Document rpd2rOgVMOym5xMgzpRXmkyna

fa* 662 MICROSTRUCTURB OP CHRYSOTILE ASBBSTOS 1 s (fr) Growth patterns J. * 1/ . void ari often observed. The samples from Quebec, FI*. 12 ahowt tht typioal orotMaotion of (h umpla Qloba and Tasmania, which contain a number or from Olobe. Arizona, Betides normal cylinder at A, fibrils consisting of a roultl-splral lattiee, are classified there are many anomalous ones. including a multi* into this group. (3) The most frequent values of the layer spiral at B, a non-hoilowed cylinder at C, two* outer and inner diameters of the samples showing a step growth at D, and an inter-fibril segment at . sharp distribution (Coalinga and Transvaal) are 220- Anomalous black and white contrast indicated by the 270 A and 70-80 A, respectively. The value of 260 A snows seems to be due to higher order reflexions which for the outer diameter obtained by Whittaker (1957), are excited simultaneously. It is seen that inter* and (partly theoretically and partly from X-ray diffraction intra-fibril parts are well preserved without serious measurements on a sample from Bell Mine, Quebec), radiation damage, as a result of the precautions referred. is in good agreement with the present value, but his to in the preceding section. We can see that the inter- value of 110 A for the inner diameter is not. (4) Most fibril sites are generally vacant. fibrils thicker than about 350 A seem to contain a These anomalous growth patterns were rarely seen discontinous step-growth. in the samples from Quebec, Coalinga and Transvaal, while they were often observed in the samples from Globe and Tasmania. (c) Behaviour of the 7*3 A fringes Dlsmslo* (a) Crystal structure Although Fadurow (1950) suggested a triclinic struc It was reported in the previous paper (Yada, 1967) ture for chrysolite, it is thought at present that the that all the circumferential 7-3 A fringes observed were .principal form of chrysotile is clino-chrysotile, while spiral or multi-spiral. Although such spiral structures ortho- and para-chrysotile are less common (Whit were found in all the samples studied, it was shown for taker & Zussman, 1956). The result found in the present the first time that a perfectly concentric structure also experiment that the 4-5 A fringes are not always exists in some samples. In Fig. 13, (a) shows the spiral parallel to the fibril edge seems to be important in structure and (6) the multi-spiral one, seen in the connexion with the crystal structure of chrysotile. The sample from Transvaal. Fig. 14 shows the concentric tilting of these fringes is equally compatible with either structure in the same Transvaal sample, in which two of the following structures. (1) a helical roll with a rec pairs of through-focused images are shown, and tangular lattice, retaining the monoclinic structure (a forming the concentricity. It is noted that the upper clino-chrysotile in which the a axis is not exactly parallel fibril clearly has concentric structure, though it is a little to the fibre axis and the b axis not exactly perpendicu elliptical in shape. The 4-5 A fringes corresponding lar to the fibre axis) and (2) a stra`-Jht roll with a non- to 020, in the radial direction, are clearly visible in the rectangular lattice (an anorthic stricture in which the upper images, while they are scarcely seen in the lower b axis is perpendicular to the fibre axis but the a axis is ones. The disappearance of these fringes seems to be not parallel and the e axis not perpendicular to the fibre caused not by a lack of resolution but by a deviation axis). A choice between these two possibilities could be from the Bragg condition due to the tilting of the made if the 2-6 A transverse fringes corresponding 010 layers. Table 2 shows the distribution of the con to 200 in the central region of the fibrils were clearly centric and spiral structures for three samples. It is observed. Unfortunately, however, it was difficult to seen that the percentage of fibrils showing the concen resolve these fringes, as mentioned in the preceding tric structure is fairly high, being nearly a half in the section. However, by a dark field method in which a Transvaal sample, while it is low for the Coalinga and 200 reflexion is selected by the use of a small objective Tasmania samples. Although the total numbers of aperture and the corresponding dark field images comcross sections examined .hre not sufficiently large for pared with a normal bright field image, it is possible to statistical analysis, the observed differences seem to be decide which of the helical roll and the straight roll it significant and are presumably related to the different the true structure. growth conditions. * I (d) Distribution of diameters From the histograms for the distributions of the inner and outer diameters shown in Fig. 15, the fol According to a preliminary test using such a method it appears that the former is the case.* With a model of the helical roll it is quite easy to understand the experi mental fact that the tilt angle of the 4-5 A fringes varies over a range of up to about ten degrees from lowing points are noted: (I) The distribution of the diameters is considerably different from sample to sample, presumably depending on the growth condi tions in each locality. (2) In samples for which the outer diameters Bre distributed over wide ranges, there fibril to fibril, though no periodicity of the angular distribution was observed in the present experiment. There is still the possibility of a polytype structure. The non-parallel 7-3 A fringes seen in the synthetic chrysotile (Fig. 9) are the result of such a polytype. is a tendency for the peak position in the histogram to . This. type..pf structure, however, does not necessarily .'-'shift to a larger valuc:.Iris'uQhsamples' fibrils both with'V- wyl'-a-Jfcv' ;; -,-v!------'----:----------------- . . : -.>._ '>yV- a very small inner diantcler ind .without the central The details will be reported elsewhere. 8000 6554 i ' PRODl ICED RY mo n