Clays and Clay Minerals, Vol. 26, No. 3, pp. 181-188, 1978.

S E M I Q U A N T I T A T I V E C H E M I C A L A N A L Y S I S OF ASBESTOS FIBERS A N D CLAY M I N E R A L S WITH AN A N A L Y T I C A L E L E C T R O N MICROSCOPE

HISATO HAYASHI Research Institute of UndergroundResources, Mining College, Akita University, Akita 010, Japan
SABURO AITA AND MIKIO SUZUKI Application Department, JEOL Ltd., Nakagami, Akishima 196, Japan

(Received 31 May 1977)

Abstract By using a transmission electron microscope equipped with an energy dispersive spectrometer, it Was possible to detect the morphological, structural, and chemical characteristics of individual asbestos fibers and clay minerals without any realignment of the equipment. A rapid and convenient procedure for semiquantitative analysis is proposed. Analyses are given for 21 hydrous silicates, asbestos and clay minerals, by both ordinary chemical and energy dispersive methods. The energy dispersive results were comparable to those obtained by chemical analysis. Application of this procedure to asbestos fibers proved this method to be practical and valid for characterization of asbestos in environmental and tissue samples. Key Words---Amphibole, Asbestos, Chrysotile, Morphology, Structure.

INTRODUCTION Energy dispersive analysis of X-rays in the transmission electron microscope has emerged during the past few years as a highly useful analytical technique in a variety of research fields such as air pollution, pathology, forensic chemistry, and experimental petrology. This technique has enabled the compositional determination of individual microparticles of less than 1 /zmz. Electron microprobe analysis of major elements in bulk samples has been described by a number of authors. The analytical techniques for bulk specimens are inapplicable to X-ray microanalysis of ultrathin specimens in the electron microscope, because the characteristic X-ray intensity depends on the thickness of the specimen as well as on the chemical composition. The effect of the specimen thickness on the characteristic X-ray intensity was eliminated by Marshall and Hall (1968). They showed that the intensity ratio of a characteristic peak to the intensity of the continuum provides a measure for the concentration of an element, which is independent of the film thickness. Literature exists on application of the analytical electron microscope to environmental and tissue samples whose volumes of less than 1/xm 3 (Langer et al., 1971, 1974; Pooley, 1972, 1976; Henderson et al., 1973; Rubin and Maggiore, 1974); but few of these investigations have determined chemical components quantitatively for silicate minerals, especially submicron particles. Asbestos has been studied in the field of occupational health in relation to respiratory disease as well as pleural cancer (Selikoff, 1972). Recently the contamination
Copyright 9 1978, The Clay Minerals Society

of public water supplies by fibrous amphibole minerals was found in tailing discharge of an iron ore mining company located in Minnesota (Nicholson, 1974; Kramer, 1976). In order to monitor these health effects and the extent of environmental contamination by asbestos, the identification and characterization of microparticles have become important fields of study. Asbestos fibers have been intensively studied for these reasons. This study is initiated to use the analytical electron microscope for the semiquantitative determination of the major elements in 21 hydrous silicate minerals which already have been analyzed by ordinary chemical methods. SPECIMENS AND METHOD Specimens used in this study are listed in Table 1. These minerals were confirmed to be pure minerals by X-ray, thermal, and chemical analyses. The analytical electron microscope is composed of the JEM-100 C type transmission electron microscope, EM-SEG side entry goniometer, EM-ASID high resolution scanning attachment, and EM-NDS dispersive X-ray spectrometer. The instrument was operated at 40 kV. The particles of each specimen were deposited from suspension on a supporting film on an electron microscope grid, and then coated with carbon. The isolated particles were selected on the STEM image, and then analyzed using the EM-NDS dispersive X-ray spectrometer.

181

182

Hayashi, Aita, and Suzuki

Clays and Clay Minerals

Fig. 1. Transmission electron micrographs and selected area diffraction patterns obtained on amosite (A), crocidolite (B) and chrysotile (C).

Table 1.

Details of analyzed minerals.

No. 1 2 3 4 5 6 7 8 9 I0 II 12 13 14 15 16 17 18 19 20 21

Minerals Chrysotile Chrysotile Anthophyllite Anthophyllite Amosite Amosite Tremolite Tremolite Actinolite Actinolite Crocidolite Crocidolite Halloysite Pyrophyllite Talc Sepiolite Palygorskite Al-chlorite AI-Mg-chlorite Mg-chlorite Fe-chlorite

Locality Cassier mine, Canada. Yamabe mine, Hokkaido, Japan. South America. Unknown. Penge mine, South Africa. Penge mine, South Africa. Sue-machi, Fukuoka, Japan. lwate, Japan. Sekikawa-mura, Ehime, Japan. Ohgushi, Nagasaki, Japan. Kuruman mine. South Africa. Kuruman mine, South Africa. Kusatsu, Gunma, Japan. Honamimine, Nagano, Japan. Haicheng, China. Kuzuu, Tochigi, Japan. Kuzuu,Tochigi, Japan. Kamikita mine, Aomori, Japan. Hanaoka mine, Akita, Japan. Wanibichi mine, Shimane, Japan. Ichinokoshi, Toyama, Japan.

Ref. No. Cassier AA

A-IO S-22

M J-lO H-2

K-40 H-330

Analyst H a y a s h i (1971) Hayashi (1971) Komuro (1975) Hayashi (1971) Hayashi (1971) Hayashi (1971) Komuro (1975) Komuro (1975) Komuro (1975) Hayashi (1971) Hayashi (1971) Hayashi (1971) Sato (1968) Kodama (1958) Komuro (1974) Nakamura (1966) Nakamura (1968) Hayashi (1964) Hayashi (1961) Sudo (1956) Sudo (1943)

Vol. 26, No. 3, 1978

Analysis of a s b e s t o s fibers

183

Chrysotile

Anthophyllite

Amosite

Tremolite

Crocidolite

Pyrophyl I i t e

Talc

Al-chlorite

Mg-chlorite

Fe-chlorite

Fig. 2. Energy dispersive spectrometer spectra for various asbestos fibers and clay minerals.

EXPERIMENTAL RESULTS

Asbestos and related minerals

The asbestos minerals are natural, hydrated silicates which occur as fibers. Asbestos fibers are classified into two groups: amphiboles and chrysotile. Anthophyllite, amosite, tremolite, actinolite and crocidolite belong to the amphibole mineral in which cations link the infinite chains of composition SisO22(OH)2. As shown in Figure 1, each fiber of amphibole asbestos tends to form straight edges parallel to tbe fiber axis and vary from electron-dense fibers to opaque ones. On the other hand, chrysotile fibrils normally consist of a central capillary and electron opaque walls. Chrysotile yields a unique selected area electron diffraction pattern as shown in Figure 1. L a y e r lines which denote the periodicity of the fiber axis a yield measurements of 5.3 A. Figure 1 also shows two representative selected area electron diffraction patterns of amphibole asbestos which are quite different from that of chrysotile. Amphiboles, however, are quite similar in diffraction patterns (Figure 1). It is difficult to distinguish a particular mineral species among amphiboles, because these diffraction patterns resemble each other. This technique, however, is very effective for differentiating chrysotile from amphibole asbestos. X-ray emission spectra of five asbestiform minerals are shown in Figure 2. Chrysotile shows two large peaks for silicon and magnesium, and a small peak for iron. As can be seen, chrysotile and anthophyllite have similar spectra but they differ in relative amounts of magnesium, silicon and iron. Amosite has two large peaks for silicon and iron, and a small peak for magnesium. Tremolite, on the other hand, differs significantly from the above-mentioned three minerals due to

the presence of a calcium peak. The presence of a sodium peak in crocidolite is one of the most notable differences among the six species of asbestiform minerals. Chrysotile is the trioctahedral analogue of the kaolin minerals and shows certain morphological similarities to halloysite. Halloysite fibers are generally difficult to distinguish from chrysotile fibrils, but X-ray emission spectra can show easily the difference in chemistry of these minerals.

Clay minerals
Talc, pyrophyllite, sepiolite, palygorskite and four kinds of chlorites were examined with the analytical electron microscope. Talc and pyrophyllite were difficult to identify from their morphology or from their electron diffraction patterns. The tetrahedral cation population of talc and pyrophyllite is mainly silicon and the octahedral sheet contains aluminum in pyrophyllite and magnesium in talc. Therefore X-ray energy spectra for talc show a large silicon peak and a peak for magnesium while that for pyrophyllite has a large peak for silicon and an aluminum peak (Figure 2). The sepiotite-palygorskite group is a series of hydrous magnesium minerals. Palygorskite is the aluminum- or iron-rich member and sepiolite is the aluminum-poor member. The electron micrographs of these minerals have revealed that the particles consist of elongated fibers which are axially parallel, sharp-edged and unevenly broken at their ends. The predominant peaks in the energy dispersive X-ray spectra are magnesium and silicon for the sepiolite fiber, and silicon, aluminum and magnesium from the palygorskite fiber. The chlorites are a group of hydrous silicate minerals whose structure permits extensive isomorphous substitution. These minerals vary in chemical composition

184

Hayashi, Aita, and Suzuki

Clays and Clay Minerals

~ o

~ o

~
o o o o

o
c~

~ o
u

~J
x

c~

c~ o=

~o
~ o

~
N O

o
0

~o
~ 0

~
0

~
0

o
0

~ o
~

~
0

~
C~

~
0

3
0

~
0

~
~

o
0

o c; o

C~N

~ o
u

'

~o

=o

~r~

r~ro

~ro

~ o

~q ~ o ~ c~ ~ ~ ~ ~ ~ ~ ~ ~ o o ~ ~ ~ ~ ~ ~ ~ ~

o~ o,

~ c;

~ o

~ 3

~ o

~ o

o c;

~ ~

~ ~

o~ o

o~

o~

rn

u~

co

o~

Vol. 26, No. 3, 1978

Analysis of asbestos fibers

185

70 6o u)

Si02 r = 0.949
o

70 f
6O

5i02

~,
0 C 0

50 4o

(~0~ ~

OODO

ff ~

.~_ =>, so
cl
tO

r=

0,973

'%

40

.u
c

30
2o

0 00~ ~00

._u
E30

0 e-

20

lO o

lO

(~

L ,o

' 20

30

' ~o

' 50

' 6o

7'0

0 o

' ,o

2'0

30

~'

' 5o

60

' 70

E,D. anatysis

(~)

E.O. analysis

Fig. 3. Weightpercent SiO=by energydispersiveanalysisversuschemicalanalysis.(A)Individualvaluesobtainedby energydispersiveanalysis. (B) Average values of individualvalues obtained by energy dispersiveanalysis.

over an extremely wide range. Although four kinds of chlorite having different chemical composition were examined with the transmission electron microscope, it was difficult to differentiate one species from another from their morphology or electron diffraction patterns. Al-chlorite, AI-Mg-chlorite and Mg-chlorite have similar spectra differing only in the relative quantities of magnesium, aluminum, silicon and iron. However, these chlorites may he differentiated from each other on the basis of Mg-A1-Si ratios. Fe-chlorite, on the other hand, differs significantly from the others due to the presence of a magnesium peak and large iron peak.

X-ray intensity provides a measure of the concentration of an element which is independent of the thickness of the specimen. In this study, the relative peak intensity ratio of each element was calculated from the following equation, with the total X-ray intensity in the specified X-ray energy range between 0 keV and 8 keV, as constant. nllNa + n12Mg + "'" + njXj + . . . +na2U = K (constant) where n~ is the ratio of a net peak intensity obtained by subtracting the background intensity, to the total X-ray intensity. X-ray intensities and background intensities were measured more than 5 times for each specimen. The duration of measuring X-ray intensity was 100-200 seconds per position. The numerical value of K varies with each specimen. The value of K in a given specimen was obtained after subtracting the water content [H20 (+) and H20 ( - ) ] from the total obtained by ordinary chemical analysis. Table 2 compares the average peak intensity ratios nj of 21 minerals and the values obtained from chemical analysis. DISCUSSION The characteristic X-ray intensity for a given element in a sample is only roughly proportional to the concentration of the element. The intensity is significantly affected by both instrumental and physical factors. The effect of the instrumental factors include the X-ray detection efficiency and the electron beam potential and current. These effects of instrumental variables can be minimized with the ratio of the measured intensities of the elements in the sample to those in standards of

Semiquantitative determination of chemical compositions

The incident electron beam on a specimen generates characteristic X-rays of elements present in the specimen. The intensity of the generated X-rays is indicated by the following equation: I = KpW; where K is a constant, p is the incident beam intensity, and W is the mass of the element in the irradiated area. In electron microprobe analysis of a thin film, the correction value for X-ray intensity is negligibly small, when the incident electron beam is focusabte on a very small area. The analytical electron microscope is capable of acquiring an X-ray spectrum from particles less than 200/~ in diameter in approximately 100 seconds. Marshall and Hall (1968) reported in their electron microprobe analysis of thin films that the concentration of an element can be obtained by measuring its characteristic X-ray intensity, and simultaneously, the intensity of a portion of the continuous X-ray spectrum. Consequently, the ratio of characteristic X-ray intensity to continuum

186

Hayashi, Aita, and Suzuki

Clays and Clay Minerals

A1203
50 = ,

Total iron
so r= 0 , 9 7 5
r = 0, 9 8 4

MgO

,~_
-->'/*0 o o

.m 50

9 ~o ~~ 30
.u 20 o cO 10 o

40
t30

--~ 30
|

20 10

.R E 2O c10 ~o o

o o

0 10 20 30 40 50

E.D, analysis
Fig. 4.

10

20

30

/*0

;o

,o

3o

/*o

so

E, D. analysis

E.D, analysis

Weight percent A1203(A), total iron (B) and MgO (C) by energy dispersiveanalysis versus chemicalanalysis. compared to the bulk specimen. Furthermore they are transparent to most of the primary X-rays produced by incident electron beam. Hence the matrix effects on characteristic X-rays (atomic number effect, absorption effect and fluorescence effect) can be neglected. In this study the ratio of the characteristic X-ray intensity of an element to the sum of the intensities of all of the elements in the sample is taken as a measure of concentration of that element. The SiO2 content calculated this way is plotted against the value obtained by chemical analysis in Figure 3A. The average value of individual SiOz content for each specimen is plotted in Figure 3B, in relation to the chemical values. Highly significant regression was obtained in both cases; Y = 0.90X + 2.685 Y = 0.95X + 0.241 from Figure 3A from Figure 3B

known composition. The physical factors are dependent on the sample composition and include the atomic number, absorption and secondary fluorescent effects. Correction for the atomic number, absorption and fluorescence effects have been developed and are widely used for thick polished specimens. Recently Waldl et al. (1975) described an empirical method for energy dispersive X-ray microanalysis of ternary and multicomponent alloy systems. They have introduced the following intensity reduction to eliminate t h e influence of instrumental fluctuations.
K i r e d __

Ii _
It

Ii
"~i I t +

Ib

where Ii is the intensity of the X-ray peak of an element i, It is the total measured intensity and Ib is the background intensity. In the case of very thin specimens, as used for the transmission electron microscope, however, the analytical technique for bulk specimens cannot be employed simply. Several techniques have been developed primarily for the quantitative determination of elemental concentration in ultrathin specimens (Marshall and Hall, 1968; Namae, 1975; Chandler and Morton, 1976). Marshall and Hall (1968) showed the concentration of an element could be determined by the ratio of the intensity of the characteristic line to that of the continuous radiation from the same analyzing point. The method proposed by Namae (1975) is simple and useful for a very thin specimen, without knowing the thickness of the specimens, when the constant K of an element is already determined. However, it is difficult to prepare the specimens in the same thickness as the standard, and to determine the constant K of each element in specimens. Since the specimens used in this study are thin enough for use with a conventional 100 kV transmission electron microscope, the activated volume by the incident electron beam is v e r y small

These regressions have correlation constants of r = 0.949 and r = 0.973, respectively. Figures 4A,B and C show the energy dispersive results for A1203, total iron and MgO versus the values obtained by chemical analysis. A linear regression analysis calculated for A1203, total iron and MgO gave the following results. Correlation coefficients are higher than 0.97. A1203: Y = 1 . 0 6 X - 1.35 Total iron: Y = 1.17X + 0.03 MgO: Y = 1.21X - 2.46 r = 0.981 r = 0.975 r = 0.984

There is a little lower energy dispersive value for total iron and MgO compared to the chemical data. Generally energy dispersive results are comparable with those obtained by chemical analysis. The procedure described here can omit the process of estimating thickness of each specimen and determining the constant K for each element in the specimens. Furthermore, using this method a good correlation is obtained between the energy dispersive and chemical determination for SiO~, A120~, total iron and

Vol. 26, No. 3, 1978

Analysis of asbestos fibers

0 Chrysotile zx T r e m o l i t e 9 Actino/ite 9 Croc/dolite

187

CQO~ NQ20+K20

Anthophy//ite

[] Arnosite

CQO,AI203, NQ20* K20

~.

mm

(A)

NIgO

--o

v 5O

v
9 Actinolife 9 Crocido/ite X Talc

r.-o-tx~--tnHD~o

Tatar

Fe+

MnO

0 Chry$otile 9 Anthophyl/ite
C a O . AI203 Na20,K20

[] Amosite a Tremolite

CQO*AI203, NQ20, K.20

z~

ffi

9 9

9 9 9

(B)
9 QO 9 9 D
-/.. XO O CO 9 9 9

MgO

~:zl~o-~x 9

v 50

TotolFe. MnO

Fig. 5. Diagrams of three component composition fields for asbestos and talc. (A) Energy dispersive results of this study. (B) Chemical analytical data quoted from several researchers. MgO, which are major c o m p o n e n t s of silicate minerals. The principal feature of this m e t h o d lies in its rapidity and the good results. H e n c e , application of this m e t h o d is considered to be eminently satisfactory for m a n y purposes, such as characterization of asbestos in environmental and tissue samples. ACKNOWLEDGMENTS The authors wish to express their thanks to Dr. H. Shimane for his technical assistance and helpful advice and to Mr. H. Y o t s u m o t o for his valuable suggestions and encouragement. This w o r k was supported in part by a Grant-in-Aid for Cancer R e s e a r c h f r o m the Ministry of Health and Welfare. REFERENCES Chandler, J. A. and Morton, M. S. (1976) Determination of elemental area concentration in ultrathin specimen by X-ray microanalysis and atomic absorption spectrometry: Anal. Chem. 48, 1316--1318. Henderson, W. J., Chandler, J. A., Blundell, G., Griffiths, C. and Dav-

188

Hayashi, Aita, and Suzuki

Clays a n d Clay Minerals

ies, J. (1973) The application of analytical electron microscopy to the study of diseased biological tissue: J. Microsc. 99, 183-192. Kramer, J. R. (1976) Fibrous cummingtonite in Lake Superior: Can. Mineral. 14, 91-98. Langer, A. M., Selikoff, I. J. and Sastre, A. (1971) Chrysotile asbestos in the lungs of persons in New York City: Arch. Environ. Health 22, 348-361. Langer, A. M., Mackler, A. D. and Pooley, F. D. (1974) Electron microscopical investigation of asbestos fibers: Environ. Health Perspect. 9, 63-80. Marshall, D. J, and Hall, T. A. (1968) Electron-probe X-ray microanalysis of thin films: Br. J. Appl. Phys. Ser. 2, 1, 1651-1656. Namae, T. (1975) A method of quantitative analysis for thin specimens by energy dispersive spectrometer fitted to transmission electron microscope: J. Electron Microsc. 24, 1-6.

Nicholson, W. J. (1974) Analysis of amphibole asbestiform fibers in municipal water supplies: Environ. Health Perspect. 9, 165-172. Pooley, F. D. (1972) Electron microscope characteristics of inhaled chrysotile asbestos fibre: Br. J. Ind. Med. 29, 146-153. Pooley, F. D. (1976) An examination of the fibrous mineral content of asbestos lung tissue from the Canadian chrysotile mining industry: Environ. Res. 12, 281-298. Rubin, I. B. and Maggiore, C. J. (1974) Elemental analysis of asbestos fibers by means of electron probe technique: Environ. Health Perspect. 9, 81-94. Selikoff, I. J. (1972) Carcinogenicity of amosite asbestos: Arch. Environ. Health 25, 183-188. Waldl, E., Walfermann, H., Ruscovic, N. and Warlimont, H. (1975) High-accuracy, empirical method of quantitative electron microprobe analysis using an energy dispersive system: Anal. Chem. 47, 1017-1019.

PesmMe- C nOMO~bm TpaHCMHCCHOHHOFO 9~eKTpOHHOrO MHKpocKona,o6opy~oBaHHOrO pacceHBa~HM 9HeprHD cneKTpOMeTpOMrCTa~O BOSMO~HbD4 o H p e ~ e n H T b M O p ~ o n o r H q e c K M e t c T p y K T y p H ~ e H X H M M q e c K H e x a p a K T e p M C T H K H OT~e~BHBIX a c 6 e C T O B ~ X B O ~ O K O H H r n M H H C T ~ X M M H e p a ~ o B 6e3 n O B T O p H O ~ ~ e H T p O B K H O ~ O p y ~ o B a H H H . H p H B O ~ T C H ~aHH~e a H a ~ H s o B 21 BO~HBIX C M ~ H K a T O B w a c 6 e C T O B H r~HHMCThIX M M H e p a ~ o B r n o ~ y q e H H ~ X C ~ O M O ~ S D O6~qHblX X H M H q e c K H x H P a c c e H B a D ~ H X 9 H e p r H D MeTO~OB. P e s y ~ b T a T ~ HCH O ~ B S O B a H M H p a c c e H B a D ~ H X 9 H e p r M ~ M e T O ~ O B 6biUM C O H O C T O B H M ~ C p e s y ~ B T a T a M M XHMHqecKMx aHa~MSOB.~PMMeHeHHe 9TO~ ~po~e~yp~ ~H M s y q e H H ~ ac6ecToBbD~ BO~O-KOH ~ o K a s a ~ o ero ~ p a K T M q H O C T 5 M ~ p H F O ~ H O C T ~ ~ H Ho~yqeHH~ xapaKTepMCTHKH a c 6 e C T O B B ~pHpo~Hh~X H T K a H e B ~ X 0 6 p a s ~ a x .

K u r z r e f e r a t - I n d e m ein T r a n s m i s s i o n - E l e k t r o n e n - M i k r o s k o p , ausgestattet mit einem Energie-Streuungsspektrometer, b e n u t z t wurde, war es m6glich, die m o r -phologischen, s t r u k t u r e l l e n und c h e m i s c h e n E i g e n s c h a f t e n v o n e i n z e l n e n Asb e s t f a s e r n und T o n m i n e r a l i e n zu ermitteln, ohne die A p p a r a t u r w i e d e r zu eichen. E i n e s c h n e l l e und b e q u e m e M e t h o d e fur s e m i - q u a n t i t a t i v e A n a l y s e w i r d vogeschlagen. Analysenresultate, e r h a l t e n d u t c h a l l g e m e i n e , c h e m i s c h e und Energie-Streuungsmethoden, fur 21 w a s s e r h a l t i g e Silikate, A s b e s t und T o n m i n e r a l i e n sind angegeben. Die R e s u l t a t e der E n e r g i e - S t r e u u n g s m e t h o d e waren v e r g l e i c h b a r m i t denen, e r h a l t e n d u r c h c h e m i s c h e Analyse. A n w e n d u n g d i e s e r P r o z e d u r auf A s b e s t f a s e r n , hat b e w i e s e n , d a B d i e s e M e t h o d e p r a k t i s c h und b r a u c h b a r fur die C h a r a k t e r i s i e r u n g v o n A s b e s t in U m w e l t - u n d G e w e b e p r o b e n ist.

R~sum~-Par l'emploi d'un microscope ~lectronlque ~ t r a n s m i s s i o n ~quipp~ d' un spectrom~tre dispersant l'~nergie,il a ~t~ possible de d~tec~er les caract~ristiques m o r p h o l o g i q u e s , s t r u c t u r a l e s et chimiques de fibres d'asbeste individuelles et de m i n ~ r a u x argileux sans r~aligner l'~quipement.Un proc~d~ rapide et commode pour l'analyse semi-quantitative est propose. Des analyses sont donn~es pour 21 silicates h y d r a t ~ e s , a s b e s t e s et m i n e r a u x argileux,~ la fois par des m~thodes chimiques ordinaires et des m~thodes dispersant l'~nergie.Les r~sultats des m ~ t h o d e s dispersant l'~nergie sont comparables ~ ceux obtenus par analyse chimique. L ' a p p l i c a t i o n de ce proc~d~ aux fibres d'asbeste prouvent que cette m~thode est pratique et valable pour la c a r a c t ~ r i s a t i o n d'asbeste dans les ~chantillons de tissu et du m i l i e u ambiant.