Tobacco Documents Online

CS National Bureau of Standards Special Publication 506. Proceedings of the Workshop on Asbestos: Definitions and Measurement Methods held at NBS, Gaithersburg, MD, July 18-20, 1977. (Issued November 1g7B) ENVIRONMENTAL PROTECTION AGENCY INTERIM METHOD FOR DETERMINING ASBESTOS IN WATER Charles H. Anderson Environmental Research Laboratory U.S. EnVironmental Protection Agency Athens, Georgia 30601 Abstract The discovery of asbestos and asbestiform minerals in water supplies and drinking water has resulted in the requirement for a reliable analytical method. In order to meet this requirement, an interim method, based upon the state-of-the-art in asbestos analytical methodology, has been prepared. In this paper, the broad elements of the method are set forth and discussed. Key Words; Analytical Chemistry; asbestos; environmental pollutants; water. Introduction Environmental concern following the discovery of asbestos and asbestiform minerals in water supplies and drinking water has resulted in a broad range of activities within the Environmental Protection Agency to improve detection sensitivity and to delineate human exposure and subsequent health effects. An important initial step is the development of a reliable analytical method for determining asbestos in water. Based upon the premise that a method should reflect the state-of-the-art of asbestos analytical methodology, an interim procedure has been written. As such, it is a working document subject to subsequent revi- sion and validation. The method relies on previously published work [1•5]1 together with the work that has been carried out at the Environmental Protection Agency's laboratories at Duluth, MN, Athens, GA, and Cincinnati, OH. In this paper, the broad elements of the method and a discussion of the rationale for some of the decisions that were made when choosing between alternatives is presented. The basic features of the method are summarized in Table 1; the complete, detailed method is available upon request from the author. 'Figures in brackets indicate the literature references at the end of this paper. Preceding page rlank 365

Table 1. Summary of EPA interim method for asbestos in water. Definition: Chrysotile - A magnesium silicate, the fibrous form of serpentine, . possessing a layered, helical cylindrical structure. Amphibole - A silicate mineral whose basic structural unit is a double silica chain, of variable composition, and layered structure. Fiber - A particle in the micron size range possessing parallel sides and a length/width ratio of greater than or equal to 3:1. Instrumentation: Transmission Electron Microscope capable of selected area diffraction. Sample: One liter of water. Sample Preparation: Filter sample through .7 um Nuclepore or .22 um Millipore using sample volume 50-500 ml. Maximum of 20 ug/cm2 total particulate. High organic requires low temperature ashing and resuspension by mild ultrasonification. Portion of Millipore placed on TEM grid, dissolve by condensation washing or, carbon coat Nuclepore, dissolve by Jaffe Wick in chloroform. TEM Examination: At 10,000-20,000 magnification. Count 100 fibers or 20 grid squares. Use field of view method if greater than 50 fibers/grid square. Identification: Chrysotile on the basis of morphology and SAED. Amphibote on the basis of morphology and SAED. Reporting: Confirmed chrysotile and amphibole fibers in MFL (million fibers/liter) Mass/liter Distribution by length, width, and aspect ratio Definition of Asbestos Before any quantitative analytical procedure could be outlined, it was obvious that the term asbestos required a definition in terms of measurable chemical and physical parameters. Of the two broad classes of asbestos, chrysotile is readily defined on the basis of 9ts unique morphology, crystalline structure, and elemental composition. Amphibole's characterization, on the other hand, is not so straightforward. The broad class of amphiboles can be defined as silicate minerals whose basic structural unit is a double silica chain, a fibrous morphology, and elemental composition corresponding to the recognized amphibole asbestos types. In the EPA method, amphibole asbestos determination is based on crystal structure, amphibole morphology, and a fiber aspect ratio of 3:1 or greater. The basis for this fiber aspect ratio is conservative and reflects the state-of- the-art in asbestos analytical methods. Although this aspect ratio is lower than that proposed by Ampian [6], it would seem that the ultimate test, insofar as environmental samples are concerned, lies in the health effects of mineral fibers of different size and aspect ratio. Although health effects data will prove difficult to obtain, it seemed prudent to use this more conservative approach. 366

The Environmental Sample As the EPA interim method would be applied to a variety of pollution sources and used for a variety of purposes, no attempt was made to furnish specific sampling instructions. Instead, only guidelines and precautions were included in the method. Asbestos is, in fact, a special type of particulate matter exhibiting a range of particle sizes, and a vertiaat distribution of asbestos concentrations may be present in a water supply. For example, Cook [7] has documented the variability of amphibole fiber concentrations in Lake Superior at the Duluth water supply intake and demonstrated that the amphibole fiber concentrations were dependent on the presence or absence of an ice cover on the lake, the direction and velocity of winds, and the depth of the thermocline. It is important, therefore, to plan a sampling program for a particular purpose and to use the results only in context of the sampling procedure. Another analysis issue was whether the sample taken in the field should be filtered with the filter and its deposited particulates sent to the laboratory; or, whether the entire water sample should be collected and furnished to the analytical laboratory. Although each approach has advantages, it was considered that the possibility of contamination, the potential for loss from the filter paper, and the general lack of control of the filtration step were overriding disadvantages of filtration in the field. Collection of a sample of the water itself was therefore suggested as the better alternative. The Analytical Approach Choice of Instrumentation In broad terms, the approach to the determination of asbestos in water uses preconcentration by filtration followed by direct microscopic identification and measurement of the asbestos fibers. Because asbestos fiber diameters are below the range of optical microscopy techniques, electron microscopic methods must be employed. Although scanning electron microscopy (SEM) has been suggested to be applicable [8], those laboratories that have compared transmission electron microscopy (TEM) with SEM have concluded that TEM is the superior tooi [1,4,9]. TEM allows examination at low (ti200x) and high (~20,000x) magnification and gives excellent brightness and contrast. Furthermore, most modern TEM instruments readily allow selected area electron diffraction (SAED) to be carried out on individual fibers; such capability allows a positive identification of the characteristic crystalline structure of chrysotile and amphiboles. An energy dispersive x-ray (EDX) detector is adaptable to the newer TEM's and can furnish additional information on the elemental composition of individual fibers that are under examination, but its use was not required in the EPA-proposed method. Preparation of Samples The analytical sample, as received by a laboratory, will consist of a 1-liter poly- ethylene bottle containing a representative sample from the environmental source. The objective of preparing the subsample and subsequent microscopic sample is to transfer the asbestos particles from the environmental source to the TEM with a minimum loss. At the same time the particle size, shape, and size distribution in the original sample should be maintained. Furthermore, the TEM sample must allow the examination of single asbestos fibers with no overlapping or obscuration by extraneous material. The initial step in the sample preparation is the filtration of a known volume of the water sample containing the suspended particles of asbestos onto a membrane filter. This filtering is a critical step whose function is not only to separate, but also to uniformly distribute the particulate matter with minimum of overlap. Some precautions are therefore necessary in this procedure. The liquid sample is agitated in a low-power ultrasonic bath prior to filtration to ensure homogeneity. A fixed volume, ranging from 50-500 mL, is added to a vacuum filtration apparatus containing a 0.1-pm Nuclepore or a 0.22-{im Millipore filter. The volume is determined by the amount of particulate matter present, 367 2063105158

and the maximum loading that can be tolerated is 20 pg/cm, or about 200 pg on a 47-mm filter. The applied vacuum should be sufficient for filtration but gentle enough to avoid the formation of a vortex. Once the filtration has been initiated, no additional water should be added nor should the sides of the funnel be rinsed. If the sample contains a substantial amount of organic material, a preliminary, ashing step is required, followed by resuspension and filtration. Low-temperature ashing in an oxygen plasma with resuspension in a fixed volume of water followed by mild ultrasonification has been found to be satisfactory. Preparation of TEM Specimen The transfer of a part of the filter on which the particulates have been deposited on the TEN grid and the subsequent elimination (by dissolution) of the filter material so that a TEM examination can be accomplished is probably the most critical step in the analysis procedure. As the examination in the TEM and subsequent calculations assumes a random orientation and little or no loss of particles, it is essential that the transfer be carried out not only without losing particles, but also with a minimum of movement. This goal becomes very difficult to achieve, largely because the asbestos fibers are in the colloidal size range; movement apparently can take place very easily. Two approaches acceptable for TEM sample preparation are: a. The condensation washer method, which is used when a Millipore filter is employed. b. The Jaffe Wick method, which is used with a Nuclepore filter. In the condensation washing technique [1,3], acetone vapors are condensed in a special reflux condenser at the position just below the TEM grids. Successful operation requires the delicate introduction of sufficient vapor to dissolve the filter in a reasonable time but not enough to cause pooling, movement, or wash-off of the deposited fibers. As a result, close control of bath temperature, cooling water temperature, and flow is required. McCrone [1] and Lishka, et al. [3] claim successful results with this procedure. Beaman [4] in a detailed study of the condensation washing technique, found, under his experimental conditions, amphibole fiber losses ranging from 37 to 60 percent. Chrysotile fibers apparently are less mobile, for Beaman found losses ranging from 0 to 21 percent. In spite of the criticisms of the condensation washer, the fact that at least two laboratories obtained successful results dictated the inclusion of the method as an alternative preparation step in the EPA procedure. In the Nuclepore-Jaffe Wick technique, the Nuclepore filter is carbon-coated in a vacuum evaporator (after filtration) before attempting to dissolve the filter material from the grid. Fixed by the carbon coating, the particles are thereby rendered immobile and less susceptible to loss. The filter material is dissolved away by a simple wicking action that can be obtained from several layers of filter paper in a covered Petri dish containing chloroform. The dissolving time, although longer than that for the condensation washer, can usually be accomplished overnight. The Nuclepore filter is well adapted to carbon coating because it has a flat surface and no disturbing, replicated structure is found in the grid film. In contrast, the Millipore filter contains a fibrous-like structure that, when replicated, interferes with the TEN examination. Cook [5] at the Duluth Environmental Research Laboratory, Nicholson [2] at Mt. Sinai, Glass [10] at Ontario Research Foundation, and chemists at our laboratory have all obtained excellent results with the Jaffe Wick preparation method. An advantage of this method is that if a fiber is lost during the dissolving step a replica of the fiber remains; thus, an internal check on the procedure is preserved. The fact that such fiber replicas are rarely if ever observed gives substance to the conclusion that no significant loss or movement takes place during the preparation process. Counting of Fibers The prepared TEN grid holding the asbestos fibers and other particulate matter is initially examined at low magnification (300x-1000x) in order to determine whether the 368

es grid preparation has been prepared satisfactorily. If the grid is too heavily loaded (>300 fibers/grid square), if the distribution is noticeably uneven, or if a majority of the grid squares have broken carbon films, a new preparation is required. For those natural waters that contain sufficient organic matter to obscure other particulates, the filtered material must be subjected to low temperature ashing, resuspension, and filtration. The analytical procedure employs standard counting techniques at 10,000-20,000x in determining the number, dimensions, and type of asbestos fibers that are present in the area that is examined. Two general approaches-random search or systematic search-were suggested for the EPA method depending on the number of fibers present. If an 80 pm x 80 pm grid square contains more than about 50-100 fibers, it is conve- nient to use the field of view method. Beaman [4] and chemists at the Athens Laboratory have found this method satisfactory for these situations. In this method, several grid squares are selected and random fields of view examined. The area of the field is known from the magnification of the microscope and the area of the projected image. The total fibers counted in the known number of fields of the known area can be then converted to million of fibers per liter (MFL) through a simple conversion factor that is dependent on the original filter diameter and the amount filtered. If only a few fibers are found in each grid square, it is more convenient to system- atically search up to ten whole grid squares and count the fibers lying within these areas. As the area of individual grid squares may vary by -10 percent, the dimensions of each grid square examined should be recorded. Ideally, 100 fibers are examined for each sample, 50 each from two grid preparations. In practice, however, some samples may contain so few fibers that considerations of time become important. In the EPA method, ten grid squares on two grid preparations are examined, and the number of fibers in this fixed area are counted when the fiber concentrations are quite low. Identification of Fibers Each fiber that is found should be subjected to further examination to determine whether it is asbestos and classified as chrysotile or an amphibole type. Chrysotile's unique tubular structure and its tendency to form bundles of single fibers makes it readily identifiable. For an unequivocal identification, however, a selected area electron diffraction (SAED) pattern of chrysotile gives a unique pattern exhibiting prominent streaks on the first layer line and a triple set of double spots on the second layer line. UICC standard asbestos fiber material is available to furnish standard comparison diffraction patterns. Amphibole fibers are identified on the basis of lath-like morphology, aspect ratio, and an SAED pattern. Although it would be desirable to identify the different amphibole asbestos types, their diffraction patterns are almost identical and their differentiation by SAED is almost impossible and clearly impractical. Amphibole identification is more difficult than chrysotile because the amphibole SAED does not have the unique characteris- tics of the chrysotile pattern and requires some judgement in interpreting the SAED pattern. Some amphibole fibers show only partial patterns that are not sufficiently complete to allow positive identification; these are classified as "probably" amphiboles. As Beaman [4] and Millette [11] have indicated, it is useful to determine the elemental composition of a fiber as an aid to identification. This is particularly true if a fiber fails to give an identifiable electron diffraction pattern and additional information is required for identification. Because the fiber width and thickness is less than that excited by the electron beam, the elemental x-ray intensities are a function of width. This variation with particle size can be partially overcome, however, by determining x-ray intensity ratios. But these ratios, because of differential absorption, are also a function of particle size. Because of the difficulty of specifying quantitative procedures based upon x-ray intensities, the EPA method suggests the use of energy dispersive x-ray analysis as a useful tool but does not require its use. As Ruud [12] has pointed out, even though a good quantitative analysis could be obtained from EDX, it should not be considered a definitive identification without an SAED pattern. 369 2063105160

The length and width of each fiber positively identified, as well as the "probables", are recorded. Precision of Analysis The analysis precision obtained within an individual laboratory is dependent upon the number of fibers counted. If 100 fibers are counted and the loading is at least 3.5 fibers/grid square, computer modeling of the counting errors shows that a relative standard deviation of only about 10 percent can be expected. In actual practice, some degradation from this precision will be observed but should not exceed ±20 percent if several grids are prepared from the same filtered sample. The relative standard deviation of analyses of the same water sample in the same laboratory will increase because of sample preparation errors, and a relative standard deviation of about ±20-30 percent can be expected. Table 2 shows the results obtained on five sets of samples of asbestos and indicates that this range can be achieved. As the number of fibers counted decreases, the precision will also decrease approximately propor- tional to N" where N is the number of fibers counted. Table 2. Precision of C-coated Nuclepore method. Type No. Ave. Standard Coefficient asbestos samples conc. (MFL) deviation of variation Chrysotile 10 23 4.7 23% Crocidolite 9 13 1.7 13% Crocidolite 10 16 2.8 17% "Taconite" 10 21 5.0 24% "Taconite" 10 28 3.4 12% Average 18% Although there have been a number of interlaboratory testing programs, few of these have been carried out using the same procedure. Those that have been done indicate that agreement within a factor of two is achieved if 100 fibers can be counted. Results obtained among three laboratories at different locations within the Environmental Protection Agency are given in Table 3. Although these data are insufficient for statistical purposes, they indicate the analysis capability obtainable at the present time. 370

S-S Table 3. Comparison of results-Nuclepore method (except as noted) positively identified fibers (MFL). Sample Asbestos type Lab A Lab B Lab C 1 Amphibole 137 150 --- 2 Amphibole 86 92 70a 3 Amphibole 130 220 140 13a 120a 4 Amphibole 44 58 58 17a 48a 5 Chrysotile 29 14 --- 17a 6 Chrysotile 66 58 60 56a 5oa a Condensation Washer. Summary The Environmental Protection Agency has written an analytical method for asbestos in water, based on what was considered to represent the state-of-the-art asbestos analytical methodology. In its present form, the method should be considered as an interim method having no official status. When the results of future research efforts and cooperative testing are available, it is expected to be proposed as a referee method for asbestos. References [1] McCrone, W. C. and Stewart, I. M., Asbestos, Amer. Lab. 6(4), 10-18 (1974). [2] Nicholson, W. J., Analysis of amphibole asbestiform fibers in municipal water supplies, Env. Health Perspectives 9, 1965-172 (1974). [3] Lishka, R. G., Millette, J. R., and McFarren, E. F., Asbestos Analysis by Electron Microscope in Proc. AWMA Water ualit Tech. Conf. American Water Works Assoc., Denver, C0, XIV=1 - XIV-12 (1975). [4] Beaman, D. R. and File, D. M., Quantitative determination of asbestos fiber concen- trations, Anal. Chem. 48, 101-110 (1976). [5] Cook, P. M., Rubin, I. B., Maggiore, C. J., and Nicholson, W. J., X-ray Diffraction and Electron Beam Analysis of Asbestiform Minerals in Lake Superior Waters in Proc. Inter. Conf. on Environ. Sensing and Assessment 34, Las Vegas, NV, 1-9 (1976). [6] Ampian, S., Asbestos Minerals and Their Nonasbestos Analogs, paper presented at Electron Microscopy of Nicrofibers Symposium, Pennsylvania State University, University Park, PA, August 23-25, 1976. [7] Cook, P. M., Glass, G. E. , and Tucker, J. H., Asbestiform amphibole minerals: detection and measurement of high concentrations in municipal water supplies, Science 185, 853-855 (1974). 371 2063105162

[8] Pattnaik, A. and Meakin, J. 0., Development of an Instrumental Monitoring Method for Measurement of Asbestos Concentrations in or Near Sources, (U.S. Environmental Protection Agency, W~gton,C, EPA 650/2=73-016, 1973). [9] Flickinger, J. and Standridge, J., Identification of fibrous material in two public water supplies, Environ. Sci. and Tech. 10, 1028-32 (1976). [10] Glass, R. W., Improved Methodology for Determination of Asbestos as a Water Pollutant, Ontario Research Foundation Report, April 30, 1976, Mississauga, Ontario, Canada. [11] Millette, J. R. and McFarren, E. F., Eds of waterborne asbestos fibers in TEM, SEM, and STEM, Scanning Electron Microscopy/1976 (Part III) 451-460 (1976). [12] Ruud, C. 0., Barrett, C. S., Russell, P. A., and Clark, R. L., Selected area electron diffraction and energy dispersive X-Ray analysis for the identification of asbestos fibers, a comparison, Micron 7, 115-132 (1976). Discussion R. LEE: I noticed in your description of the method that you rely on chrysotile which has a selected area diffraction, and has morphology. For amphiboles you rely on morphology plus selected area diffraction or chemistry. If you accept that as your definition of an asbestos particle, I think it is very important to know whether or not what you are telling me is that now I have to treat any cleavage fragment, any massive hand specimen which I grind down, in which there should be a more morphological and orien- tation difference, as an asbestos particle. Secondly I'd like to say that, before you answer, that we're going to show some preliminary data that suggest that we can give you a very close diagnostic method for distinguishing between them. C. ANDERSON: This is not my idea of what should be done or should'nt be done, this is our concept of the consensus of the state of the art of analytical methodology in asbestos as it existed when we wrote the method. The state of the analytical methodology for amphiboles is just very, very muddy. We certainly are willing to listen to your suggestions as to how we can do this better. LEE: Is there any reason to assume that all amphibole cleavage fragments are identical to amosite asbestos? ANDERSON: I think that the critical issue is what are the health effects of one versus the other. LEE: The only data we have seen on that to date was shown yesterday, indicating that grunerite had no cellular activity. ANDERSON: I saw some slides showing almost any particle has some in vitro effects. LEE: In this particular case grunerite (the non-fibrous variety) did not show any activity. ANDERSON: What were the particle characteristics of the grunerite? A. WILEY: I suggest that you change your title. Rather than identifying asbestos say that you are identifying chrysotile and amphibole. Since you can't say that it is asbestos, why not say just amphibole, period. V. WOLKODOFF: I notice in your paper, for five fibers, you would say statistically significant, and anything less than that would be not statistically significant. Do you- still hold to that? I 372

r ANDERSON: The five-fiber criterion was considered to be the state of the art, and I was happy to see Dr. Leineweber point out that at five fibers the statistics show you the range is between .48 and 10. It seems to me that five fibers is statistically significant to indicate asbestos is present. WOLKDDOFF: If you go by the Poisson distribution. But there are cases where less than five fibers is extremely important in the interpretation of particular problems to us, providing our background is zero. ANDERSON: You apply the statistics to your problems in the context of what you are worrying about. What we did was, if you find less than five in the water samples you really can't say with much confidence how much is there. WOLKODOFF: I'm glad to hear you say that and really it's a big help then. On this business of hornblende, is this your offical stance or posture that these are not to be counted? ANDERSON: I can't take any official stance. I claim in the method you will mis- identify hornblende as an amphibole asbestos. If the mineralogists want to take issue with me, let me know. We will take that out. WOLKODOFF: Have you gone into this as a subject? ANDERSON: You mean as far as differentiating various types? WOLKODOFF: Of the various types, yes. ANDERSON: No. WOLKODOFF: As far as you are concerned then, a hornblende is a hornblende. I mean, an amphibole is an amphibole. ANDERSON: Right. WOLKODOFF: Your people, like Milette and Cook and yourself, can you actually differentiate amphiboles by selected area electron diffraction? Have you gone into this subject? ANDERSON: As differentiate types, no. WOLKODOFF: As far as you are concerned, an amphibole is an amphibole. ANDERSON: Right. WOLKODOFF: I thought then perhaps that when you say that EDS is not absolutely necessary, that maybe there was a matter of cost reduction, but you are saying that for technical reasons, very much like Don Beaman pointed out. ANDERSON: Look at this from my point of view. Suppose I say, Valdimir write a method that everybody agrees with and put it down specifically enough so that people can follow it. How do you do this with an EDS system? I don't know. I don't know that much; I strongly recommend using it, but I was not really very comfortable in just saying use the EDS like the manufacturer said to use it. WOLKODOFF: For many of our problems it would be of great benefit. ANDERSON: And, of course, there is a cost consideration involved here too. WOLKODOFF: I must commend and compliment you on your paper. We _felt it was very well done, and I think with some additions and so forth it will..... ANDERSON: Thank you. 373 2063105164

e BEAMAN: I'd like to mention the EDS,. Charles, again I think there have been presented at this meeting and last year's some very serious challenges to the use of selected area electron diffraction identification and classification of amphiboles. I've heard people say it was almost impossible to classify cleavage fragments as an amphibole looking at the selected area electron diffraction pattern on the screen of the TEM. You may be able to classiky by taking a photograph and indexing it, but I think that in conjunction with the energy dispersive spectrometry you are on much firmer ground. ANDERSON: I agree, but I think what you have to remember is also the purpose of the method that we wrote. We wrote it from the point of view of not giving a complete characterization of the particulate matter that was in a water source. We consider that to be a little bit beyond the scope of an analytical method. There is a fine distinction between a very quick and dirty method and a research method in which you are characterizing the whole source and an analytical method that the broad analytical laboratory might want to use. BEAMAN: If you were going to use just the SAED, then you have to put some confidence limits on it. The numbers that you present in an interlaboratory comparison, for example, you would have to put a range on there and say that those 60 or 150 could be as low as 5 or 10 if you were to make a positive identification. J. MCALEAR: NBS Associates, I'm going to have to speak in behalf of some of the scores of laboratories who have been doing some scanning electron microscope analysis for asbestos for some years to make the point that the actual application in this area is fairly extensive using the SEM, and I think it is growing. I'm not going to take time now to make a detailed comparison here. It has been done at many places, but I think it is a very poor mistake to rule out scanning electron microscopy in a general, even interim method when such things have a tendency in fact to become regulations; become standards. I think that this needs to be objectively reviewed. ANDERSON: Let me respond to that. I came into this program having little experience in transmission electron microscopy. My major experience was with scanning electron microscopes, wavelength electron probes, and energy dispersive x-ray detectors. I too thought that the people using TEM were crazy. As a matter of fact I tried hard to see if an SEM wouldn't do the job. I will be the last to make the broad statement that SEM's are no good; I know better than that. There are higher brightness sources, the La88 source gives you an increased electron density, not too many people have been working with asbestos with better electron sources or field emission source; this can give you an increased yield of x-rays..... I am trying to be objective, but you look at what people have done and compared SEM with TEM and they all come up with the same conclusion about the superiority of the TEM. MCALEAR: We have many customers who use both TEM and SEM and I don't believe the votes are in on this as yet by a long shot. ANDERSON: The whole difference is the size range that we are considering. We are considering asbestos in water and the asbestos fibers are very small-about 250A wide. I. STEWART: There was a comment about the statistical significance of results and, as I understand it, your phi is basically an attempt to be realistic and say that there will be backgrounds. Now, the gentleman from Johns Manville surprised me by mentioning there is zero background. We have done a lot of blanks with nothing in them but we do not call them zero background, which I think is totally unrealistic with asbestos. The values that have been published in the literature range from 30 fibers per grid square, reported by Tony Richards of Turner Bros., down to this claim for zero or near zero. Now if you take your 20 grid squares, that means you have six hundred fibers, at which point you are really talking about noise-to-signal ratio. J. KRAMER: I'd like to address the question of SAED confirmation or chrysotile and the amphiboles. I think you ought to be complimented on the details of your general method of preparation, which I think all people need, and they can go through step by step and determine whether this works in their lab or not. But in the literature and here in terms of electron diffraction confirmation we have seen two different wall paper patterns. 374