NOTES HIGH RESOLUTION ELECTRON MICROGRAPHS OF IMOGOLITE

Transcription

1 Clay Minerals (1970) 8, 487. NOTES HIGH RESOLUTION ELECTRON MICROGRAPHS OF IMOGOLITE Imogolite was discovered by Yoshinaga & Aomine (1962) in the clay fraction of a glassy volcanic ash soil ('Imogo9 in Japan. Chemical analyses gave the fox~aula 1-1 SiO~. AltOs. 2" H~O (+) without correction for allophane as an impurity with which it is usually associated (Wada & Yoshinaga, 1969). Imogolite appeared in the electron microscope as micron-length threads with a diameter of 100 to 300 A (Yoshinaga & Aomlne, 1962). More recently, they were found to consist of fibre units with separations in the order of A (Yoshinaga, Yotsumoto & Ibe, 1968; Yoshinaga, 1968; Wada & Yoshinaga, 1969) or A (Russell, McHardy & Fraser, 1969). The present study was undertaken to provide additional data on the morphology and dimensions of the unit of imogolite by means of high resolution electron microscopy. Imogolite was separated from macroscopic gelfilms collected from two pumice beds, Kanuma (Nagasaka, Utsunomiya, Toehigi) and Kurayoshi (Shuki, Kurayoshi, Tottori). They were treated first with 30% H~O~, and then successively with Na2S~O,-NaHCOs-Na citrate and 2% NazCOs (Jackson, 1956) for removal of organic matter and extractable oxides. The latter two treatments also had, to some extent, the effect of removing associated amorphous silicates (Yoshinaga, 1968; Wada & Greenland, 1970). After completing the above treatments, the samples were washed successively with water, alcohol and acetone, and then air-dried. In order to carry out direct observation, the pulverized, air-dry samples were suspended in water and treated for 15 min with an ultrasonic probe at 400 kc. A small drop of the suspension was dried in air on a microgrid reinforced with carbon (Fukami & Adachi, 1965). This made it possible to observe imogolite threads spanning the hole of the microgrid without any supporting film. For thin sections, the pulverized samples were immersed in a methylmethacrylate-n-butylmethacrylate mixture (3:7) and, after adding 2,4-dichlorobenzoyl peroxide (1"5% v/w), allowed to stand for 6 hr at 45 ~ C. The embedded imogolite threads were sectioned by means of an ultramicrotome equipped with a glass knife. The thin A sections were mounted on a microgrid, and the embedding plastic was dissolved with chloro. form prior to observation. The instruments used were the JEM-120 and JEM-100B electron microscopes at an accelerating voltage of 100 or 120 kv and magnification of x 50,000.

2 488 Notes Plate 1 shows an electron micrograph of dispersed threads of imogolite. Most of the threads consist of one or more fibre units which are bent yet run almost in parallel. An interesting feature of the micrograph is that the individual lines never occur singly, but as two or more parallel lines. This suggests that the fibre unit is tubular, with a pair of parallel lines representing the walls of the tube. Similar features can be observed in the electron mierographs taken for imogolites from different sources (Yoshinaga, 1968). The appearance of the areas (encircled in Plate 1), where between two and four units are thought to be in parallel alignment, supports the view that the unit fibre is a hollow tube rather than that each line is a fibre with a single core. A broken tube can be seen at one point, indicated with an arrow. The outside diameters of the tubes were found to be A, and the inner diameters 7-10 A. The observed unit has been dehydrated by irradiation of the electron beam in vacuum. This was confirmed by comparing the electron diffraction with the X-ray diffraction data (Wada & Yoshinaga, 1969). The outside diameters of the tubes could then be correlated with the d-value for an X-ray diffraction band of imogolite, 18"4 to 18.8 A enhanced either by heating at ~ C or by evacuation. If the d-value 18.6 A is equated with the interplanar distance dl0 of the hexagonal lattice which would be formed by a close packing of the tubes, their outside diameter would be 21.5 A. The accuracy with which the latter can be measured, however, does not permit detailed interpretation. Nor is the appearance of the thicker bundles of units very helpful in resolving the question of packing of the units. Their variable and complex features may reflect those in the packing of the units and in the twisting of the threads. Moreover, image formation by the electron beam is known to be complex, and this adds to the difficulties of interpretation for specimens with a thickness of several tens of angstroms. An electron micrograph of the thin section is shown in Plate 2. The orientation of the imogolite threads is random in the section, but the units in each thread can be expected to align in the same direction. If the units are tubular, and are sectioned perpendicularly to their axes, they should appear as clusters of rings with an outside diameter of A. Such clusters are not likely to occur frequently, as the axis of each tube must be oriented parallel to the electron beam, not necessarily for the whole thickness of the section but at least 200 A. A considerable number of the rings appears showing good contrast, although their clustering is not so spectacular as might be expected. Whether this is caused by the poor alignment of the units or by difficulty in preserving their association as a thread in the thin section has not been resolved. After the dissolution of the embedding plastics, the latter difficulty could be expected for threads with a length of A. The general appearance of the mierograph also seems to s~aggest only fragile association of the units. The two structure models so far proposed for imogolite (Russell, McHardy & Fraser, 1969; Wada & Yoshinaga, 1969) cannot be reconciled as such with the present observation. The value of diffraction for the elucidation of the structure of imogolite

3 PLATE 1 An electron micrograph of dispersed threads of imogolite. Threads comprising a few tubular units running parallel are encircled, and a broken single tube is indicated with an arrow. To face page 488

4 PLATE 2 An electron micrograph of imogolite in thin section.

5 Notes 489 is limited by its paracrystalline nature, but this new information on the morphology and dimensions of the structural unit should be relevant to any future studies. Faculty of Agriculture, Kyushu University, Fukuoka, Japan. Faculty of Agriculture, Ehime University, Matsuyama, Japan. Japan Electron Optics Laboratory Company Ltd., Tokyo, Japan. 16 February K. WADA N. YOSHINAGA H. YOTSUMOTO, K. IBE, S. Area REFERENCES FUKAMI A. & ADAcm K. (1965) J. Electron Mierose. Chiba Cy. 14, 112. JACKSON M.L. (1956) Soil Chemical Analysis-Advanced Course, p. 31. Published by the author, Madison, Wisconsin. RUSSELL J.D., MCHARDY W.J. & FRASER A.R. (1969) Clay Miner. O, 87. WADA K. & GmEENLAND D.J. (1970) Clay Miner. 8, 241. WADA K. & YOSHINAGA N. (1969) Am. Miner, 54, 50. YOSHINAGA N. (1968) Soil Sci. Pl. Nutr. Tokyo 14, 238. YOSHINAGA N. & Ao~ S. (1962) Soil Sci. Pl. Nutr. Tokyo 8 O), 22. YOSHINAGA N., YOTSUMOTO H. & IBE, K. (1968) Am. Miner. 53, 319.