
which is parallel to the intersection edges of the crystal faces defining the crystal
zone. -
11
-------
5 ABBREVIATIONS
DMF - Dimethyl formamide
ED - Electron diffraction
EDXA Energy dispersive X-ray analysis
FWHM Full width, half maximum
HEPA - High efficiency particle absolute
MEC Mixed esters of cellulose
PC • Polycarbonate
PCM - Phase contrast optical microscopy
SAED Selected area electron diffraction
SEM - Scanning electron microscope
STEM - Scanning transmission electron microscope
TEM - Transmission electron microscope
UICC - Union Intemationale Contre le Cancer
6 EQUIPMENT AND APPARATUS
6.1 General
General laboratory equipment, such as glass beakers, disposable pipets,
disposable plastic beakers and measuring cylinders, is required, with the addition of
the specific items listed below. Some analyses do not require all of the equipment
listed.
12
-------
6.2 Sample preparation
6.2.1 Laboratory balance, sensitivity 0.0001 gram
6.2.2 Muffle furnace, temperature range up to 800°C
6.2.3 Fused silica tray, approximately 15 cm x 9 cm
6.2.4 Laboratory magnetic stirrer
6.2.5 Teflon coated magnetic stirrer bars
6.3 Rapid Screening Method
6.3.1 SINK-FLOAT® Standard, density 2.75±0.005 g/cc at 23°C. Cargille
Laboratories, Inc., Cedar Grove, New Jersey 07009.
6.3.2 Centrifuge, capable of 3600 rpm and accommodating four or more
15 mL centrifuge tubes
6.3.3 Water aspirator
6.3.4 Stereo-binocular microscope, lOx to 40x magnification
6.3.5 Polarized light microscope
6.3.6 Scanning electron microscope, with energy dispersive x-ray analysis
system
6.4 Measurement of Respirable Fibers by TEM
6.4.1 Peristaltic pump capable of pumping 15-25 mI/minute
6.4.2 Glass filtration system, 25 mm diameter
6.4.3 Transmission electron microscope, as specified in ISO 13794
6.4.4 Energy dispersive x-ray analysis system, as specified in !SO 13794
13
-------
7 REAGENTS
7.1 High density, eitherl- 1-2.2- tetrabromoethane or tribromomethane
7.2 Ethanol, reagent grade
7.3 Reagent water, either freshly-distilled or deionized water, filtered through an
MCE filter of maximum porosity 0.22 pm. and meeting the requirements of
ASTM D 1193 for reagent water.
Note: For analyses incorporating TEM specimen preparation, it is important that
the reagent water be freshly produced and filtered, in order to minimize
bacterial interferences on TEM specimens.
7.4 Dimethylformamide 1 reagent grade
7.5 Acetone, reagent grade
7.6 Glacial acetic acid, reagent grade
7.7 Diaminoethane, reagent grade
7.8 Hydrochloric acid, concentrated reagent grade
8 SELECTION AND PRE-TREATMENT OF SUB-SAMPLE FOR ANALYSIS
8.1 Types of Sample
Samples presented for analysis may include:
(a) Exfoliated vermiculite used as loose-fill attic insulation or as horticultural
soil conditioner;
(b) Potting soil containing vermiculite, peat moss. fertilizers and other
constituents;
(c) Beneficiated crude vermiculite ore, which is the form in which vermiculite
is transported from a mine to an exfoliation facility;
(d) Soil samples containing exfoliated vermiculite, originating from land to
which vermiculite products have been applied in the past;
14
-------
DRAFT
(e) Soil samples containing crude vermiculite ore, such as could be found on
the ground in the vicinity of vermiculite mines or vermiculite exfoliation
facilities;
(f) Samples of insulation incorporating materials in addition to vermiculite.
8.2 Obtaining a Representative Sub-Sample for Analysis
Products such as potting soil often have a substantial water content, and
such products shall be dried before analysis. The sample shall be weighed before
and after drying to obtain the weight of water, so that the final results can be
expressed in terms of the original weight or dry weight of the sample.
If amphibole is present in a vermiculite product, the size range of the
fragments of amphibole is usually approximately the same as that of the
vermiculite flakes, because during the beneficiation process the material is
segregated into several different size categories. The fragments of amphibole are
distributed randomly throughout the vermiculite, and the number of these
fragments is generally much lower than the numbers of vermiculite flakes.
Accordingly, if a reproducible analysis is to be obtained, it is necessary to select a
sub-sample of vermiculite sufficiently large that a statistically-valid number of the
amphibole fragments are included. The weight of sub-sample required is
dependent on the size grade of the vermiculite, and the percentage of vermiculite
in the sample. Table ‘1 gives recommended approximate weights of vermiculite
that should be used for the initial sub-sample. For products containing vermiculite,
a visual estimate of the proportion of vermiculite in the product should be made
and the starting weights in Table 1 should be proportionately increased.
Table 1. Recommended Sub- Sample Weights of Vermiculite for Analysis
Size of Vermiculite Flakes, mm
Recommended Minimum Starting
Weight for Analysis, grams
<2
5
>2- <5
10
>5
50
15
-------
The sub. sample shall be obtained from the original sample by the cone and
quarter method. On a clean surface, such as a sheet of aluminum foil, form the
sample into a cone. Using a thin flat sheet of metal or rigid plastic, divide the cone
into two parts, vertically from the apex. Form either of the two fractions into a
cone, and repeat the procedure until one of the separate fractions is of a suitable
weight for analysis.
8.3 Pre-Treatment of Sub-Samples
8.3.1 Background
In order to obtain the best detection limits and reproducibility for
determination of asbestos, pre- treatment is required for samples other than
exfoliated vermiculite. The pre-treatments are summarized in the flow-chart shown
in Figure 1. This analytical method is based on selective removal of exfoliated
vermiculite by water flotation, so any sample that contains unexfoliated vermiculite
shall be exfoliated prior to analysis. If the analyses are intended to measure only
amphibole asbestos, exfoliation in a muffle furnace may be used. As is the case
for industrial exfoliation, chrysotile is unlikely to survive the thermal treatment
received during laboratory exfoliation, since chrysotile degrades at temperatures
exceeding approximately 500°C.
‘8.3.2 Exfoliated Vermiculite
Exfoliated vermiculite, as marketed for application as loose attic fill insulation
or for horticultural use, requires no pre-treatment before analysis. Vermiculite
insulation samples originating from attics, and samples identified as horticultural
vermiculite may be submitted directly to the analytical procedure.
8.3.3 Potting Soil
Vermiculite may represent only a fraction of the weight of potting soil. The
other components are generally organic with some plant nutrients. Water may
also represent a significant proportion of the weight. The results of the analysis
should generally be expressed in terms of dry weight, because the water content
may vary depending on the storage conditions. Weigh a container (a disposable
container formed from aluminum foil is suitable), place the sub-sample in the
container and weigh again. Dry the sub-sample for a period of approximately
16
-------
I
Attic insulation and
Horticultural Vermiculite
j
Weigh Sub-Sample
Beneficiated
Vermiculite Ore
4 __4.
Proceed to Analysis
Figure 1. Pre-treatment of samples containing vermiculite
Cone and Quarter
to Produce
Sub-Sampleof Appropriate Size
I
Exfoliate
Thermally
at 800°C
4
Exfoliate
Chemically
( 30% H 2 0 2 )
4,
Ash at 480°C
to Remove
Organic Materials
L I
17
-------
10 hours, either on a slide warmer or in an oven at a temperature of approximately
60°C, and then re-weigh. Calculate the weight of water evaporated from the
sample.
Transfer the sub- sample to a fused quartz tray or other suitable open
container for ashing at 480°C. Place the sub- sample in a muffle furnace operating
at a temperature of 480±10°C for a period of approximately 10 hours. Weigh the
residual ash, and calculate the percentage as a proportion of the original dried
sub. sample. The residual ash may be submitted directly to the analytical
procedure.
8.3.4 Lawn Fertilizer
Vermiculite-based lawn fertilizers contain water-soluble plant nutrients and
sometimes organic herbicides. Vermiculite generally represents only a small
proportion of the weight. Although it is possible to dissolve out most of the
nutrients and herbicides by extraction with hot water,
8.3.5 Beneficiated Crude Vermiculite Ore
The analytical procedure takes advantage of the low density of exfoliated
vermiculite to allow separation of the majority of the vermiculite by flotation on
water. Samples of beneficiated crude vermiculite ore must therefore be exfoliated
prior to analysis. There are two ways by which beneficiated vermiculite can be
exfoliated.
(a) The vermiculite can be exfoliated at high temperature in a muffle furnace,
which simulates the manner in which vermiculite is exfoliated
commercially. This procedure is rapid and inexpensive. Weigh the
sub-sample of vermiculite. An example of a sub-sample of beneficiated
vermiculite is shown in Figure 2. Set the temperature of the muffle
furnace to 800°C, and place a fused silica tray into the furnace. Have
available a large glass or metal container available to receive the
exfoliated vermiculite. A sheet of aluminum foil, formed into a container,
has been found satisfactory. Using crucible tongs, remove the silica tray
from the muffle furnace. The tray will be at a red heat. Sprinkle a small
amount of the vermiculite sample into the silica tray as shown in
Figure 3. Return the silica tray to the muffle furnace, as shown in
Figure 4, and close the furnace door for approximately 15 seconds to
complete the exfoliation. Remove the silica tray from the muffle furnace,
18
-------
Figure 3 Sprinkling of Beneficiated Vermiculite into Heated Silica
Tray
Figure 2 Example of Sub.Sample of Beneficiated Vermiculite
19
-------
Figure 4 Portion of SubSample Before Exfoliation in Muffle
Furnace
Figure 5 Example of Exfoliated Sub-Sample
20
-------
and pour the exfoliated vermiculite into the container. Repeat this procedure as
many times as necessary, until all of the vermiculite has been exfoliated. The
exfoliated product is shown in Figure 5
(b) The vermiculite can be exfoliated chemically, using 30% hydrogen
peroxide solution. This procedure permits the vermiculite to be exfoliated
at low temperatures. As shown in Figure 6, place the vermiculite
sub- sample into a glass beaker sufficiently large to accommodate the
volume of the exfoliated product. Add 30% hydrogen peroxide, equal to
approximately twice the volume of the vermiculite, and allow the
container to stand for approximately 48 hours at room temperature. The
sub- sample will exfoliate as shown in Figures 7 and 8. Dry the exfoliated
product either on a hotplate or in an oven.
8.3.6 Soil Samples Containing Crude Vermiculite Ore
Soil samples containing crude (unexfoliated) vermiculite ore probably also
contain organic constituents which need to be removed prior to analysis.
Exfoliation in a muffle furnace is the optimum approach, because in one operation
the vermiculite is exfoliated and the organic constituents are oxidized. However, if
chrysotile is suspected to be present and is of concern, the organic materials may
be oxidized in a muffle furnace at a temperature of 480°C and the vermiculite may
then be exfoliated chemically using 30% hydrogen peroxide as described in 8.3.5.
8.3.7 Soil Samples Containing Exfoliated Vermiculite
Soils which have been amended with products containing exfoliated
vermiculite will probably also contain substantial amounts of organic materials.
The organic constituents shall be oxidized in a muffle furnace at 480°C.
8.3.8 Samples of insulation incorporating materials in addition to vermiculite.
Vermiculite-based insulation and fireproofung products often contain as much
as 60% of other materials such as gypsum and calcium carbonate, and also
possibly chrysotile. Constituents such as gypsum and calcium carbonate can be
removed by treatment in 10% hydrochloric acid, without affecting the vermiculite.
If chrysotile is present, the treatment will cause the refractive indices to be
lowered slightly, but the optical propăties of any amphibole asbestos present will
not be affected.
21
-------
Figure 6 SubSample of Beneficiated Vermiculite
Before Chemical Exfoliation
22
Figure 7 Sub.Sample of Beneficiated Vermiculite
After Chemical Exfoliation
-------
Gently break the product into fragments of about 0.5 cm dimension, using a
mortar and pestle. The purpose of this crushing is to facilitate the dissolution of
the carbonates and gypsum by hydrochloric acid. Place the product into a beaker
with a Teflon coated magnetic stimng bar. Add an excess of 10% hydrochloric
acid. The amount of acid to be added is dependent on the weight of the sample
and the proportion of carbonates and gypsum. For gypsum, the solubility in water
is approximately 2.4 gIL. Perform the dissolution procedure at room temperature,
because gypsum is unusUal in that. its solubility in water reduces with increase of
temperature. Stir the suspension for approximately 15 minutes, and filter using a
pre-weighed polycarbonate filter of maximum pore size 0.8 pm. Dry and wwgh the
filtered residue.
9 PROCEDURE FOR ANALYSIS
9.1 IntrodUction
In some situations, it is required only to determine if amphibole asbestos is
present, and if so, to determine the percentage by weight. In this case, the rapid
screening analysis as described in 9.2 should be used. The rapid screening
analysis also incorporates an optional TEM procedure to be used if no amphibole
Figure 8 Appearanceof Vermiculite After Chemical Exfoliation
23
-------
asbestos is detected in the residue from the analysis. This TEM examination can
give additional assurance that asbestos is not present.
In other situations, the number of respirable fibers per gram of the respirable
dust fraction, or the number of respirable fibers per gram of original sample is
required. The analysis as described in 9.3 should be used. If all three
measurements are required for the same sample, the analysis described in 9.4
should be used.
9.2 Rapid Screening Analysis to Determine the Weight Percent of Amphibole
Asbestos
9.2.1 General
The rapid screening analysis is designed to determine the weight percent of
amphibole asbestos in an exfoliated vermiculite sample, in which the particle sizes
range up to some millimeters in dimension. Figure 9 shows a flow-chart which
summarizes the analytical procedure. -
9.2.2 Separation of Vermiculite from other Components by Flotation on Water
Place 800 mL of reagent water into a 1000 mL glass beaker. Using a spoon,
place a portion of the exfoliated vermiculite sub-sample into the beaker, and
immerse the vermiculite several times by pushing it under the surface using the
spoon. Remove the floating vermiculite and discard it. Continue to wash portions
of the vermiculite in this manner until all of the sub-sample has been treated.
Carefully remove alt fragments of vermiculite from the surface of the water, and
allow the suspension to settle for 60 minutes. After this period of time, all
amphibole fibers thicker than approximately 3 pm will have settled to the bottom
of the beaker. Using a pump or syphon, transfer the supernatant liquid to a
second beaker. Using ethanol, wash the sediment from the first beaker into a
glass petri-dish and dry the sediment by placing the petri-dish on a slide warmer at
a temperature of approximately 60°C. Use of an oven for drying the sediment is
not recommended, because of the hazards associated with evaporation of ethanol
in a closed environment. Transfer the sediment to a pre-weighed dish, and weigh
the dish to obtain the weight of the sediment. Figure 10 shows an example of
sediment after the water sedimentation procedure.
24
-------
9.2.3 Optional Preparation of TEM Specimens From the Aqueous Suspension of
Vermiculite
If amphibole asbestos is detected in the centrifugate, it can be assumed that
respirable amphibole asbestos fibers are present in the aqueous suspension, and
that some would be present in airborne dust generated from the vermiculite. If
amphibole asbestos is not detected in the centrifugate. there is still a possibility
that fine amphibole asbestos fibers, too small for detection by the stereo binocular
microscope or PLM, could be present. This possibility can be confirmed or
-discounted by examination of particles in the aqueous suspension by TEM.
Prepare analytical filters by the procedure described in 9.3.4. It is beyond the
scope of this document to describe the preparation of TEM specimens from
membrane filters; these procedures are fully described in ISO 13794
9.2.4 Density Adjustment of Heavy Liquid
Prepare a heavy liquid of density 2.75 g/mL, by addition of reagent ethanol to
either 1-1.2. 2-tetrabromoethane or tribromomethane (bromoform). Pure
1-1-2-2-tetrabromoethane has a density of 2.96 gImL, and tribromomethane has a
density of 2.89 g/mL.
Use of a SINK-FLOAT® standard is the most convenient and rapid method for
adjustment of the liquid density. Figure 11 shows a SINK-FLOAT standard. This
device is a short, sealed, glass tube containing weights, the overall weight of
which is calibrated such that it does not sink or float in a liquid of the specified
-density. Determine the required volume of heavy liquid, and place approximately
90% of this volume of 1-1-2- 2-tetrabromoethane or tribromomethane into a glass
beaker. Place the SINK-FLOAT® standard into the beaker, and add reagent ethanol
until the SINK-FLOAT® standard is suspended in the liquid without either sinking or
floating, as shown in Figure 1 2 . It is recommended to add the ethanol in small
amounts, and to stir with a glass rod after each addition to ensure thorough mixing
before adding further ethanol. If too much ethanol is added, the SINK- FLOAT®
standard will sink, indicating that the density is lower than 2.75 gImL. In this
case, more of the heavy liquid can be added to adjust the density upwards to the
correct value.
If a SINK-FLOAT® standard is not available, the density of the heavy liquid
can be adjusted using a calculated addition of ethanol. Figure 13 shows the
concentration of ethanol that must be present in a mixture with either of these
heavy liquids in order to achieve a density of 2.75 g/mL. Before use, the density
of the liquid produced by this method shall be confirmed by the density bottle
method, and the density adjusted if necessary.
25
-------
Yes No
Hand-Pick Fiber Bundles, Weigh, and
Identify by PLM and/or SEM-EDXA
Filter Aliquots of Suspension
for Optional TEM Examination
Figure 9 Flow. Chart for Rapid Screening Method for Determination of Weight Percent Asbestos in
Vermiculite
Suspend Vermiculite in Filtered Distilled Water
and Remove Floating Vermiculite
Filter Suspended and Sedimented Particles
using 0.4 pm Pore Size Polycarbonate Filter
+
I Dry Filter and Separate Filter Cake from Filter I
Disperse Filtered Material
in 2.75 Density Liquid and Centrifuge
Weigh Centrifugate and Examine by Stereo-Microscope
Weighable Amounts of Large Fiber Bundles Observed
4,
‘ I ,
Particle Count by PLM and SEM-EDXA
to Determine Weight % Amphibole, or
Stereo-Binocular Microscope Estimation
26
-------
DRAFT
Figure 10 Example 01 Sediment After Water Sedimentation
27
-------
Figure 11 SINK-FLOAT Standard For Adjustment of Liquid Density
Figure 12 Position of SINK.FLOAT• at Liquid Density of 275 glmL
28
-------
3.0
2.8
2.6
2.4
Percent Ethanol
Figure 13 Volume Percent Ethanol Required for Adjustment of Heavy Liquid Density
0 5 10 15 20
25
E
C)
U)
w
29
-------
9.2.5 Separation of Amphibole Fragments by Centrifugation in a Heavy Liquid
If possible, it is preferable to perform the heavy liquid separation on the
entire weight of sediment from the water sedimentation. Using 15 mL centrifuge
tubes, it may be necessary to divide the sediment into more than one centrifuge
tube. It is recommended that no more than approximately 3-4 cm depth of
sediment be processed in each 15 mL centrifuge tube. Add the density-adjusted
heavy liquid to each of the centrifuge tubes until the level is approximately 0.5 cm
from the top of the tube. Place the centrifuge tubes into the head of the
centrifuge, balancing the head with one or more tubes containing heavy liquid if
necessary. The dimensions and rotation speed of the centrifuge determine the
centrifugation time necessary. The time required for sedimentation of particles can
be calculated from the formula:
18 108
.ln(R/S)
60.(a- b).w 2 .d 2
Where: t = the time in minutes for particles of diameter d to sediment
= the coefficient of viscosity of the heavy liquid in poise
a = the density of the particle in g/cc
b = the density of the heavy liquid
w = the angular velocity of the centrifuge in radians/second
d = the diameter of the particle in micrometers
R = the outer radius of the centrifuge
S = the inner radius of the centrifuge
30
-------
Particle Diameter (jim)
Figure 14 Example of Sedimentation Times for Particles Centrifuged in Liquid of Density 2.75
2
E
0
I
100
0.1
10
1
1
10
31
-------
Figure 14 shows an example of times for sedimentation, using a liquid of density
2.75 gImL, and a typical bench top centrifuge. The centrifuge parameters used in
Figure 13 are 3600 revolutions per minute, with an outer radius of 14.5 cm and an
inner radius of 3.5 cm.
Figure 15
Example of Centrifuge Tubes After
Centrifugation
V
32
-------
With the example centrifuge, centrifugation for 10 minutes ensures that all
amphibole particles larger than approximately 3 pm will have sedimented. During
centrifugation, the residue separates into a sedimented fraction (centrifugate) and a
floating fraction, with very little material remaining suspended. Figure 15 shows
an example of centrifuge tubes after centrifugation of the sediment from water
suspension.
The next step in the analysis is to remove the floating and suspended
fractions, leaving a clean centrifuge tube from which the centrifugate can be
washed out. Using a small spatula, remove as much of the floating material as
possible, particularly any large particles, as shown in Figure 16. Discard this
material. The rest of the floating material and any particles remaining suspended in
the heavy liquid is removed using a suction tube. Set up an Erlenmeyer flask as
shown in Figure 17. Turn on the water aspirator, and lower the thinned
polyethylene tube into the centrifuge tube as shown in Figure 18. Aspirate all of
the floating material first, and then lower the polyethylene tube to remove nearly
all of the supernatant liquid, avoiding any disturbance of the centrifugate at the
bottom of the tube. Around the inside of the top of the centrifuge tube, some of
the floating material will remain. This must be removed carefully using paper
towel, in order to avoid contaminating the centrifugate with the floating material.
Wrap a strip of paper towel, approximately 5 cm in width and 20 cm long, around
the end of a small spatula. Holding the centrifuge tube almost horizontal so that
residual floating material does not accidentally fall into the tube while it is being
removed, insert the end of the spatula with the rolled paper towel into the
centrifuge tube, and wipe around the inside with an upwards motion as shown in
Figure 19. As the paper towel collects material, tear off a portion to expose clean
paper, and continue the action of removing all of the floating material from the
inside of the centrifuge tube. When all of the floating material has been removed,
it is necessary to wash the centrifugate several times with ethanol in order to
remove all traces of the heavy liquid. Fill the centrifuge tube with ethanol from a
wash-bottle. Aim the stream of the wash-bottle at the base of the centrifuge tube
a shown in Figure 20, in order to disturb and disperse the cake of particulate.
Repeat this procedure with any other centrifuge tubes used for sample processing.
Pour out the heavy liquid in the tubes that were used for balancing the
centrifuge, and this clean heavy liquid may be re-used. Fill these centrifuge tubes
with ethanol, for use in balancing the centrifuge during washing of the centrifugates
Centrifuge the tubes for 3 minutes to allow the particulate to sediment. Set
up a second Erlenmeyer flask suction tube system. Turn on the water aspirator,
and lower the thinned polyethylene tube into the centrifuge tube. Aspirate all of
ethanol, taking care not to disturb the centrifugate at the bottom of the tube.
33
-------
DRAFT
Figure 17 Example of Suction Device to Asprate Liquid From
Centrifuge Tubes
Figure 16 Removal of Large Floating Particles Using Spatula
34
-------
Figure 19 Removal of Residual Floating Material From Inside of
Centrifuge Tube Using Strips of Paper Towel
Figure 18 Removal of Floating Particles and liquid Using Aspirator
35
-------
-. ;. —
I
j_.mt : 5 _7 -
Ii _ 4
Figure 20 Washing of Centrifugate With Ethanol
Figure 21 Removal of Ethanol Washings Using Aspirator Tube
36
-------
- . /2 ’
;t \’
:t 4
Figure 22 RinsEig of Centrifugate from Centrifuge Tube Using
Ethanol
Repeat this ethanol washing procedure two more times. Hold the centrifuge tube
inverted at an angle of approximately 45° over a pre-weighed, 47mm diameter.,
plastic petri-dish. Using the wash-bottle with ethanol, aim the jet at the bottom of
the centrifuge tube as shown in Figure 22 and wash all of the centrifugate from
the tube to the petri-dish. If more than one centrifuge tube was used because the
sample size was large, combine all centrifugates into the one petri-dish. During
combination of centrifugates, it may be necessary to decant some of the ethanol
from the petri.dish in order to avoid overflow. Place the petri-dish on the slide
warmer to evaporate the ethanol. After the centrifugate is dry, weigh the dish to
obtain the weight of the centrifugate.
9.2.6 Stereo-Binocular Microscope Examination of the Centrifugate
The centrifugate contains particles with densities exceeding 2.75 glcc, which
includes any amphibole particles present in the original sub-sample. There are
three possible outcomes which define the extent to which further analytical work
is necessary. The procedure shall be either (a), (b) or (c).
(a) If the sample originated from Libby , Montana, the centrifugate will
contain a major proportion of large fiber bundles that are gray.green in
37
-------
color, and are easily visible under the stereo- binocular microscope at
magnifications up to 40. If a sub-sample of sufficient size was used,
numerous fiber bundles should be present in the centrifugate, as shown
in Figure 23. The analyst will generally have no difficulty recognizing
these fiber bundles. To quantify the fiber bundles, the analyst must
determine whether the more efficient approach is to pick the fiber
bundles from the centrifugate for weighing, or to remove non-asbestos
particles from the centrifugate and weigh the balance of the material.
The fiber bundles picked from such a centrifugate are shown in
Figure 24. After the fiber bundles have been weighed, representative
bundles shall be selected for identification by either PLM, SEM or TEM.
The morphology, color and optical properties of the amphibole asbestos
fibers in vermiculite originating from Libby are characteristic, and with
experience, the analyst need go no further than mounting representative
fiber bundles in a high dispersion liquid of refractive index 1.630, in
which the very fine fibers exhibit dispersion staining colors of magenta to
gold (parallel) and blue (perpendicular). Representative fiber bundles may
be examined by SEM or TEM, and the EDXA spectra obtained may be
used as the basis for identification.
(b) If the sample originated from a mine other than Libby, Montana. few
amphibole asbestos fiber bundles, if any, may be observed in the
centrifugate during the stereo-binocular microscope examination.
However. the centrifugate may contain a large proportion of
non-asbestiform amphibole fragments. These fragments are not
asbestos . Non-asbestos amphibole fragments are recognized by cleavage
planes parallel to the length of the crystal, intersecting at angles of
approximately 56° and 124°. In well- crystallized material, these angles
can be recognized by examination of the ends of elongated fragments.
such as in the centrifugate from another vermiculite sample shown in
Figure 25. The total amount of non-asbestiform amphibole may be
estimated by hand- picking of fragments and weighing, using the same
procedure as defined in (a). If required, identification and quantification
of the individual non-asbestiform amphiboles present are best performed
by SEM, since there is an overlap in the optical properties of amphiboles
such as actinolite and homblende, and mixtures of amphibole types may
be present.
(C) One situation that sometimes occurs is that, during the
stereo-microscope examination, only a few amphibole asbestos fiber
bundles may be visible in the centrifugate, along with fragments of
non asbestiform amphiboles and other minerals. In this case, it is unlikely
38
-------
Figure 23 Example of Centrifugate Containing Asbestos Fiber
Bundles
1—I I Il-ti I I i i1111111 IIII1I1IIIii11IlJ1.I
Figure 24 Asbestos Fiber Bundles Hand Picked From Centrifugate
39
-------
•g :1.1111 1:11 Fl
Figure 25 Example of Non Asbestiform Actinolite Detected in the
Centrifugate from a Vermiculite Sample
that random sampling of particles for either SEM particle counting or PIM
examination would include any of these asbestos fiber bundles, and a
false-negative result would be reported. If the aggregate of the
amphibole asbestos fiber bundles is within the range of the chemical
balance, the best approach is to pick them from the centrifugate and
weigh them. The statistical validity of the calculated concentration may
be limited by the low number of fiber bundles, but in general this is
inconsequential since the result is close to the limit of detection.
If it is found that the aggregate weight of the fiber bundles Is below the
sensitivity of the balance, it is necessary to approximate their weight
concentration by other methods. Two approaches to determining an
estimate of the amphibole asbestos concentration are available, as
described in (1) and (2).
(1) an estimate of the upper bound of the amphibole asbestos
concentration may be made by assuming the sensitivity of
the balance as the weight of amphibole asbestos. In many
cases, this may be sufficient for the purpose;
NonAsbostIfO rn Tremolite, Brazil”
40
-------
(2) An approximation of the number of particles in the
centrifugate may be made by estimation of the average
particle size and assuming that they all have a density of
2.75. The weight percentage of any observed amphibole
asbestos fiber bundles may then be approximated by a
simple ratio of the number of amphibole asbestos fiber
bundles to the calculated number of particles in the
centrifugate. While this approach yields only an
approximation of the concentration, the approximate nature
of the result is generally inconsequential because the value
obtained is close to the limit of detection.
In the event that a low concentration of amphibole asbestos is reported,
representative fiber bundles shall be identified either by SEM or PLM. In
the majority of cases, amphibole asbestos can be identified satisfactorily
by PIM alone.
9.3 Determination of Concentration of Respirable Fibers by Transmission Electron
Microscopy
9.3.1 Introduction
In this procedure, the sample of exfoliated vermiculite is first dispersed in
water, and all floating vermiculite is removed. Using a peristaltic pump,
cold water is then introduced at the base of the container at a rate such
that the vertical velocity in the suspension container is slightly higher
than the rate of fall of the largest respirable size particle. The cold water
remains at the bottom of the beaker and the suspension is displaced.
The displaced suspension overflows the container and is collected in a
second container. TEM specimens are then prepared from the displaced
suspension , and the balance of the suspension is filtered through a
pre-weighed filter. After drying, the filter is weighed to determine the
total weight of respirable particles. Readily-available laboratory apparatus
is used to perform this measurement. Alternative apparatus may be
used, provided that the basic parameters of the measurement remain
constant. Figure 26 shows a flow chart which illustrates the procedure.
If the purpose is to measure only the respirable fiber concentration, and
there is no interest in the weight percent measurement, the
sedimentation and TEM procedure in Figure 26 should be followed.
41
-------
9.3.2 Separation of Coarse Vermiculite
Place 800 mL of reagent water into a 1000 mL glass beaker. Using a
spoon, place a portion of the exfoliated vermiculite sub-sample into the
beaker, and immerse the vermiculite several times by pushing it under the
surface using the spoon. Remove the floating vermiculite and discard it.
Continue to wash portions of the vermiculite in this manner until all of
the sub-sample has been treated. Carefully remove all floating ftagments
of vermiculite from the surface of the water, using a spatula. A piece of
paper towel touched to the surface of the water has been found useful in
removing the final traces of floating vermiculite.
9.3.3 Separation of Respirable Fibers by Displacement Sedimentation
After all of the floating vermiculite has been removed, make the
suspension up to a volume of 1 liter using reagent water at roomS
temperature or several degrees above room temperature. Place the
beaker into a calibrated ultrasonic bath for 2 minutes. Remove the
beaker from the ultrasonic bath, and mix the contents by air bubbling
using filtered air. Attach the polyethylene tube to the beaker as
illustrated in Figure 27, and connect it to the peristaltic pump set at a
suitable volume flow rate to produce the correct upward displacement
velocity. Figure 28 shows the falling speeds in water of specific size
particles in the density range exhibited by amphiboles. Respirable
particles of aerodynamic diameter 10 pm, with a density of 3.4 (the
maximum density for the amphibole asbestos varieties) have a physical
diameter of approximately 5.42 pm. Particles of this diameter have a
falling speed in water of 3.82 x 1 Q3 cm/s , and this upward velocity must
be established during the displacement sedimentation. Attach the cold
water supply to the peristaltic pump, and activate the pump. Depending
on the size of the container used for the vermiculite suspension, the
displacement process may take Lip to 1 hour for completion. The water
supply to the peristaltic pump must remain cold for this period. One way
of ensuring this is to cool the water supply container by immersion in a
larger container containing ice cubes. Progress of the displacement
process may be monitored visually if the cold water supply is colored.
Normal red food color has been found satisfactory as an indicator.
42
-------
9.3.4 Preparation of TEM Specimens From Displaced Suspension
After all of the vermiculite suspension has been displaced, pour the displaced
suspension into a 1 liter glass beaker, and homogenize the suspension by agitation
using filtered air. Filtration of the aqueous suspension is a very critical procedure
because it is important to obtain uniform deposits of particulate on the analytical
filters. The following procedure shall be used.
(a) Set up the filtration system and connect to a vacuum source;
(b) add freshly distilled water to the filtration unit base component until there
is a raised meniscus;
(c) place a 5 pm pore size cellulose ester filter on to the water meniscus.
The filter will centralize. Apply the vacuum very briefly in order to bring
the filter into contact with the base component; /
(d) add freshly distilled water to the top of the cellulose ester filter, and
place the analytical filter (either a 0.2 pm maximum pore size
capillary-pore polycarbonate filter or a 0.22 pm maximum pore size
cellulose ester filter) on to the water surface. Apply the vacuum very
briefly again in order to bring both filters into contact with the base
component;
(e) install the filtration reservoir and clamp the assembly together.
(f) Before filtering the aqueous suspensions, prepare a funnel blank by
filtration of 40 ml of freshly-distilled water. This sample is a control to
ensure that the filtration equipment is clean and the reagent water is not
contaminated by fibers.
43
-------
Suspend Vermiculite in Filtered Distilled Water ]
I I
Remove Floating Vermiculite i
Place in Ultrasonic Bath and
Agitate by Bubbling Filtered Air
I Sediment Non-Respirable Particles
Filter Aliquots of
Aqueous Suspension
Filter Sedimented
Particles and Combine
With Suspended Particles
4
Heavy Liquid
Density Separation
Analyze Centrifugate by
PLM and/or SEM-EDXA
Weight Percent
Amphibole
Figure 26 Summary of Analytical Method for Determination of Asbestos in Vermiculite
w
Filter Balance of
Suspended Particle J
Prepare TEM Specimens and Analyze I
.4,
Respirable Fibers/Gram of Resipirable Dust
Respirable Fibers/Gram of Vermiculite
r
44
-------
102
I
I
a,
Partide D amet (InTO
Figure 27 Falling Speeds of Particles in Water
____
101
101.
/
p
I v -
10
11
r
I —0- Partide Density = 2.85 g!cc
L— --_Partide Density = 3.40 g cc
- A . . . .
I
0.1 1 10 100 1000
45
-------
DRAFT
- . -
_ f __, cJ — - -. V —
‘1 v ,
ijr ____
- _____ I - %
[ .- : • -• . — :.• .
_ _ _ _ _ _ _ _ _ - - — t 1 k . .
1
N 1
- c -r c : ’. . -
- - . -t_ r—- -- - . 4 CJt’ r
4 — - t— 4 -c t - -.- C—
Figure 29
Displacement Sedimentation at Approximately 40%
Completion
Figure 28 Apparatus for Displacement Sedimentation PrOcedure
46
-------
(g) The volume of the aqueous suspension to be filtered depends on either
the particulate concentration or the asbestos fiber concentration. The
volume of the aqueous suspension required to produce an analytical filter
with a suitable particulate or fiber loading for analysis often cannot be
predicted, and it is usually necessary to prepare several analytical filters
corresponding to filtration of different aliquots. The number of grid
openings on the TEM specimens that require examination in order to
achieve a particular analytical sensitivity are shown in Table 2.
Table 2- Examples of the minimum number of grid openings of TEM specimens required to
be examined to achieve a particular analytical sensitivity and limit of detection
Analytical
sensitivity
(10 Fiberslg)
Limit of
detection
(10 Fiberslg)
Volume of Suspension Filtered (ml)
0.01
0.03
0.1
0.3
1
0.1
0.3
551
184
56
19
6
0.2
0.6
276
92
28
10
4
0.3
0.9
184
62
19
7
4
0.4
1.2
138
46
14
5
4
0.5
1.5
111
37
12
4
4
0.7
2.1
79
27
8
4
4
1
3
56
19
6
4
4
2
6
28
10
4
4
4
3
9
19
7
4
4
4
4
12
14
5
4
4
4
5
15
12
4
4
4
4
7
21
8
4
4
4
4
10
30
6
4
4
4
4
NOTES
In Table 1 • it Is assumed that the Initial sample weight was 50 grams. the volume of water used to disperse
the sample Is 1 liter, the active area of the analytical filter is 199 mm 2 . and the TEM grid openings are square
with a linear dimension of 85 pm. The limit of detection is defined as the upper 95% confidence limit of the
Poisson distribution for a count of zero structures. In the absence of background, this is equal to 2.99 times
the analytical sensitivity. Non-zero backgrounds observed during analysis of blank filters will degrade the limit
of detection.
47
-------
(h) The aqueous suspensions are generally not stable; it is therefore
necessary to prepare all analytical filters immediately. Uniform deposits
of particulate on the analytical filters cannot be assured if liquid volumes
smaller than 5 mL are filtered using filtration systems of 199 mm 2 active
area; accordingly, where it is required to filter volumes smaller than
5 mL, the aliquot shall be diluted with freshly-distilled and filtered water
to a volume exceeding 5 mL.
(I) Pour the aliquot of the suspension into the filtration reservoir, and apply
the vacuum. If the volume of the aliquot is larger than the capacity of•
the filtration reservoir, do not allow the level of liquid in the reservoir to
fall below 5 cm depth before the remaining volume is added. Failure to
observe this precaution may result in disturbance of the filtered
particulate and non- uniform deposition.
(j) With the vacuum still applied, unclamp the filtration assembly and
remove the filtration reservoir. Using clean tweezers, remove the
analytical filter and transfer it to a petri-dish. Allow the filter to dry
before placing the cover on the petri-dish.
(k) For the beaker blank, prepare only one analytical filter by filtration of the
entire 40 mL suspension.
(I) After filtration of the aliquots for preparation of TEM specimens, filter the
balance of the suspension through a pre-weighed, 0.4 pm pore size
polycarbonate filter. Dry the filter with the particulate deposit and
re-weigh to determine the weight of respirable particulate material and
fibers. This measurement is needed in order to express the concentration
of respirable fibers relative to the weight of total respirable particulate
material -
NOTES
It Is recommended to prepare several analytical filters from the suspension. I i the particulate or fiber
COnCeflUatiOfl is thought to be such that it is required to fitter an aliquot of lower volume than 1 mL,
use a dilution procedure in whIch 1 ml. of the original suspension is transferred to a clean beaker and
diluted with freshly-distilled water to a total volume of 100 mL. After stirring to ensure complete
mixing, auquots oil ml, 3 mL 10 mL and 30 ml. from this diluted suspension can then be filtered.
corresponding to volumes of 0.01 mL, 0.03 mL . 0,1 mL and 0,3 mL of the original suspension. From
the original dispersion, volumes of 1 mL and 3 mL can also be filtered, giving 6 analytical filters with a
concentration range of a factor of 300. The requirement for washing of the filtration apparatus Is
minimized If the aliquots are filtered in order of increasing concentration.
48
-------
It is beyond the scope of this method to provide detailed instructions for
preparation of TEM specimens from membrane filters; these instructions are
published in ISO 13794. It is recommended that aliquots of the aqueous
suspension of vermiculite be filtered using the method specified in ISO 13794. If
polycarbonate filters are used, they shall be cleaned to remove the asbestos
contamination frequently present on this type of filter (Chatfield, 2000). Prepare
TEM specimens from the filters using the methods specified in ISO 13794.
9.3.5 Examination of TEM Specimens
Criteria for examination of TEM specimens are specified in ISO 10312 and
ISO 13794. For the purpose of deriving risk estimates, only asbestos structures
longer than 5 pm need be considered. The above ISO Standards specify that a
magnification of approximately 10,000 is sufficient for determination of the
concentration of asbestos structures longer than 5 pm. Identify amphiboles
according to the International Mineralogical Association classification (Leake,
1997)
9.4 Combined Procedure
If both the concentration of respirable fibers and the weight percent of
amphibole asbestos are required, follow the full flow chart as specified in
Figure 25, along with the more detailed instructions for determination of the
individual parameters.
10 DATA REPORTING
10.1 Rapid Screening Analysis to Determine Weight Percent of Amphibole
Asbestos
In the test report, all relevant measurements shall be reported, including:
(a) the initial weight of the sub. sample;
(b) the weight loss on drying (if applicable);
(c) the weight loss on ashing (if applicable);
(d) the weight of sediment after water sedimentation;
(e) the weight of centrifugate after centrifugation;
(f) the manner by which asbestos in the centrifugate was quantified; and
one of the following:
49
-------
(1) the weight of handpicked asbestos;
(2) the number of centrifugate particles and the number of asbestos
fiber bundles;
(3) the assumed sensitivity of the chemical balance;
(g) the weight percent of amphibole asbestos in the original sub- sample,
calculated by procedures (1), (2) or (3).
10.2 Concentration of Respirabte Fibers
Report all of the ana yticaI parameters, and from these parameters calculate
the analytical sensitivity S , in structureslgram, using the formula:
(Aa.Vd)
s=
(k.A 9 .V 1 .M 5 )
where: s = Required analytical sensitivity in structures/gram
Aa = Active area of analytical filter in mm 2
Vd = Volume of water in ml used for dispersal of sample
k = Number of grid openings examined
A 9 = Area of TEM specimen grid opening in mm 2
V = Volume of aqueous dispersion filtered in mL
W , = Weight of solid material
Report the lengths, widths and identifications of all respirable fibers detected
that have compositions consistent with any asbestos varieties specified in
Section 4 (Definitions). Using the calculated analytical sensitivity, report the
concentration of respirable fibers per gram of original sub-sample, and per gram of
total respirable particulate material. Also report the 95% confidence interval of
these measurements, and the numbers of fibers on which the measurements are
based.
50
-------
11 ACCURACY AND PRECISION
11.1 Rapid Screening Analysis to Determine Weight Percent of Amphibole
Asbestos
The accuracy of this analysis is limited only by transfer losses during
processing, and by the sensitivity of the laboratory balance. The precision is
limited by the initial size of the sub-sample, and the statistical effects of large
asbestos fiber bundles when there are only small numbers present, or when one or
more asbestos fiber bundles represent a large proportion of the weight of asbestos
detected.
11.2 Concentration of Respirable Fibers
There is no independent method to establish the accuracy of measurements
of the concentration of respirable fibers. The precision of measurements, for
measurements based on water suspensions of fibers, is usually limited by the
Poisson distribution if filtrations are performed using the specified procedures.
Accordingly, the precision can be improved by examination of greater areas of the
TEM specimens in order to collect data on larger numbers of fibers.
12 REFERENCES
Chatfueld, EJ. and Lewis, G.M. (1980): Development and application of an
analytical technique for measurement of asbestos fibers in vermiculite. In:
Scanning Electron Microscopy/i 980/I, (0. Johari, Ed.). SEM Inc., AMF O’Hare,
Chicago, Illinois 60666, U.S.A.
Chatfueld, E.J. (2000): A rapid procedure for preparation of transmission electron
microscopy specimens from polycarbonate filters. ASTM STP 1342, Advances in
Environmental Measurement Methods for Asbestos, Michael E. Beard and Harry 1.
Rook, Eds.,242-249
International Mineralogical Association (1978): Nomenclature of amphiboles
(compiled by B.E. Leake), Canadian Mineralogist, 16:501
International Mineralogical Association (1997): Nomenclature of amphiboles:
Report of the Subcommittee on Amphiboles of the International Mineralogical
Association Commission on New Minerals and Mineral Names (compiled by B.E.
Leake), Mineralogical Magazine, April 1997, Vol 61, pp 295-321
51
-------
International Organization for Standardization (1999): Iso 13794, Ambient air -
Determination of asbestos fibres- Indirect-transfer Transmission electron
microscopy method
52
-------
APPENDIX A. RECOVERY OF HEAVY LIQUIDS FOR RE-USE
Al Background
The heavy liquids used in this analytical method are quite expensive, and for
both economic and environmental reasons they should not be discarded after use.
Accordingly, a method for recovery of these liquids is required.
For use in screening analyses, it is usually sufficient to remove the larger
fragments of material by filtration through a coarse Whatman paper filter, followed
by pressure filtration through a 0.22 pm porosity mixed esters of cellulose filter.
Both liquids are compatible with this type of filter, but compatibility with filter unit
components must be considered before such components are used. The liquid will
acquire a coloration on the first use, but this does not interfere with screening
analyses. If the analytical procedure includes measurement of fine fibers by
transmission electron microscopy, the retention of particles by a 0.22 pm porosity
filter may be insufficient to prevent cross-contamination between samples, and for
these types of analyses the heavy liquid must be purified by distillation.
A.2 Purification of Heavy Liquids by Distillation
The boiling point of l-l-2-2-tetrabromoethane is approximately 243.5°C, but
unfortunately it decomposes at this temperature, liberating bromine. It is therefore
not possible to distill this liquid at the boiling point at atmospheric pressure. If
vacuum distillation equipment is available, l-l-2-2.tetrabromoethane can be
distilled at .reduced atmospheric pressures. The boiling point of tribromoethane
(bromoform) is approximately 150.5°C, and it can be distilled at atmospheric
pressure without decomposition. Table Al shows the vapor pressure of
1-1-2-2-tetrabromoethane and tribromoethane at various temperatures below the
atmospheric pressure boiling point. Suitable distillation conditions can be selected
from Table Al.
A non-boiling still is the optimum method for recovery of the heavy liquids,
because this method of distillation avoids carry-over of solid particles into the
distillate by the spray generated from the boiling liquid during conventional
distillation. A non-boiling still can be operated at a temperature significantly below
the boiling point at atmospheric pressure, but still have an acceptable distillation
rate. Even with good design, the distillation rate is relatively slow, but the
advantage is that the distillate is very pure and contains far fewer particles than
distillates from conventional distillation. Satisfactory distillation rates are obtained
at vapor pressures of 40 mm and higher.
53
-------
Table A. 1 Vapor Pressure - Temperature Data for 1.1-2- 2-Tetrabromoethafle and
Tribromomethafle (Bromoform)
Vapor Pressure, mm
Temperature,_°C
1-1-2- 2 .Tetrabromoethafle
Tribromomethane
1
65
(Solid)
10
110
34
40
144
63.6
100
170
85.9
400
217.5
127.9
150.5
760
243.5
A suitable non-boiling still can be constructed very simply, and a diagram of
one design is illustrated in Figure Al. Liquid in the base of the container is heated
by an annular heater attached to the base. The condenser consists of a 250 mL
glass flask through which there is a circulation of cold water. Vapor from the
heavy liquid condenses on the outer surface of the condenser, and the condensed
liquid falls into a glass funnel which passes through the base of the container. The
power supply to the heater is controlled by a light dimmer. A simple still of this
design can be operated at a temperature of approximately 150°C, and an example
of the construction is shown in Figure A2.
54
-------
I Cooling water supply (in)
2 Coolin water supply (out)
3 Thermometer, 0-200°C
4 Glass flask
5 Distillation chamber
6 Glass funnel
7 Uquld to be distilled
8 Annular heater
9 Power to heater
10 Power to heater
11 TrIpod Stand
12 Collection flask for distillate
Figure Al Example of Design of NonBoiling Still for Recovery of Heavy Uquids
4
5
6
8 -i
11—
—9
55
-------
DRAFT
Figure A2 Example of Non.Boiling Still For
Recovery of Heavy Uquids
56
------- |