STATE OF ILLINOIS DWIGHT H. GREEN, Governor DEPARTMENT OF REGISTRATION AND EDUCATION FRANK G. THOMPSON, Director DIVISION OF THE STATE GEOLOGICAL SURVEY M. M. LEIGHTON. Chief URBANA REPORT OF INVESTIGATIONS— No. 115 DIAGNOSTIC CRITERIA FOR CLAY MINERALS W. F. BRADLEY Reprinted from The American Mineralogist, 30, (1945) PRINTED BY AUTHORITY OF THE STATE OF ILLINOIS URBANA, ILLINOIS 1946 ILLINOIS GEOLOGICAL SURVEY LIBRARY APR 6 1946 ORGANIZATION STATE OF ILLINOIS HON. DWIGHT H. GREEN, Governor DEPARTMENT OF REGISTRATION AND EDUCATION HON. FRANK G. THOMPSON, Director BOARD OF NATURAL RESOURCES AND CONSERVATION HON. FRANK G. THOMPSON, Chairman NORMAN L. BOWEN, Ph.D., D.Sc, LL.D., Geology ROGER ADAMS, Ph.D., D.Sc, Chemistry LOUIS R. HOWSON, C.E., Engineering CARL G. HARTMAN, Ph.D., Biology EZRA JACOB KRAUS, Ph.D., D.Sc, Forestry ARTHUR CUTTS WILLARD, D.Engr., LL.D. President of the University of Illinois CI^OLOGTCAL SURVEY DIVISION M. M. LEIGHTON, Chief ILLINOIS STATE GEOLOGICAL SURVEY 3 3051 00005 7368 SCIENTIFIC AND TECHNICAL STAFF OF THE STATE GEOLOGICAL SURVEY DIVISION 100 Natural Resources Building, TJrhana M. M. LEIGHTON, Ph.D., Chief Enid Townley, M.S., Assistant to the Chief Velda a. Millaed, Junior As'St. to the Chief Helen E. McMorris, Secretary to the Chief Effie Hetishee, B.S., Geological Assistant Research A€ 2. The association of ethylene glycol with halloysite. (a) natural endellite; (b) endellite treated with ethylene glycol; (c) halloysite resulting from aging of (a) ; (d) halloy- site treated with ethylene glycol; and (e) a plot of Hendricks and Teller's function for units of 10.8 and 7.2 A, plotted to the approximatfe- scale of the diffraction diagrams. The arrpw indicates the position of normal reflection. 708 W. F. BRADLEY other active liquids (1) would also have indices ranging from 1.52 to 1.55. Halloysite derives from endelUte upon loss of the molecular water. The resultant material is very porous, retains its original bulk, and readily absorbs water, but the structural change from the endeUite crystalliza- tion to a kaolinitic crystaUization is irreversible, or at best incompletely reversible. Wet halloysite specimens give essentially the halloysite dif- fraction diagram. The diffraction diagram of halloysite wet with ethylene glycol differs in a distinctive way from those of endellite, halloysite, or kaolinite. The prismatic interferences of endellite and halloysite are unchanged, but of the possible interferences related to the base, only one line, at 3.6 A, ap- pears sharp and well-defined. In the region corresponding to periodicities from about 11 to about 7 A^ there appears only a broad weak band, in which the intensity seems to be somewhat higher near its edges. The function derived by Hendricks and Teller (9) for mixed layer minerals applies only to equally probable units of the same form factor. In the absence of suitable information about the form factors it is im- practical to attempt to compute a theoretical intensity distribution for this case (as will later be done for a more apt example), but a comparison of the plot of Hendricks and Teller's function with the diffraction dia- gram in Fig. 2 indicates clearly that a random or near random interleav- ing of 10.8 and 7.2 A units would produce the same sort of diagram as is observed. Actually, it seems indicated that simple 7.2 A units outnumber the complex 10.8 A units, but the characterization of halloysite as the collapsed structure subject to random penetration of active hquids, seems justified. %■: % Illite Although iUites invariably contain more or less molecular water, as is apparent from thermal analysis curves (10), the crystallization is de- termined by mica principles, and the position in the structure of any water not merely absorbed on exterior surfaces is not known. Treatments of illite with organic liquids do not therefore result in any structural changes, and diffraction diagrams do not differ from those of the un- treated materials. The frequently encountered instances in which illites are observed to be mixed with extraneous layers are discussed in connec- tion with the mixed layer minerals. Mixed Layer Minerals Several instances have been reported (11, 12, 13, 14, 15) in which it seemed that a specimen under consideration was actually composed of irregularly alternating layers of two distinct minerals. The treatment of DIAGNOSTIC CRITERIA FOR CLAY MINERALS 709 diffraction effects from such mixed layers by Hendricks and Teller (9) has afforded a basis on which one can hope to examine in detail some of these suspected instances. A particularly fortunate case has been that of some bravaisite from the type locaUty, U. S. Nat. Museum Specimen No. 4918.* A study has shown not only that this material is about equally divided between two species, but also that the two form factors are similar enough to permit a reasonably accurate construction of the calculated diffraction effects. Grim and Rowland (10) concluded from thermal analysis curves that this material included both illite and montmorillonite. If it be assumed, then, that this material be made up of a random stacking of equally prob- able illite and montmorillonite layers, one case can be developed for the natural material itself, and a second case can be developed for the mix- ture of ilhte with glycol-complexed montmorillonite which would result from glycol treatment. In connection with the previous study of complexes of montmoril- lonite with various types of organic liquids (1), observed 00/ diffraction sequences from a series of complexes of varying cell heights have been available. Estimated intensities for the powder diffraction rings for this entire group have been plotted together against sin d/\, and Fig. 3 is a Fig. 3. Compositite curve of estimated relative intensities of CC/ powder diffraction rings versus sin d/\ taken from a group of organic molecule-montmorillonite complexes of various c-axis periodicities. smooth average curve summarizing the composite data. Although each individual diagram has been rather strongly influenced by the organic matter, the allowable degree of error for the present purposes is so great that this curve adequately represents not only the contribution of the montmorillonite layers, but also the contribution of the similarly con- stituted, but rigid, ilhte layers. To arrive at the angular distribution of scattered radiation from the random mixed layer sequences assumed, then, it is only necessary to * Furnished through the courtesy of Dr. W. F. Foshag. 710 W. F. BRADLEY multiply the ordinates of Hendricks and Teller's function for each case by the ordinates of Fig. 3. In Fig. 4 curves derived in this way for a mix- ture of 10 A and 15 A units, and of 10 A and 17 A units are compared with the diffraction diagrams of bravaisite and of glycol treated bravais- > o < A. < , < < O T ■o 1 < t < r « < o ■ aJ ^ ,/ x, . - ' ,A .00 .02 04 o« oe IS .20 .22 24 20 Fig. 4. Bravaisite, Noyant Allier, France, U. S. Nat. Mus. no. 4918 (above) and the same treated with ethylene glycol (below), each compared with its synthesized curve for random mixed layers to scale approximately equivalent to the diffraction diagram. Erect arrows indicate positions of normal or near normal reflection. Inverted arrows indicate nor- mal reinforcements in the mixing function which are extinguished by the form factor. DIAGNOSTIC CRITERIA FOR CLAY MINERALS 711 ite, respectively. Reasonable agreement in the position, relative inten- sity, and apparent breadth of all the important maxima is evident. No other instance has been encountered in which the relative frequen- cies of illite and montmorillonite layers are nearly enough equal to con- form to Hendricks and Teller's equation. However, the principle can probably be extended in a qualitative way to cases in which one species predominates over the other to a moderate extent. If for example, illite occurred more frequently than montmorillonite, we would expect from Fig. 5. Powder diffraction diagram of "glimmerton," Sarospatak, Hungary, from U. Hofmann, treated with ethylene glycol (above), compared with illite purified from shale from Alexander Count}^, Illinois, after the same treatment (below). inspection of the function that in general maxima would approach the position of the sharp interference maxima of illite. For the natural mate- rial, where the form factor extinguishes one of the small-angle maxima of the complex function, the diagram would only more nearly resemble illite. For glycol-treated material, however, both of the small-angle maxima should persist and should merely approach each other. Among the natural materials in which two small-angle maxima are developed by glycol treatment are several which have heretofore been described as ilUtes or at least as mica-like clays. These include the ''ghm- merton" from Sarospatak, Hungary (16), the ilUte purified from an un- derclay in Grundy County, Illinois (17), a "metabentonite" from High Bridge, Kentucky (10), and possibly some glauconites. In Fig. 5 the dia- gram for glycol-treated "glimmerton" is reproduced. For specimens with lesser proportion of montmorillonite layers the diagrams are less clear. The ratio of mica to montmorillonite layers in the "glimmerton" is prob- ably near 2; in the other cases, somewhat higher. A diagram of a glycol- treated iUite, unmixed with extraneous layers, is included for compari- son. Mixed layer minerals of other types also no doubt exist, of which par- 712 W. F. BRADLEY ticularly a mixture of montmorillonite and illite with montmorillonite predominating should be amenable to interpretation by this method, but no clear cut occurrence has yet been recognized. Miscellaneous Minerals The number of hydrous clayey or clay-like minerals which have been described and named is too great to attempt to catalogue them here. Probably the greatest number of these are the varieties which are actu- ally members of the montmorillonite group. Most of the specimens of this sort reported on by Grim and Rowland (10) have been made availa- ble for examination in this study. The characteristic 17 A glycol complex is readily prepared from all of the so-called beidellites discussed by them, except their number 8 A (U. S. Nat. Museum no. R7595), from the non- tronites, the white Hector magnesium clay, smectite, chloropal, and volkonskoite. Loosely related to these materials are two other mineral types in which comparable structural units appear, the vermiculites and atta- pulgite. Vermiculites. — Although the ideal structure of vermiculite (12) im- mediately suggests that the interlayer water could be displaced by or- ganic liquids, no attempts to make such complexes from the familiar crystalline materials have yet been successful. However, a series of cryptocrystalline weathered amygdules collected by Dr. R. M. Grogan from basic Keweenawan flows along the north shore of Lake Superior, which have been identified as vermiculites by .x-ray diffraction methods, do form the same sort of 17 A glycol complex as is observed for montmoril- lonites. In view of the known tendency for vermiculites to occur mixed with chlorites, it seems quite possible that such mixing, particularly if it were lateral within layers, might inhibit the introduction of glycol into the natural large crystals even though it would be readily introduced into a fine-grained, or into an ideal specin en. Attapulgite. — The association of any individual water molecule with the silicate portion of attapulgite is probably due to hydrogen bonding, analogous with that exhibited by the water associated with montmoril- lonite, but the disposition of the silicate chains (18) confines the water to narrow channels and precludes the changes in dimensions that are observed for swelling clays. The water channels have cross sections of about 3.7X6.0 A, which might accommodate single strings of ethylene gycol, but in the absence of dimensional changes, only minor intensity variations are to be expected upon glycol treatment, and no obvious variations were noted in the diffraction diagrams. Refractive indices, however, are sharply reduced. Well-oriented flakes with 7 = 1.540, DIAGNOSTIC CRITERIA FOR CLAY MINERALS 713 7 — a = .032, showed after treatment 7=1.50 with comparable bire- fringence. The calculated mean index for the silicate skeleton above is about 1.44. The observed indices can therefore apparently be considered to arise from extensive extraction of the hydrogen-bonded water accom- panied by only very limited penetration of glycol into the channels. Summary The association of certain organic liquids with clay minerals affords a basis for the classification of such minerals into the principal groups, and at the same time serves to confirm the general features of the respec- tive structures. ACKNOWLEDGMFNT The author is indebted to Dr. R. E. Grim who furnished many of the mineral specimens employed, and who determined the refractive indices that are cited. References 1. Bradley, W. F., J.A.C.S., 67, 975-981 (1945). 2. Gieseking, J. E., Soil Sci., 47, 1-13 (1939). 3. Hendricks, S. B., /. Phys. Chem., 45, 65-81 (1941). 4. Davidson, R. C, Ewing, F. J., and Shute, R. S., Nat'l Pet. News, 35, No. 27, R318- 321 (1943). 5. MacEwan, B.M.C. Nature, 154, 577-578 (1944). 6. Hendricks, S. B., and Alexander, L. T., /. Am. Soc. Agron., 32, 455-458 (1940). 7. Alexander, L. T., Faust, G. T., Hendricks, S. B., Insley H., and McMurdie, H. F., Am. Mineral., 28, 1-18 (1943). 8. CoRRENS, C. W., and Mehmel, M., Zeits. Krist., 94, 337 (1936). 9. Hendricks, S. B., and Teller, E., /. Chem. Phys., 10, 147-166 (1942). 10. Grim, R. E., and Rowland, R. A., Am. Mineral., 27, 746-761, 801-818 (1942). 11. Gruner, J. W., Am. Mineral., 19, 557-575 (1934). 12. Hendricks, S. B., and Jefferson, M. E., Am. Mineral., 23, 851-862 (1938). 13. Gruner, J. W., Am. Mineral., 29, 422-430 (1944). 14. Jackson, M. L., and Hellman, N. N., Proc. Soil Sci. Am., 6, 133-145 (1941). 15. Nagelschmidt, G., Mineral. Mag., 21, 59-61 (1944). 16. Magdefrau, E., and Hofmann, U., Zeits. Krist., 58, 31-59 (1937). 17. Grim, R. E., and Bradley, W. F., /. Am. Cer. Soc, 22, 157-164 (1939). 18. Bradley, W. F., Am. Mineral, 25, 405-410 (1940).