~ gun i ”o m ABSTRACT CHARACTERIZATION OF A MONTMORILLONITE IN A NORTHERN MICHIGAN PODZOL by Gerhard John Ross lllite, chlorite, interstratified chlorite-vermiculite and chlorite-montmorillonite have weathered in situ through a vermiculite stage to montmorillonite. This weathering sequence is prevafent in the sandy soils of Northern Michigan and results in accumulation of montmorillonite in the Az-horizon of these soils in this area. This montmorillonite is classified as a beidellite since it is dioctahedral and has a relatively high aluminum content and a high charge, most of which originates from substitution of Al for Si in the tetrahedral layer. This beidellite interpretation is supported by data from the Greene-Kelly test, X-ray diffraction, differential thermal, infra-red, and chemical analyses. The formula for the beidellite is: -0.0h -O.k5 3 +. E1162 Fe0.18 M90.2%] E'OAS “3.53 .010 ”“2 “$.19 which has the high layer charge of 126 me/lOOg. This beidellite has a strong tendency to fix potassium as is evident from a 40% reduction in cation exchange capacity when saturated with potassium, heated to llOOC, and resaturated with ammonium to replace potassium. The acid dissolution technique (Osthaus, I956) for determination of tetrahedral aluminum does not apply to the beidellite studied. CHARACTERIZATION OF A MONTMORILLONITE IN A NORTHERN MICHIGAN PODZOL By Gerhard John Ross A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Soil Science I965 ACKNOWLEDGMENTS The author wishes to express his sincere gratitude to Dr. M. M. Mortland for introducing the author to the problem studied herein and for further guidance in conducting this study. He also wishes to thank Dr. E. P. Whiteside for his advice and assistance. The author appreciates the Graduate Assistantship offered to him by Michigan State University, enabling him to pursue and complete this study. 7':*~k*‘k‘k***~k***~k TABLE OF CONTENTS CHAPTER Page I. INTRODUCTION . . . . . . . . . . . . . . . . I II. LITERATURE REVIEW . . . . . . . . . . . . . 3 Montmorillonites - Structure and Classification 3 Beidellite . . 5 Relationship with Illite and Chlorite . . 7 Weathering and Alteration of Illite and Chlorite in Soils. . . IO III. MATERIALS l3 Collection of Samples. l3 Soil Profile Description IS IV. LABORATORY PROCEDURES. I7 Dispersion and Mineral Fractionation I7 X- ray Diffraction. l8 Greene- Kelly Test. . . l9 Differential Thermal Analysis. 20 Infra- red Analysis . 20 Total Specific Surface Area. 2] Cation Exchange Capacity and Its Reduction by Potassium Fixation. 2] Hydration Properties . . 2h Acid Dissolution for Determination of Tetrahedral Aluminum . 2h Total Chemical Analysis. . 25 Calculation of Montmorillonite Formulas . 26 Electron Microscope Examination. 26 V. RESULTS AND DISCUSSION . 27 X- ray Diffraction Analyses of Clay Fractions . . . . . . . 27 Application of Greene- -Kelly. Test . . . . . #3 Differential Thermal Analysis . . . . . . A7 Infra-red Analysis . . . . . . . . . . . . 52 TABLE OF CONTENTS, Continued CHAPTER VI. VII. Cation Exchange Capacity and Its Reduction by Potassium Fixation. Hydration PrOperties: Basal Spacings With Water . Chemical Determination of Tetrahedral. Aluminum in the‘LO.2/1CIay Fraction of the Deer Park Az-horizon. . Chemical Analysis of¢10.2AIClay Fraction . . . . . . . . . . . Calculation of Structural Formulas Electron Microsc0pe Examination of the Fine Clay Fraction of the Deer Park Az-horizon . GENERAL DISCUSSION CONCLUSIONS. LITERATURE CITED . Page 57 6h 69 73 75 85 89 96 98 TABLE IO LIST OF TABLES X-ray diffraction powder data for the Mg-AZ-horizon fine clay with Ca-B.J.M. beidellite with interlayer glycerol Peak temperatures (0C) in the differential thermal analysis of 40.2% clays. Frequency (cm'l) of infra-red absorption bands. Cation exchange capacity and total specific surface ofz:O.2/(clays Cation exchange capacity and 00] spacing of¢10.2A clays after K- -saturation, heating to IOOOC overnight, and re- saturation with NHh . o Basal spacings in angstrom units (A) formed during water uptake by the 4.0. leclays. . . . . . . Dissolution of aluminum of the410.;u fraction of the Deer Park A -horizon at various acid treatment tImes. Chemical analysis data in percent oven-dry weight of40.2/u clays Elemental analysis of A2-horizon montmorillonite. . Formulas for dioctahedral montmorillonites Page Al 50 58 S9 50 67 70 7h 79 86 FIGURE IO. ll. I2. 13. LIST OF FIGURES Locations of sites sampled in preliminary investigation X-ray tracings of.:2;4clay fractions from locations in Figure I X-ray tracings of.¢2/Lclay fraction of the indicated horizons of the Deer Park profile. X-ray diffraction tracings of¢:0.2/Iclay fraction . . . . . . . . . . . X-ray diffraction tracings of 2:0.2/u clay fractions illustrating Greene-Kelly test . . . Differential thermal analysis curves of the40.2/u clay fraction - Infra-red analysis curves of the 40. wt clay fraction in the 600 to I200 Cm' region. Infra-red analysis curves of the¢:0.2 clay fraction in the 3400 to 3800 Cm' region. X-ray diffraction tracings of K-saturated, heated, and NHu-resaturated clays X-ray diffraction tracings showing basal spacings of410.2;4clay fraction at I00% and at h0% relative humidity. . Dissolution curve for determination of the tetrahedral aluminum of the440.2}4clay fraction of the Deer Park Az-horizon. . Electron micrographs of the fine clay fraction of the Deer Park Az-horizon. Electron micrographs of the fine clay fraction of the Deer Park Az-horizon. Page 28 30 35 38 AA #9 54 56 62 66 7I 87 88 I. INTRODUCTION Franzmeier (I962), in his study of a chrono-sequence of Podzols in Northern Michigan, found that the layer silicates of the parent material are composed of kaolinite, chlorite and illite, but that montmorillonite is almost the only layer silicate present in the AZ-horizon of the soil studied. He postulated a weathering sequence in the 2:I silicates as follows: chlorite;::::>‘A illite//’ . . ___J . . vermIcul Ite\__ montmorI I IonIte This sequence implies the removal of the brucite layer from chlorite and of the interlayer potassiUm from illite. Considering the structure of the chlorite and of illite, it follows that the montmorillonite formed would: (I) have relatively high substitution of Al for Si in the tetrahedral layer; (2) have relatively high cation exchange capacity predominantly originating in the tetrahedral layer; and (3) fall in the befdellite end of the montmorillonite-beidellite series. This investigation was designed to test the foregoing hypothesis. This was done by studies of the struCture, chemical composition, and reactiomsof the clay fraction from the AZ-horizon of these Northern Michigan Podzols, and by its comparison with structure, prOperties, and reactions of Wyoming bentonite, nontronite, and the clay fraction of the underlying soil and its parent material. II. LITERATURE REVIEW Montmorillonites - Structure and Classification In I897, Damour and Salvetat prOposed the name montmorillonite for a clay mineral which had the approximate composition: AISiOZ . Al203 . (l+x) H20 and which could be dehydrated reversibly at a low temperature. Ross and Shannon (I925) proposed the formula (Mg, Ca) . AI203 S SiO2 . nHZO to account for the M90 and CaO which were present in all of the newer amiyses. Later X-ray studies showed that Mg is frequently present as an essential constituent while Ca is not incorporated in the structure. The structure of montmorillonites is based on that of pyrophyllite (Hoffmann, EndeII, and Wilm - I933, I93h). This mineral, as shown below, has the sequence of atomic planes perpendicular to the pseudo-hexagonal axes. Pyrophyllite Atomic planes Silicate layer 602- (tetrahedral) lISi’4+ Gibbsite layer LI 02' 2 [0H]' (octahedral) h Al 3+ LI 02‘ + 2 L6H] Silicate layer A Sit++ (tetrahedral) 6 02‘ Combined, the atomic planes shown above form a neutral structure, which may be considered as one lattice layer. These layers may be extended indefinitely along a horizontal plane. Since each of these layers are neutral, there is only a small attractive force between them, resulting in the excellent cleavage of the mineral. To account for the swelling characteristics and relatively large cation exchange capacity of montmorillonite, Hendricks (I942) and Marshall (I945) suggested that in some montmorillonites, such as Wyoming bentonite, negative charges arise from replacement of Al3+ by Mg2.+ or Fe2+ in the octahedral layer. In other montmorillonites, such as beidellite and nontronite, negative charges are due to substitution of Al'+ for Si+ in the tetrahedral layer. The classification of montmorillonites is based essentially on their structural characteristics. Accordingly, dioctahedral and trioctahedral montmorillonites are recognized (Stevens, I942, Ross and Hendricks, I945). In dioctahedral montmorillonites only two-thirds of the positions in the octahedral layer are filled by ions in six-coordination, whereas in trioctahedral montmorillonites all positions are filled. Arrangement of the dioctahedral montmorillonites according to the SizAl ratio gives rise to the montmorillonite- beidellite series (Marshall I935). The maximum ratio of 2.4:l corresponds to that of montmorillite (Wyoming bentonite); the minimum ratio of l.4:l approaches that of beidellite (Black Jack Mine, Idaho). The classification of dioctahedral montmorillonites in a single,continuous,isomorphous series has recently been challenged by Grim and Kulbicky (l96l). From analytical data of 42 montmorillonites, they concluded that two different aluminous types werepresent, namely, Cheto- and Wyoming types. These types differr primarily in the larger amount of magnesium present in the octahedral layer of the Cheto types. Cation exchange capacities are also higher in the Cheto type montmorillonite. Beidellite Larsen and Wherry (I925) proposed the name beidellite for a clay mineral from Beidell, Colorado. Because this name was applied to many minerals distinctly different from the type mineral, Ross and Hendricks (I945) redefined beidellite as the aluminum-rich end-member of the isomorphous montmorillonite-beidellite series. However, X-ray diffraction and thermal data showed that most of these beidellites were mixtures containing other minerals (Grim and Rowland, I942). Several mineralogists have discredited the name beidellite (Grim, I953; Brindley, I955; Frank-Kamenetskey, I958; Ross, I959). Others have included beidellite in their classifications (Mac-Ewan, l95l; Brown, I955; Caillere and Henin, I957; Mackenzie, I957; Strunz, I957). In I959, Ross proposed that the use of the name beidellite be discontinued since there was no confirmed occurrence of a mineral with a structure and composition similar to that required by Ross and Hendricks' definition. Because of the adverse criticism concerning the identity of beidellite, Weir and Greene-Kelly (I962) carefully re-examined a clay sample from the Black Jack Mine, Idaho. This sample was part of a specimen earlier analyzed by Shannon (I924) and listed as a beidellite by Ross and Shannon (I925). From X-ray diffraction, thermal data, chemical analysis, and electron micrographs, Weir and Greene-Kelly (I962) concluded that the B.J.M. clay mineral is composed of a single mineral species, and that it is a genuine end-member beidellite of the dioctahedral montmorillonites. Greene-Kelly (l955a) suggested that montmorillonites and beidellites should be divided at the composition at which the lattice charges arising from octahedral and tetrahedral substitution equal one another. If the lattice charges due to tetrahedral substitution are greater than those due to octahedral substitution, the mineral should be named beidellite. If the reverse is the case, the mineral should be named montmorillonite. Greene-Kelly (l953b) also devised a test to distinguish between beidellite and montmorillonite. This test depends on the observation that Li-saturated montmorillonite collapses irreversibly when heated to ZOO-300°C., whereas beidellite similarly treated re-expands upon glycerol solvation. Relationship With Illite And Chlorite From their analyses of a large number of montmorillonites, Ross and Hendricks (I945) pointed out that as the beidellite end of the montmorillonite-beidellite series is approached there is a decided tendency toward the formation of mixed- layer type minerals containing potassium. In more fundamental terms the number of non-exchangeable interlayer ions, essentially of potassium, increases with the increase in replacement of silicon by aluminum in tetrahedral coordination. According to the authors, this indicates that there may be a complete gradation between the beidellite-type clay minerals and moderately high potassium mica-like minerals. Considering the structural similarity between mont- morillonite and illite, White (I950) has postulated that if sufficient of the potassium ions could be removed from the illite without marked decomposition of the mineral, it should be structurally equivalent to a member of the montmorillonite series (beidellite); and Nagelschmidt and Hicks (I943) have stated that the replacement of all exchangeable bases by potassium in minerals of the mont- morillonite group should lead to the formation of illite. Foster (I954) calculated structural formulas for illites, montmorillonites, and beidellites and found that illites are characterized by a high total charge of 0.75 and 0.95 on the latticelayers, of which approximately two-thirds is in the tetrahedral layers. The total charge on the montmorillonites ranges from about 0.30 to 0.50 and is predominantly in the octahedral layer. Beidellites have a total charge which is commonly lower than that found on the illites and higher than that on the montmorillonites. The origin of this charge is predominantly in the tetrahedral layers. Thus, as Foster (I954) pointed out, beidellite of all the montmorillonites has the composition of the clay that would be formed if the potassium in muscovitic illite were removed and replaced by exchangeable cations like sodium or calcium. However, at that time (I954) the natural occurrence of beidellite had not been authenticated. The removal of the brucite layer from chlorite should result in an expanding clay mineral similar to that formed by the removal of interlayer potassium from illite. Both would be characterized by a high aluminum for silicon substitution in the tetrahedral layer and a high total tetrahedral charge. Ample evidence (cited in the next section) shows that chlorites in soils weather to montmorillonites. In the course of weathering and removal of the brucite layers, intermediate stages may be represented by regularly interstratified chlorite-montmorillonite (Bradley and Weaver, I956), by randomly instratified chlorite- montmorillonite (Weaver, l956a), or by swelling chlorite IO (Lippmann, I954). The fact that the resulting montmorillonites commonly are dioctahedral pointed to the existence of a naturally occurring dioctrahedral chlorite. This existence has recently been confirmed. (Brydon, et al, I96l). Weathering and Alteration of Illite and Chlorite in Soils Murray and Leininger (I956) studied the changes in clay minerals of glacial tills under the influence of weathering. The three profiles studied represented immature, intermediate, and mature stages of weathering in a temperate climate. The results of their study showed that illite and chlorite I Change progressively to montmorillonite as the intensity of weathering increases. Droste (I956), in a similar study, also concluded that illite and chlorite of the parent mineral alter respectively to illite-montmorillonite and chlorite- vermiculite mixed layers. Further degradation of these mixed layers with attendant formation of montmorillonite continues progressively with increasing intensity of weathering. These studies yield essentially the same conclusions as those given elsewhere in the literature (Barshad, I955, I959; Bayliss and Loughnan, I963; Brydon, I964; Butler, I953; ll Droste and Tharin, I958; Droste, §£_al, I962; Frye, Willman and Glass, I960; Harrison and Murray, I959; Jackson, I959; Jackson and Sherman, I953; Kodoma and Brydon, I964). Murray and Leininger (I956) postulated that the initiating mechanism for the change from chlorite and illite to montmorillonite is oxidation of ferrous iron in the octahedral layers. This oxidation causes a decrease in the net charge which weakens the bonds between the sheets and thus allows solutions to enter. These solutions may then remove iron, magnesium, and hydroxyl ions from chlorite and potassium from illite. From X-ray analyses, Droste (I956), Droste, §£_§l, (I962) and Bayliss and Loughnan (I963) concluded that illite is more resistant to weathering than chlorite. They explained this by assuming that the brucite sheet in chlorite is more accessible to leaching solutions than the potassium, which fits tightly in the hexagonal cavities formed by adjacent tetrahedral sheets of illite. Furthermore, alteration of chlorite may retard alteration of illite in the early stages of chlorite and illite alteration (Droste, I962). This interaction is related to the reaction of illite with H+ ions from water. Garrels and Howards (1959) presented this reaction by the equation, l2 K-mica4-H+ : H-mica+-KT They also pointed out that the alteration of mica also involves the loss of aluminum and silicon from the tetrahedral sheets. If these ions are not removed from the environment, equilibrium conditions soon will be reached. Such equilibrium conditions may be influenced by ions released from alteration of chlorite. In an environment of intensive eluviation, potassium, aluminum, and silicon ions released from the lattice are continually removed resulting in formation of expanding clay minerals. Thus the intensive leaching conditions commonly present in Az-horizons of podzols should be particularly conducive to weathering of chlorite and illite to montmorillonite. Brown and Jackson (I958) and Franzmeier and Whiteside (I963) showed that montmorillonite (apparently weathered from illite, vermiculite, and chlorite) is the dominant clay mineral present in the AZ-horizons of some Northern Wisconsin and Northern Michigan podzols. Brown and Jackson (I958) identified this clay mineral as a montmorillonite having many vermiculite characteristics, such as high charge, collapse to I4A upon slight heating, and partial collapse to IOA with potassium saturation. III. MATERIALS Collection of Samples Samples were collected from the Az-horizons of soils at seven locations in Emmet and Cheboygan Counties, Michigan. The locations are shown in Figure l. The X-ray tracings of the clay fraction (42p) of each sample were examined for purity of montmorillonite in order to select the soil to be sampled for further study. The soil at location No. l was selected after examination of the X-ray tracings. A view of the soil profile at that site is shown in Plate l. A soil pit, about three by six feet, was dug and the profile described. The description was written according to standard conventions (Soil Survey Staff, I951) except that the International Society Color Council - National' Bureau of Standards (ISCC-NBS, Kelly and Judd, I955) color names were used. Samples were taken from each horizon. Samples of Wyoming bentonite, No. 25, from Upton, Wyoming, and of nontronite, No. 33a, from Garfield, Washington were used for comparison in the various studies. l3 Plate 1 IS Soil Profile Description Deer Park Sand Vegetation: Red oak, aspen, red pine, white pine, paper birch, red maple. Physiography and Relief: The soil described occurs on a low dune (3-5% sl0pe) along an old beach ridge. Ground water: deep Moisture: moist Stoniness: none Location: NE l/4 of SW l/4 of SE l/4 of Sec. 35, T38N, RIW. Benton Township, Cheboygan County, Michigan Depth Horizon (inches) Description A] 0-l Sand; brownish gray (l0YR3/l); very weak, fine, granular; very friable to nearly loose; medium acid (pH 6.0); abrupt, smooth boundary. A2 I-8 Sand; yellowish gray to very pale orange (l0YR8/2); loose; very strongly acid (pH4.8); abrupt, irregular boundary with some tongues extending to a depth of 36 inches. 82] 8-15 Sand; light brown to strong yellowish brown (7.5YR5/6); very weak, fine, - ranular; very friable; strongly acid TpH5.5); abrupt, irregular boundary. This horizon is also adjacent to the A2 tongues. l5-23 23-30 30-40 40+ l6 Sand; light brown to strong yellowish brown (7.5YR5/6); very weak, fine, granular; weakly cemented; slightly acid (pH6.3); gradUal, wavy boundary. This horizon is most pronounced at the bottom of the A tongues where it may extend down E0 36 inches. Sand; light brown to dark orange yellow (7.5YR6/6); very weak, fine, granular;nearly loose; slightly acid (pH6.3); gradual, wavy boundary. Sand; light brown to dark orange yellow (7.5YR6/6); single-grain nearly loose; slightly acid (pH6.3); gradual, irregular boundary. Sand; pale orange yellow (7.5YR8/4); sin Ie grain; loose; slightly acid; (pH .5). IV. LABORATORY PROCEDURES Most of the procedures used in the identification and characterization of the clays in this study are described in detail in Soil Chemical Analysis - Advanced Course (Jackson, I956) and in Soil Chemical Analysis (Jackson, I958). Soil samples from the A2 and C] horizons of the profile described above, and samples of Wyoming bentonite and of nontronite were anlayzed in detail. Samples from the remaining soil horizons were prepared only for X-ray analysis. Dispersion and Mineral Fractionation The soil samples were air-dried and passed through a 2-mm. sieve. At least l000 gm. samples were separated for subsequent analyses. The soil samples were treated with H202 for removal of organic matter (Jackson, I956, p. 35) and with NaOH for removal of amorphous material (Jackson, I956, p. 529). Free iron oxides were removed from the soil samples by the sodium dithionite-citrate-bicarbonate method of Aguilera and Jackson (I953). All samples were dispersed by 5 minutes I7 l8 boiling with 2% Na2C03 (Jackson, I956, p. 73). The¢£2 micron and <10.2 micron clay fractions were separated with a tube centrifuge (Jackson, I956, pp. l40-l4l). X-ray1Qiffraction A Norelco X-ray unit with wide-range goniometer and Brown recorder, CuKZ radiation (from a Cu targent X-ray tube through 3 Ni filter) Operated at 35 Kv. and 20 m amps. was used for scanning parallel-oriented samples on ceramic plates or glass slides. Powder diffraction patterns were obtained with a cylindrical X-ray diffraction camera with a diameter of ll4.59mm. and X-ray photographic film mounted on its circumference. To avoid quartz and feldspar reflections from ceramic plates, a prescribed amount of the clay from the Az-horizon was oriented on a glass slide (Jackson, I956, p. I84). Suitable amounts of the clay from the other samples were deposited on ceramic plates according to the method of Kinter and Diamond (I956). This latter technique orients the plate shaped clay particles, so that the 00l planes of most of the clay minerals are in a position to diffract X—rays thus enchancing this reflection over that of an ordinary powder pattern. l9 X-ray diagrams were recorded for Mg-saturated, glycerol-solvated,paralIel-oriented samples and for K-saturated specimens heated to 300°C and 550°C. Randomly oriented powder samples were prepared essentially according to the procedure outlined by Jackson (I956, p. I88). The powder sample was placed in.a thin capillary tube of soft glass and X-rayed for approximately three hours. The 060 reflections were determined from X¥ray powder diagrams and from X-ray diagrams obtained from oriented flakes. The latter method has been discussed by Rich (I957). An oriented clay layer was deposited in a small aluminum d;sh by evaporation of sodium-saturated and dialyzed clay suspension. The oriented clay layer was peeled off the dish and mounted at right angles to the X-ray beam at the center of‘a diffractometer. Reflections were scanned in the range of 55-650, 2 9 with CuK2 radiation. Greene-KellyiTest A test devised by Greene-Kelly (I953b) was used to distinguish octahedrally substituted montmorillonite from 20 tetrahedrally substituted montmorillonite from beidellite and nontronite. Samples of the Az-horizon fine clay, Wyoming bentonfte, and nontronite were lithium saturated by adding IN LiCT under suction. Excess salt was removed-by washing with distilled water, and the samples were heated to 250°C overnight. The sames were then saturated with glycerol and X-ray diagrams were recorded. Differential Thermal Analysis Differential thermal analyses of the clays were made using an instrument manufactured by the Robert L. Stone Company, Austin, Texas. The samples were heated at a rate of I3°C per minute to l000qC. The fine clay of the Az-horizon was thermally analyzed in air and in nitrogen gas to determine thermal effects due to oxidation of organic material. Infra-red Analysis This analysis was made using a Beckman lR7 infra-red Spectrophotometer. Films of the Na-saturated clays, having parallel orientation of basal cleavage planes (00I) were prepared by evaporating a suspension in aluminum dishes. The films were mounted at right angles to the infra—red beam and scanned from 600 to 400 Cm']. In the 800 to 2000 -I Cm region the samples were also examined after dissemination in KBr pellets at concentrations of 0.6 percent clay. Total Specific Surface Area This determination was made by means of the ethylene- glycol method as proposed by Bower and Geschwend (I952). Cation Exchange Capacity and Its RedUction By Potassium Fixation A method has been pr0posed to determine the vermiculite content in clay mixtures. (Submitted for publication by Jackson, M.L. and Alexiades in the Soil Sci. soc. Amer. Proc). This determination is based on the reduction of CEC due to irreversible collapse of K-saturated vermiculite heated overnight at ll0°C and subsequently saturated with NHL " Cation-exchange capacity (CEC) is determined as follows: An aliquot containing a l00-mg. organic matter-free, iron oxide-free, CaC03-free (Jackson, I956) clay sample is placed 22 in a 20-ml centrifuge tube and washed three times with NaOAc of pH 5 (warmed in hot water bath each time) to insure freedom from Na2C03 and CaC03. Ca saturation is obtained by washing five times with N CaCl2,washing once with H20 and five times with 99% methanol and finally exchanging the Ca with Mg by five washings of IO ml. each of N_MgCl2 solution (after 5 minutes of rotational shaking, the tube is filled with distilled water each time before centrifugation). The exchangeable Ca is determined on the extract diluted to l00 ml. with H20 (approximately 0.5 N MgCI2 solution) by means of flame emission at 5600 A- The determined Ca is expressed as m.e. per loo 9. of the oven-dry sample, and is designated as CEC (Ca/Mg). K saturation of a separate (or the same) l00-mg. sample is obtained by five washings with N_KCI, removal of exceSs salt by washing once with water and five times with 99% methanol. The K saturated sample is dried from 99% methanol in an oven at IIOOC. overnight. The unfixed K is exchanged with N_NHhCl solution as follows. The clay is soaked five minutes with rotary shaking in the first washing, and triturated with a rubber-tipped rod in the 23 second washing. Subsequent washings (for a total of five) are carried out by rotating the suspended sample for five minutes each time and centrifugation. The displaced K is determined by flame emission at 7740A. The determined K is expressed as m.e. per I00 9. of the oven-dry sample, and is designated CEC (K//NHu). The difference between these two determined CEC values gives the interlayer charge of vermiculite(as developed below) and vermiculite percent is given by: CEC(Ca/Mg) - CEC(K//NHu) x loo (1) I53.9 % vermiculite : The foregoing procedure was followed to determine total and reduced CEC of A2” and C-horizon soil clays, Wyoming bentonite, and nontronite. Difficulties were encountered in determining Ca in 0.5 N Mg Clz and K in 0.5 N NHuCI solutions by flame emission. These difficulties were avoided by determining Ca in 0.5 N Mg Clz on a Perkin-Elmer Model 303 Atomic absorption spectrOphotometer and by using I‘N NHhAc (pH 7), instead of IN NHhCI, to exchange K which was then determined by flame emission on a Coleman flame photometer. 24 Hydration Properties Oriented K, Na, Mg, and Ca-saturated samples of the 40.2“.clay fraction of the A2-horizon, Wyoming bentonite, and nontronite were X-rayed when moist (l00% R.H.) and after drying for 3 weeks at 40% R.H. and 20°C. Acid Dissolution for Determination of Tetrahedral Aluminum The acid dissolution method of Osthaus (I956) as modified (Brydon, g£_al, l96l) was used to determine tetrahedral Al. Ten 80 mg samples (larger samples if more clay were available) of AZ-horizon soil clay and 50 ml. of hot 30% HCl were added to Erlemeyer flasks in a boiling water bath. The samples were treated for 5, IS, 30 minutes, I, 2, 3, 4, 5, and 7 hours. After each treatment time the sample was immediately quenched in cold water. The super- natant solution was separated by centrifuge and added to a 200 ml. volumetric flask. The clay was washed twice with distilled water, and the washings were added to the flask. The aluminon method (Jackson, I958, pp. 297-300) was used to determine Al. The dissolved Al was subtracted from the 25 total AI initially present. The logarithm of the undissolved Al was plotted versus time and the straight line obtained at 3 hours was extrapolated to zero time to obtain the value of tetrahedral Al. Total Chemical Analysis The semi-micro system described by Jackson (I958, Chapter II) was followed for total chemical analysis of the clays- Total Si and Al were determined after fusion with Na2003 (Jackson I958, Fig. ll-2, p. 280). The molybdosilicic\ acid colorometric procedure (Jackson I958, pp. 294-297) and the aluminon method (Jackson, I958, pp. 297-300) were used for Si and Al, respectively. The HF decomposition procedure (Jackson I958, Fig. ll-l, p. 279) was followed for elemental analysis of K, Na, Mg, Ca, and Fe. Both K and Na were determined by flame emission, using a Coleman flame photometer; Mg and Ca were analyzed using a Perkin-Elmer Model 303 atomic absorption spectrOphotometer. Fe was determined using the o-phenanthroline method (Jackson I958, pp. 389-391). 26 Calculation of Montmorillonite Formulas The methods of calculation of mineral formulas used were essentially the same as those described by Ross and Hendricks (I945). The methods, including modifications, are explained and illustrated by a calculation of an analysis in the discussion of results. Electron Microsc0pe Examination An RCA electron microsc0pe was used to obtain transmission micrographs of (0.2}xclay particles from the Deer Park AZ-horizon. A Formvar specimen supporting film was formed essentially according to the method described by Jackson, I956, p. 422. The clay suspension was treated with H202 to remove organic growth, washed with NaCI to remove H202 and dialyzed until free of solutes according to the rapid dialysis method (Jackson I956, p. 423). The suspension was then sprayed onto the organic supporting film as a fine fog produced in a nebulizer. V. RESULTS AND DISCUSSION X-ranyiffraction Analyses of ClanyractiOns Figure I shows the locations of the soils from which the clay fraction (4.2/1) of Az-horizons were obtained. Figure 2 shows the corresponding X-ray diffraction diagrams of oriented Mg-saturated, glycerol-solvated clay films. The strong peaks at l7.7A indicate that montmorillonite is the dominant clay mineral in the.¢2,;clay fraction of the Az-horizons of all the sites sampled. Small peaks at about 3IA in the X-ray diagrams of samples I, 2, and 7 reflect regular interstratification of montmorillonite with chlorite or vermiculite. The similarity of the X-ray diffraction diagrams obtained from the soils at the various locations, points to a uniform weathering sequence resulting in formation of montmorillonite in the AZ-horizons of the sandy soils in this area of Northern Michigan. Deer Park sand The soil at location I was selected for further study. X-ray diffraction diagrams of the442p.clay fractions are 27 'I—"t’"' l 4---- _ ——¢p---1P--- I I J 3‘ III” dA 'Y 0 CO} d —- +-db- I _ ‘- 28 ' CHILE -1 - _.L ME Locations of sites sampled in prelImInary InvestIgatIon Figure l. 29 Figure 2. X-ray diffraction tracings oszZM clay fractions. Treatment: Mg-saturated, glycerol solvated. Samples taken from Az-horizons at correspondingly numbered locations shown in Figure l. Scale factor is 8. 30 2° 29 3.34 K I4.3 X 0 o o O O 3.24 A 4.25 A 7.I5 A IO.I A I7.7 A, I 3I shown in Figure 3 for the A2, B22, and C] horizons. The patterns designated A2 (tongues) refer to a sample taken just above the lower boundary of tongues of the AZ-horizon which extend to a depth of over 36 inches. .Ci;horizon The X-ray diffraction patterns of the Mg-saturated, glycerol-solvated clay of the Cl-horizon gives a peak at T7.7R which shows the presence of some montmorillonite. The peak at l0.l2 indicates illite. Disappearance of the l4.48 peak and re-enforcement of the l0.lg peak upon K-saturation and heating show that the l4.4X peak is due mainly to vermiculite, although some discrete cthrite is indicated by this peak after heating to 300°C. Considerable randomly interstratified chlorite is present, probably as chlorite-montmorillonite and chlorite-vermiculite, as indicated by the asymmetrical l0.IR peak. Disappearance of the 7.I58 peak at 550°C. shows the presence of kaolinite. Chemical analysis of this clay gave a K20 content of 3%. Therefore, this clay is composed of at least 26% illite which is not evident from the small Io.IX peak. Approximately 20% quartz and l0% kaolinite is also present. The remaining 40% is taken up in approximately 32 equal proportion by vermiculite, montmorillonite, and chlorite. Most of the chlorite present appears to be randomly interstratified with vermiculite and montmorillonite. 822 horizon The X-ray diffraction patterns of the Mg-saturated glycerol solvated clay of the 822 horizon shows a relatively strong peak at l4.38 and a small peak at 7.IX. Upon heating, the l4.33 peak shifts to a broad l0.IX peak, indicating Al-interlayering in vermiculite (Shawhney, I960); at 550°C. the 7.IR peak disappears, indicating kaolinite. No discrete chlorite is shown to be present, bUt some chlorite, randomly interstratified with vermiculite, may contribute to the broadening of the I0.IR peak upon heating. The presence of illite is not apparent from the X-ray patterns, but could be at least 25% according to chemical analysis of the clay fraction of a similar soil by Franzmeier (I962). The approximate amounts of other clay minerals are estimated as follows: vermiculite 35%, quartz 20%, randomly interstratified chlorite-vermiculite l5%, kaolinite 5%. A2 (tongues) The X-ray patterns show strong peaks at l7.7R and o I4.3X which collapse to l0.lA upon heating, indicating 33 that the clay in the tongues of this horizon is largely vermiculite and montmorillonite. Some kaolinite is also present. Approximate quantities of the clay minerals present are estimated as follows: montmorillonite 30%, vermiculite 20%, illite 20%, (from chemical analysis of the.40.2u fraction of the Az-horizon) quartz 20%, kaolinite 5%, and randomly interstratified chlorite-vermiculite and chlorite-montmorillonite 5%. AZ-horizon The X-ray patterns of this horizon give a strong I7.7R peak which collapses to I0.IR upon heating indicating that montmorillonite is about the only layer silicate present in the clay of this horizon. However, chemical analysis shows that the.£;u.clay of this horizon contains approximately 20% illite. Disappearance of the small 7.IA peak at 550°C. gives evidence of some kaolinite. Approximate quantities of clay minerals estimated are: montmorillonite 55%, illite 20%, quartz 20%, kaoTinite 5%. Assuming that the intensity of weathering increases from the C] horizon upward to the surface of the soil, the horizons of the profile, shown in Figure 3, clearly illustrate the weathering sequence: Figure 3. 34 X-ray tracings of 2 clay fraction of the indicated horizons of the Deer Park profile. Treatments: l) Mg-saturated, glycerol solvated; 2) K-saturated and heated to 300°C; 3) K-saturated and heated 550°C. Scale factor is 8. HR m¢ haw 35 3.37 K mas X 3.3a l I M K .— xi. A Lie 2 5.05 I DEER PARK A2 (tongues) l7.T K ”.3 3' IO.I TI W I WWM&AJ.WA _..I/ I z :"WW‘W‘MNMM 3 NUW~w VNM/“MwmeMwWJhxwflvwflf III I c. I, ,1" . WWII” ‘W'me WWWJ ‘ I II I l I I 'll I . " M m 2 M .vamw' wwlw/ )7 : 5 3 My] kTJMM‘I;W¥WmWIN#-J'Wf WW ”if Figure 3 36 illite randomly interstratified i““-‘~‘~ chlorite-vermiculite and vermiculite'1;:3montmorillonite chlorite-montmorillonite chlorite In the presence of adequate K, Al, and Mg, the reaction may be reversible. Beavers, §£_al, (I954) concluded that the dominant montmOrillonite component in the surface horizons of some Illinois soils is largely derived from loess blown in from the montmorillonitic soils and sediments of the Great Plains region. However, the clay mineral distribution in the Deer Park profile, especially the preponderance of vermiculite and montmorillonite in the A2 tongues at depths near the upper boundary of the C] horizon, indicates that the montmorillonite in the Az-horizon of this profile has formed in situ and is not due to deposition of montmorillonite- rich, wind-blown material. Clayyfraction smaller than 0.2 microns Figure 4 shows X-ray diffraction patterns of the fine clay obtained from the AZ-horizon of Deer Park sand. X—ray patterns of the fine clays from Wyoming montmorillonite, Figure 4. 37 X-ray diffraction tracings of<;0. 2A clay fractions. Treatments: l) Mg- -saturated glycerol-solvated; 2) K- saturated and heated to 300°C; 3) K- saturated and heated to 550°C. AC is Wyoming bentonite; B is nontronite; is A horizon clay; D is Ci-horizon clay. insets show 060 spacings of A, B, and C clays. Scale factor is 8. 38 WWW. INC-5 WW I 5 575/ I 5 5. A. ). Wfi Wm, . 5 55 - _ .- 55 55%. 555 .. 55.55 Wm - 2.. 5.5.5 5.555); Figure 1+ 39 nontronite and the CI-horizon of Deer Park and are included for comparison. All clay fractions are smaller than 0.2 micron. The AZ-horizon fine clay was deposited on glass slides to avoid quartz and feldspar reflections from ceramic plates. The other clays were oriented on ceramic plates. As shown in Figure A the X-ray diffraction patterns of the Az-horizon clay, nontronite, and Wyoming bentonite are similar and emphasize the dominant montmorillonite content of the Az-horizon clay. The X-ray pattern of the Mg-saturated, glycerol- solvated, A -horizon fine clay shows only little evidence 2 of a l0.lA peak for illite, although this clay contains 2.6% non-exchangable K20 and thus at least 20% illite, as was determined by chemical analysis. Disappearance of the small 7.IA peak at 550°C. indicates the presence of a small amount of kaolinite. Since this clay was deposited on a glass slide, the small 3.34A peak shown in the Mg-glycerol X-ray pattern reflects the presence of a small amount of quartz in the fine clay. The symmetry of the OOl peak and the rationality of the higher orders of the 001 plane indicate the absence of interstratification. The approximate amounts of clay minerals in the fine clay fraction of the Az-horizon are estimated to be: hO montmorillonite 70%, illite 22% (from chemical analysis), quartz approximately 7% and less than 3% kaolinite. The X-ray diffraction patterns from the4;0.2,k Cl-horizon clay appear similar to those from the¢;2/. CI-horizon clay shown in Figure 3 and discussed in the previous section. The X-ray patterns of the smafler- size clay indicate less quartz and kaolinite, these two minerals commonly being more concentrated in the coarse clay fraction. ‘The (060) spacings shown in Figure 3 are near l.50A and indicate that these clays are dioctahedral. Randomly oriented powder samples Table l shows X-ray powder data for the Mg-saturated A2¥horizon fine clay with interlayer glycerol. Calculated and observed values for beidellite, reported by Weir and Greene-Kelly (I962), are compared with values obtained for the fine clay of the Az-horizon. The diffraction data reported by Weir and Greene-Kelly include several reflections not observed in the powder diffraction patterns of the soil clay. Most of'these reflections are also in addition to those observed for montmorillonites (MacEqan, 196]). However, Greene-Kelly 41 X-ray diffraction powder data for Mg-A Table 1. horizon fine clay and Ca—B.J.M. beidellite wit inter- layer glycerol. o d(ca1e.)R d(obs.)A d(obs.)8 (orthorhombic Ca-B.J.M. Mg-Az- pseudo hexagonal) beidellite l horizon-clay I hkl -- 17.57 10 17.70" 10 001 4.48 \ 020,110 4.42 10 4.45 10 4.35 021 3.99 3.95 10 022 3.56 3.54 IO 023 3.15 3.17 2 3.33* 10 024 2.77 2.76 l 025 2.59 200,130 2.57 8 2.57 10 2.56 201 2.48 2.52 10 202 2.37 2.36 8 2.36 5 203 2.234 204 2.24 2 2.24 2 2.24 040,220 2.22 041 2.19 2 2.17 042 1.695 240,310,150 1.693 6 1.693 6 1.687 241 1.664 1.663 8 1.656 6 242 1.628 1.623 6 243 1.581 1.573 1 244 1.526 1.528 1 245. -- 1.498 10 1.494 10 060,330 L12 o o d(cale.)A D(obs.)A d(obs.)g (orthorhombic Ca-B.J.M. Mg-Az- pseudo hexagonal) beidellite l horizon-clay 1 hkl 1.295 1.293 8 1.293" 6 400,260 1.244 1.243 8 1.245 6 420,350,170 1.121 1.122 2 440,080 0.979 0.979 2 460,530,190 0.863 0.864 4 390,600 *lnterpreted as quartz reflection 43 was able to index these additional reflections as the 021, 201, 041, and 241 series by reference to an orthorhombic pseudo-hexagonal unit cell. Calculated values of spacings for such a cell are given in the first column of Table l and the appropriate indices in the sixth. The scale of relative intensities (I) ranges from a minimum of 0 to a maximum of 10. The agreement between calculated spacings and spacings observed for the B.J.M. beidellite and Az-horizon fine clay is good and gives evidence that the Az-horizon fine clay is a beidellite. Application of Greene-Kelly Test The X-ray diagrams in Figure 5 show that the Li- saturated Wyoming montmorillonite, when heated to 250°C. overnight, does not re-expand upon blycerol solvation but has reverted to an internally-compensated and non-expanding perphyllite-type mineral having an 001 spacing of 9.5A. Nontronite and the Az-horizon fine clay, similarly treated, expand to the 17.7A, 001 montmorillonite spacing. The broadening of the peaks in the IDA, 5A, and 3.5A regions, shown in the X-ray pattern of the Az-horizon fine clay may be attributed to the presence of illite in the sample or the collapse of some layers to 9.SA or to both. However, the symmetry of the 001 peak indicates that such collapse has been negligible. Figure 5. 2.1. X-ray diffraction tracings of 4.0.2;4c1ay fractions illustrating Greene-Kelly test. Treatments: Li-saturated, heated to 250°C, and then glycerol solvated. A is Wyoming bentonite; B is nontronite; C is Az-horizon clay. Scale factor is 8. 45 2’26 9 1747 A EF'ZO Figure 5 1 46 According to Greene-Kelly (1953), the positive test obtained in Wyoming montmorillonite indicates that in this clay mineral most of the charge is located in the octahedral layer. Upon dehydration, the predominantly octahedral charge attracts the small positive lithium ions into vacant octahedral sites, thus irreversibly reducing the charge with attendant loss of cation exchange capacity, surface area, and swelling prOperties. A negative test indicates that most of the charge is located in the tetrahedral layer. Lithium ions are not retained in the octahedral layer, and no loss of cation exchange capacity, surface area and swelling pr0perties occurs. The test serves to distinguish octahedrally substituted montmorillonites from beidellites and nontronites. Such montmorillonites give a positive test; beidellites and nontronites give a negative test. Therefore, according to the test, the Az-horizon fine clay is a beidellite. The dominant octahedral charge of Wyoming montmorillonite and the dominant tetrahedral charge of nontronite, which are known, confirm the effectiveness of the test. 47 Differential Thermal Analysis Figure 6 shows differential thermal analysis curves for the fine clays of Wyoming bentonite, nontronite, and the AZ-horizon. The carve of the Ca-B.J.M. beidellite (Weir and Greene-Kelly, 1962) is included for comparison. The soil clay was heated in air and in NZ to determine thermal effects due to oxidation of organic material. Table 2 contains data for peak temperatures. All of the curves show the large low-temperature endotherm between 100°C. and 3000C. which is characteristic of minerals belonging to the montmorillonite group. Wyoming montmorillonite gives an endothermic peak, due to dehydroxylation, at 7400C. and another at 915°C., due to breakdown of the dehydrated structure. Recrystallization to oxides and spinel gives an exotherm at 970°C. The AZ-horizon fine clay heated in N2, gives its main dehydroxylation peak at 560°C., a less pronounced endothermic peak at 8300C., and an exothermic peak at 960°C. The sample heated in air gives a broad exothermic peak at 380°C. which may be interpreted as being due to oxidation of organic matter remaining after the H202 treatment. Nontronite gives a dehydroxylation peak at 530°C. and an exothermic peak at 985°C. The high temperature endotherm is absent which may be explained by simultaneous decomposition of the dehydrated structure and recrystallization to hematite or magnetite. 48 Figure 6. Differential thermal analysis curves of the.50.20 clay fraction. A is Wyoming bentonite; B is Az-horizon clay (heated in N2); C is nontronite; D is A -horizon clay (heated in air); E is beidel ite from the Black Jack Mine, Idaho (Weir and Greene-Kelly, 1962). All samples Na-saturated except E which was Ca-saturated. 0 49 L l l l L l l I 100 200 300 400 500 600 700 800 900 1000 TEMPERATURE °C Figure 6 50 Table 3. Peak temperatures (DC) in the differential thermal analysis of 40.2p clays 4L Clay Mineral lst End. 2nd End. 3rd End. Exo. Na-Wyoming montmorillonite 135 740 915 970 Na-Az-horizon clay (in N2) 150 560 830 960 Na-Az-horizon clay (in air) 155 565 880 980 (not distinct) Na-nontronite 155 530 -- 985 Ca-B.J.M. beidellite 140,210 560 -- 970 51 The differences in dehydroxylation endotherm temperatures have been attributed to variation in composition of the octahedral layer (Grim and Rowland, 1942; Kelley and Page, 1942). According to these authors the temperature of dehydroxylation increases from 420°C. to 650°C. to 890°C. as the composition of the 2:1 octahedral layer varies from high Fe to high Al to high Mg. However, in many studies (Greene- Kelly, 1961), including the one reported here, montmorillonites containing little Fe but much Al give a main dehydroxylation peak near to that of nontronite. It also seems to be established that the presence of Fe alone does not lower the peak temperature since the presence of 1.9 Fe ions per unit cell in an iron-rich montmorillonite from Japan does not alter the thermal curve (Sudo and 0ta, 1952). Greene-Kelly (1961), comparing thermal curves of several dioctahedral clay minerals belonging to the montmorillonite group, directs attention to the fact that all except octahedrally-substituted montmorillonites give their main dehydroxylation peak in the 500°C. region and all except montmorillonite owe their charge predominantly to tetrahedral substitution. The same conclusion may be drawn from a comparison of the curves in Figure 6. It seems logical, therefore, to correlate decrease in dehydroxylation temperature with increase in tetrahedral substitution. 52 Infra-red Analysis Figure 7 and Table 3 show infra-red absorption data of the clays as indicated. The data obtained for the clays used in this experiment are compared to data taken from the literature for the clays marked with an asterisk. The broken curves in Figure 7 and the solid curves in Figure 8 are of samples examined as oriented films; the solid curves in Figure 7 are of samples disseminated in KBr pellets at concentrations of 0.6 percent before examination. Although some progress has been made in relating variations in infra-red absorption to variations in structure and composition of clay minerals within the montmorillonite group, much uncertainty still remains. (Farmer and Russell, 1964). Therefore, except for the assignment of certain data to types of bonds and vibrations shown in Table 3, no attempt is made here to analyze in detail the structure and composition of the clays used in this experiment by means of their infra-red spectra. Instead, the data obtained in this analysis are compared to data from the literature for clay minerals having a well-known structure and composition. Thus agreement of data for the Az-horizon fine clay with data for beidellite gives evidence as to the identity of the Az-horizon clay. Agreement of data for Wyoming bentonite and nontronite with data from the literature for the same clay minerals indicates the reliability of the comparisons. 53 Infra-red analysis curves of the<:0.2p clay fraction. Treatments: Na-saturated and air-dried. Solid curves represent spectra from KBr pellets; broken curves represent spectra from films. A is Wyoming bentonite; B is AZ-horizon clay; C is nontronite. Figure 7. PERCENT TRANSMISSION 1200 1100 54 l l 1 1000 900 855 FREQUENCY 10M") Figure 7 1 800 800 700 Figure 8. 55 Infra-red analysis curves of.40.2p clay fraction. Treatments: Na-saturated and air-dried. Curves represent spectra from KBr pellets. EIis Wyoming bentonite; B is AZ-horizon clay; C is nontronite. 56 z 9 (D ‘.’2 55 U) z < 0: 1‘ 3655‘ .— z m U 0: LIJ a 36 3.9-J 11 3585 l I J 1 I 4000 3800 3600 3400 FREQUENCY 10M") Figure 8 3200 57 As shown in Table 3, good agreement exists between data obtained in this study and data obtained from the literature for Wyoming bentonite and nontronite. The absorption bands for the Az-horizon fine clay, especially the strong Si-O peak at 1040cm'] and the O-H peak at 3655cm-1 correspond to (those given for beidellite in the literature. The doublet peak at 780cm'1 and 800cm"I in the spectrum of the fine clay of the Az-horizon indicates the presence of some quartz. This agreement provides evidence that the fine clay of the AZ-horizon is a beidellite. Cation Exchange Capacity and its Reduction by Potassium Fixation Table 4 shows the results of cation exchange capacity (CEC) and total specific surface measurements. The CEC were determined using Ca as the saturating cation and Mg C12 as the replacing reagent. The relatively low CEC of the CI-horizon fine clay indicates that this clay contains some non-expanding clay minerals. The specific surface of this clay appears high, which may be due to incomplete removal by Na0H treatment of amorphous material having a large specific surface. An experiment was carried out to investigate the potassium fixation capacity of the clays listed in Table 5. The clays were saturated with potassium, washed with methanol, and dried 58 “:0m_ .__omm:m ocm coEcmmv ocsumcou__ mzu Eocm coxMu mumok 30mm mmmm Nmmm mmmm comm mmmm mc_;ouocum I10 owe 0mm NNo mNo - - mam mmm mmm mmm om“ own um_nzon om“ mmm mmm com com m_m Nucmzo oom omk mmk mam mmw m5m NNm 0mm 0mm «km mmm mc56cmn mam New omm on Nam oam 1-0-x mNo_ 1 NNO. - m_o_ m_o_ w:o_ omo_ .qo. 0:0— :mo_ omo_ omo. - :mo. “mo. 6:5;05mtum .mo_ wmop om__ m___ om__ o-_m .0.nmu_COLuCOZ ®u_CO.._uCOZ #GHZCOqum muzconwcmm #0“: — —®_u_®m >m—U mCFEmEIF 951.803.? coNCOrWNK mbcmm co_qu0mn< bmc-mcmc_ oLu mo A_1Eov >ocmsoocm .m m_nmh 59 Table 4. Cation-exchange capacity and total specific surface of.¢0.2u clays CafTon exchange Sampje capacity Specific surface mé7100g. M‘fg. Azrhorizon 91 730 Wyoming bentonite 87 1137 Nontronite 102 995 Cl-horizon 25 402 Table 5. Cation exchange capacity and 001 spacing of 410.2;Iclays after K-saturation and heating to 1100C overnight and resaturation with NH4 *Redfiction in Sample Cation exchange cation exchange 001 spacing capacity, capacitygg a me/lOOg. me71009. (K) AZ-horizon 55 36 random inter- stratification of ollapsed (10 ) and expanded layers Nontronite 90 12 12.8 Wyoming bentonite 87 0 14.8 Cl-horizon 18 7 -- 60 at 110°C. overnight. The cation exchange capacities were then determined by exchanging the K with NHhAc as replacing reagent. The values obtained are shown in column 1 of Table 5. The difference between the CEC values before (Table 4) and after this treatment (Table 5) gives the reduction of CEC shown in column 2, Table 5. Samples of the clays were also oriented on ceramic plates, saturated with KCl, washed with 99% methanol to remove excess salt, and then dried at 110°C. overnight. After oven drying, the clay films were saturated with NH4 by adding NHhCl under suction, washed with water to remove excess salt, and then X-rayed at 100% relative humidity. The 001 spacings are shown in Figure 9 and Table 5. These data show that the reduction in CEC is correlated with a decrease in 001 spacings. The decreased CEC, therefore, may be attributed to incomplete expansion of K-saturated, oven-dried 2:1 swelling clays upon NH4 saturation, resulting in incomplete replacement of NH# for K in the exchange reaction. Thus, under the conditions of this experiment, the K not exchanged by NHA may be designated as fixed K. It is well known that vermiculite collapses irreversibly upon K saturation and drying. Therefore, the reduction in the CEC of the Cl-horizon fine clay may be explained by the irreversible collapse of vermiculite shown to be present in X-ray diffraction tracings of K-saturated, heated, and NH4 resaturated clays. A is Az-horizon clay; B is nontronite; C is Wyoming bentonite. Scale factor is 8. Figure 9. 62 Figure 9 2° 20 63 this clay by X-ray data (Figure 4). If the assumption is made that this irreversible collapse is restricted to vermiculite the amount of this clay mineral present in the sample could be calculated from the reduction in cation exchange capacity. However, a comparison of the reduction inCEC of the fine clays from the Az-horizon, nontronite, and Wyoming bentOnite, shows that the CEC of Wyoming bentonite remains the same, whereas the CEC of the Az-horizon fine clay and nontronite are reduced. Since these data were consistent in three repeated determinations, and since the X-ray tracings of neither the AZ-horizon clay nor the nontronite (Figure 4) indicate the presence of vermiculite in these clays, experimental error and irreversible collapse due to vermiculite cannot account for the reduced CEC values observed. Therefore, the assumptionthat K—fixation is restricted to vermiculite is in error and cannot be used as the basis for calculating the, amount of vermiculite in the sample. Further analysis of the data in Table 5 and Figure 9 shows that decrease of CEC and of 001 spacing occurs in the A2-horizon clay and in nontronite which have high total and tetrahedral charge. The CEC of the Wyoming bentonite, which has a lower and predominantly octahedral charge is not affected by the treatment. Therefore, the logical conclusion is that 64 K-fixation and reduction in CEC is correlated with the source and amount of the charge on the clay mineral. The tetrahedral charge, being close to the interlamellar K ions has a greater capacity to fix these ions than has the octahedral charge which is farther removed from the interlamellar surfaces. The higher total charge of the fine clay of the AZ-horizon may account for the greater reduction of its CEC as compared to that of nontronite. Hydration Properties: Basal §9acings With Water Figure 10 showsX-ray diffraction patterns at the relative humidities indicated for AZ-horizon soil clay, Wyoming bentonite and nontronite saturated with different cations. Table 6 contains the corresponding basal spacings in Angstrom units (2). Random interstratification due to interlayering of different ”hydrates” with different spacings is shown in the broadening of the 001 peaks which are designated (br) in Table 6. Assuming that the broadened peaks resulted from random alternation of two different hydrates, the centers of the peak represent the approximate weighted mean of the spacings actually present. 65 Figure 10. X-ray tracings showing basal spacings of.;0.2}1c1ay fraction at 100% and at 40% relative humidity. Treatments: Saturated with Na(1), K(2), Mg.(3), Ca(4). A is A 7horizon clay; B is Wyoming benton te; C is nontronite. Scale factor is 8 66 100 °/o R. H. 10 F 67 m.:_ cn o.:_ o.N_ m.:_ :.w_ _.m_ w.:_ to :.m_ mumcocucoz m.:_ m.:_ m.N_ :.N_ m.m_ m.m_ m.:_ up N.m_ mu_co___coeucoE mc_Eo>3 tn m.m_ 56 m.:_ up m.m_ _.q_ .6 N.N_ m.m_ o.m_ m.m_ c0N_co;-~< mu mi, M 021 mu sz x m2 cowumo LWWm_couc_ Doom ..I.m x0: ooow ..I.m 500. >650 A.I.mv mm_u_oE:; m>_um_oc ucmcomm_o ozu um m>m_o.1w.o.v ogu >3 oxmua: coumz mc_csp 60800» A__ox1ocomco ocm c_03 EOE» mumo 1 m Aomm_v .m._.< m_mcoc_z mococomom 80cm mumo 1 N >onum m_:u cm ooc_muno moon 1 _ Na.o oo.~ No.0 mm._ m_»o 03.0 oo.m c65m655mmz5 .6_055tmu Fmficocucoz . mmu_cocucoz mm.o o_.N N_.o mm._ mo.o om.o cm.m ecumc_:mm3 .b_o_mcmu Nmu_coLucoz. om.o oo.~ m_.o ma.o oa._ mm.o mo.m Anew. >566. mzumZm_I c_v :c_mcoom_3 .oz m:.o mo.~ mN.o m_.o No._ m:.o mm.m Apcmm 5.6a Lama :55 mea___ea_mm _emm_eu_z .oz m:.o _o.~ .o.o No.0 mm._ mm.o m:.m Osmo— moc_z xomw xom_m Nm.o No.~ mm.o e_.o mm._ mo.o mm.m .005e60cea me556>3- mc_Eo>3 .cOua: mmu_co___coeucoz ~m.o mo.~ mm.o m_.o mm._ mo.o Nm.m ~655coaeea aa_eo>z z -w a: mu _< _< _m mcoWDmo ~+ m+ m+ m+ :+ >u__m005 omcmsoxo mo co_umc_pcooU co_umc_otooo ocm m.aEmm mo_com moco_m>_nqm pmcmozmuoo _mcpo;wcuoh1 mmu_co___coeucoE _mcom;muoo_o to» mm_:Ecou .o. o.oma Figure 12: Electron micrographs of the fine clay fraction of the Deer Park AZ-horizon Figure 13. J Electron micrographs of the fine clay fraction of the Deer Park Az-horizon VI. GENERAL DISCUSSION X-ray diffraction tracings show that a montmorillonite is the predominant layer silicate in the4;2;ifraction of the AZ-horizons of some sandy soils in Northern Michigan. The tracings of the.:2/u fraction of the Deer Park profile indicate that the monrmorillonite has formed in situ and these tracings illustrate the following weathering sequence. Illite randomly interstraglfl;8\\\\\\\\\ . _ . . 5 . . chlorite vermiculite and vermiculite chlorite-montmorl 1 V chlorite .___\ 1 . 5__montmorillonite This weathering sequence is prevalent in Northern Michigan resulting in accumulation of montmorillonite in the Az-horizons ofsandy soils of various ages ains evident in the Deer Park sand at location No. 1 (approximately 2250 years) and in the Blue Lake sand at locations No. 6 and 7 (approximately 10,000 years). The 060 spacing and the structural formula indicate that this montmorillonite is dioctahedral. The structural formula also shows that over 90 percent of the total charge arises 89 90 from substitution of Al for Si in the tetrahedral layer and the remainder from substitution of Mg for Al anf Fe in the octahedral layer. Greene-Kelly test confirms these charge relationships. Therefore, this material may be interpreted as a beidellitic member of the montmorillonite group. Data from X-ray powder diffraction and differential thermal analysis of the fine clay of the Az-horizon agree with those reported by Weir and Greene-Kelly (1962) for the typical end-member beidellite from the Black Jack Mine, Idaho. Data from infra-red analysis correspond to those reported for beidellite by Farmer and Russell (1964). These agreements provide additional evidence for the beidellite interpretation. The szaturation, heating, and NHq-resaturation treatments reduce the cation exchange capacities (CEC) of both the fine clay of the AZ-horizon and nontronite but do rot affect the CEC of Wyoming bentonite. This indicates a correlation of increased K-fixation with increased tetrahedral and total charge. Because none of the other analyses showed the presence of vermiculite in these clays, the determination of vermiculite in clay mixtures by the method of CEC reduction cannot be accepted, unless vermiculite is defined by this method alone. 91 Hydration studies show that at 100% relative humidity the fine clay of the Az-horizon forms a somewhat smaller spacing with the Na, Mg, and Ca than do Wyoming bentonite and nontronite saturated with the same ion species. This smaller spacing may be related to high total and tetrahedral charge. The limited swelling of the fine clay of the A2- horizon and also the tendency to fix potassium appear to be characteristic of beidellite in soils (Marshall 1936, 1949). The chemical determination of tetrahedral aluminum by acid dissolution gave anomalous results. These may be related to relatively rapid dissolution of tetrahedral Al due to structural defects of the montmorillonite in the fine clay of the AZ-horizon. Earlier beidellite interpretations have been questioned because X-ray diffraction and differential thermal analyses showed that in many cases the so-called beidellite was an interlayered mixture of illite and montmorillonite (Grim 1953). In view of the high illite content (22% calculated on a muscovite basis) the same doubt may be raised regarding the beidellite interpretation in this study. However, in this study X-ray diffraction analyses do not indicate random or regular interstratification of montmorillonite with illite as shown by the symmetry of the 001 peak and the rationality of the higher orders of the 001 plane. 92 The apparent absence of illite-montmorillonite inter- stratification may be explained by the presence of mica in montmorillonite as "cores.'l Such cores have been identified in electron micrographs by Venkata Raman and Jackson (1964) and in photomicrographs by Mortland (1958) of vermiculite in which the smooth surfaces of the mica cores could be distinguished from the rough surface of the surrounding vermiculite. A similar arrangement of unweathered mica cores in montmorillonite may be visualized according to the simplified diagram shown below. ica cores \_,,5__/5’___»::4;1 $87‘2:1 silicate layer, 9.53 W °\\expanded montmorillonite spacing Because there is a tendency for interlayer constancy of the sum of interlayer sorption surface and K-occupied interlayers, any given mica interlayer segment tends to remain completely filled with K or else becomes completely affected by interlayer swelling (Mehra and Jackson, 1959). If these conditions prevail, mica and montmorillonite are present as non-interstratified, discrete phases (Mortland, 1958). 93 This postulated arrangement of montmorillonite and mica may contribute to the tendency for K fixation, since the mica cores maintain alignment of the montmorillonite silica sheet cavities in which K may be strongly bonded. The presence of illite as mica cores also justifies the calculation of percent illite using the K20 content of muscovite. At this stage of weathering similar ”islands” of chlorite may have weathered to expanding layer silicates, since chlorite appears to be less stable than illite under these weathering conditions (Droste, §£_al, 1962; and Bayliss and Loughnan, 1963). The apparent absence of vermiculite (commonly the first weathered product of mica) while mica is still present, indicates that this mica is very resistant to further weathering. Weir and Greene-Kelly (1962) have recommended four requirements for the use of the name beidellite: l. Beidellites are montmorillonite minerals, and the term should only be used as a species name for the appr0priate member of the montmorillonite group. 94 2. The term beidellite should be used for the aluminum rich members of the montmorillonite- beidellite series of minerals, as pr0posed by Ross and Hendricks (1945), but the composition of the ideal end-member should be restricted to that of an exactly dioctahedral mineral, as prOposed by MacEwan (1951) and Brown (1955). 3. Beidellites and montmorillonites should be divided at the composition at which the lattice charges from octahedral and tetrahedral substitution equal one another (Greene-Kelly, 1955). 4. Naturally occurring beidellite specimens should not ideally contain non-exchangable potassium. The first three of these requirements are_fu1filled by thezL0.2u clay fraction of the Deer Park AZ-horizon, but the ideal condition in the fourth requirement is not satisfied. However, X-ray data indicate no interstratification of illite with montmorillonite, and considering the argument presented previously concerning the presence of illite as mica cores, the montmorillonite in this clay fraction may I, 95 be regarded as a discrete phase containing no non-exchangeable potassium. Therefore, the montmorillonite present in the 20.2;iclay fraction of the Deer Park Az-horizon is classified as the species beidellite of the montmorillonite-beidellite series. VII. CONCLUSIONS 1. Illite, chlorite, and randomly interstratified chlorite-vermicute and chlorite-montmorillonite have weathered in situ through vermiculite to a dioctahedral montmorillonite in sandy soils of various ages over a relatively large area in Northern Michigan. 2. Because this montmorillonite mineral is rich in aluminum and has a high charge of which over 90 percent originates from substitution of Al for Si in the tetrahedral layer, it is classified as a beidellite of formula, 1:“1.62 Fe.18 149.21fl 1211.55 513.55 010 EH32 M149 3. This beidellite has a strong tendency to fix potassium. 4. Vermiculite contents in clay mineral mixtures cannot be calculated from reduction in cation exchange capacities following K saturation, heating to 110°C., and resaturation with NH“. 96 97 5. The acid dissolution technique for determination of tetrahedral aluminum as used in this investigation does not apply to the beidellite studied. LITERATURE CITED Aguilera, N. H., and Jackson, M. L. 1953. Iron oxide removal from soils and clays. Soil Sci. Soc. Amer. Proc. 17:359-364. Barshad, I. 1955. Chemistry of the Soil. ed. by Bear, F. E. Reinhold Publishing Corporation, New York. Ch. 1:1-52. Barshad, l. 1959.. Factors affecting clay formation. Clays and Clay Minerals. Pergamon Press, New York. 6:110-132. Baylin, P. and Loughnan, F. C. 1964. Mineralogical trans- formations accompanying the chemical weathering of clay-slates from New South Wales. Clay Min. Bull. 353-361. Beavers, H. H., Johns, W. D., Grimm, R. E. and Odell, R.T. 1955. Clay minerals in some Illinois soils developed from loess and till under grass vegetation. Clays and Clay Minerals. Nat. Ac. Sci., Nat. Res. Couno, Publ. 395:356-373. ' Bower, C. A. and Geschwend, F. B. 1952. Ethylene glycol retention by soils as a measure of surface area and interlayer swelling. Soil Sci. Soc. Amer. Proc. 16: 342-345. Bradley, W. F. and Weaver C. E. 1956. llChlorite-vermiculite.” Amer. Min. 41:497-505. - Brindley, G. E. 1955. Clays and Clay Technology. Calif. Divis. Mines, Bull. 169:33. ~ Brown, G. 1955. Report of the Clay Minerals Group Sub- committee on Nomenclature of Clay Minerals. Clay Min. Bull. 2:294. . Brown, 8. E. and Jackson, M. L. 1958. Clay mineral distribution in the Hiawatha sandy soils of Northern Wisconsin. Clays and Clay Minerals. Nat. Ac. Sci., Nat. Res. Counc. Publ. 566:213-226. Brydon, J. E., Clark, J. S. and Osborne, V. 1961. Dioctahedral chlorite. Can. Mineralogist 6:595-609. 98 99 Brydon, J. E. 1964. The alteration of mica and chlorite during podzolization. Abstracts of “The Clay Minerals SOciety”. Wisconsin, p. 5. ' Caillere, S. and Henin, S. 1957. Pr0positions pour normaliser la nomenclature des mineraux argileux. Bull. Groupe franc. Aryiles. 9:77-83. Damour, A. A. and Salvetat , , 1847. "H drosilicate d'alumine trouve a Montmor1llon (Vienne ” Ann. Chim. (Phys.), 111, 21:376-383. Droste, J. B. and Thorin, J. C. 1958. Alteration of clay minerals in Illinoian till by weathering: LBull. Geol. Soc. Amer. 69:61-67. Droste, J. B., Bhattacharya, N. and Sunderman, J. A. 1962. Clay mineral alteration in some Indiana soils. Clays and Clay Minerals. Pergamon Press, New York. 11:329-341. Farmer, V. C. and Russell, J. D. 1964. The infra-red spectra of layer silicates. 20:1149-1173. Foster, M. D. 1953. Geochemical studies of clay minerals. II. Relation between ionic substitution and swelling in montmorillonites. Am. Min. 38:994-1006. Foster, M. D. 1954. The relation between ”illite“, beidellite, and montmorillonite. Clays and Clay Minerals. Nat. Ac. Sci., Nat. Res. Counc. Publ. 327:386. Franke-Kamenetsky, V. A. 1958. The Investigation and Utilization of Clays (D.P.‘Bobrovnik et al, ed.) Izdatelstvo Lvov Univ. :713. Franzmeier, D. P. 1962. A chronosequence of podzols in Northern Michigan. Ph.D. thesis, Michigan State University, East Lansing. Franzmeier, D. P. and Whiteside, E. P. 1963. A chrono- ' sequence of Podzols in Northern Michigan. Agr. Exp. St. Quar. Bull., Mich. St. Univ. V. 4 , No. l, pp.l-57. Frye, J. C., William, H. B., and Glass, H. D. Gumbotil, Accretion-gley, and the we thering profile: 111. State. Geol. Survey Circ. 295:1-3 . 100 Garrels, R. M. and Howard, P. 1959. Reactions of feldspar and mica with water at low temperature and pressure. Clays88 and Clay Minerals. Pergamon Press, New York. Greene-Kelly, R. 1953b. Identification of montmorillonoids. J. Soil Sci. 4:233-237. Greene-Kelly, R. 1955. Dehydration of the montmorillonite minerals. Mineral Mag. 30: 604. Greene-Kelly, R. 1961. The montmorillonite minerals in The Differential Thermal Investigation of Clays. Mackenzie, R.C. ed. The Min. Soc., London, Ch. 5:140-165. Grim, R. E. and Rowlands, R. E. 1942. Differential thermal analysis of clay minerals and other hydrous materials. Am. Mineral. 27: 801. Grim, R. E. 1953. Clay Mineraology. McGraw-Hill Book Co., New York :39. Grim, R. E. and Kulbicki, G. 1961. Montmorillonite: high temperature reactions and classification. Am. Min. 46:1329-1369. Harrison, J. L., and Murray, H. H. 1959. Clay mineral stability and formation during weathering. Cla s and Clay Minerals. Pergamon Press, New York. 6:14 -153. Hendricks, S. B. 1942. Lattice structure of clay minerals and some properties of clays. J. Geol. 50: 276- -290. Hofmann, U., Endell, K. and Wilm, D. 1933. Kristallstruktur and Quelling von Montmorillonit. Zeitschr. Kristallographie. 86 (A): 340- 347. Hofmann, U., Ende11,K. and Wilm, D. 1934. RontgenographisChe and kolloidschemische Untersuchungen uber Ton: Zeitshr. Angew. Chemie. 47: 539- 547 Jackson, M. L. and Sherman, G. D. 1953. Chemical weathering of minerals in soil. Advances in Agronomy. Academic Press, New York. 55219-318. Jackson, M. L. 1956. Soil chemical analysis - Advanced Course. 'Univ. of Wisconsin. 101 Jackson, M. L. 1958. Soil Chemical Analysis. Prentice Hall, Englewood Cliffs, New Jersey. Jackson, M. L. 1959. Frequency distribution of clay minerals in major great soil groups as related to the factors of soil formation. Clays and Clay Minerals. Pergamon Press, New York. 6:133-143. Jackson, M. L. 1964. Chemical Composition of Soils, in Chemistry of the Soil. Bear, F. E. ed. Reinhold Publ. Corp., New York. Kelley, W. P. and Page, J. B. 1942. Criteria for the identification of the constituents of soil colloids. Soil Sci. Soc. Am. Proc. 7:175-182. Kelly, K. L. and Judd, D. B. 1955. The lSCC-NBS method of designating colors and a dictionary of color names. National Bureau of Standards Circular 553. U.S. Govt. Ptg. Off., Washington. Kinter, E. B., and Diamond, S. 1956. A new method for preparation and treatment of oriented-aggregrate specimens of soil clays for X-ray diffraction analysis. Soil Sci. 81:111-120. Kodoma, H. and Brydon, J. E. 1964. Interstratified montmorillonite-mica clays from subsoils of the Prairie Provinces of Canada. Abstracts of “The Clay Minerals Society“. Wisconsin, p. 16. Larsen, E. S. and Wherry, E. T. 1925. Beidellite, a new mineral name. Journ. Wash. Acad. Sci. 15:465. Lippmann, F. 1954. llCorrensite.” Heidelberg. Beitr. Min. 4:130-134. Mac-Ewan, D. M. C. 1951. The montmorillonite minerals (montmorillonoids) X-ray Identification and Structure of Clay Minerals. Mineral. Soc., London, Ch. 4:86. Mackenzie, R. C. 1957. Modern methods for studying clays. Agrochimica, 1:308. Mackenzie, R. C. 1960. The evaluation of clay mineral competition with particular reference to smectites. Silicates lndustr. 25:12-18, 71-75. 102 Marshall, C. E. 1935. Lazer lattices and base-exchange clays. Z. Krist. 91: 33-449. E. Marshall, C. 1936. Soil science and mineralogy. Soil Sci Soc. Am. Proc. 1:23-31. Marshall, C. E. 1949. The Colloid Chemistry of the Silicate Minerals. Ac. Press, New York. Mehra, 0. P., and Jackson, M. L. 1959. Constancy of the sum of mica unit cell potassium surface and interlayer sorption surface in vermiculite-illite clays. Soil Sci. Soc. Am. Proc. 23:101-105. Mortland, M. M. 1958. Kinetics of potassium release from biotite. Soil Sci.-Soc. Am. Proc. 22:503-508. Murray, H. H. and Leininger, R. K. 1956. Effect of weathering on clay minerals. Clays and Clay Minerals. Nat. Ac. Sci. Nat. Res. Counc. Publ. 395:322-336. Nagelschmidt, G. and Hicks, D. 1943. The mica of certain coal-measure shales in South Wales.. Mineral. Mag. 26:297-303. ‘ Osthaus, B. 1956. Kinetic studies on montmorillonites and nontronite by the acid dissolution technique. Clays and Clay Minerals. U. S. Nat'l. Acad. Sci.-Nat'1. Res. Council, Pub. 456:301-321. ~ Reference Clay Minerals 1951. A.|.P. Research Project 49. American Petroleum Institute, Columbia University, New York. No. 7:51-55. Rich, C. I. 1957. Determination of (060) reflections of clay minerals by means of counter type X-ray diffraction instruments. Amer. Min. 42:569-570. Ross, C. S., and Shannon, E. V. 1925. ChemicaJ composition and Optical pro erties of beidellite. Washington Acad. Sci. Journ. 15: 67-468. Ross, C. S. and Hendricks, S. B. 1945. Minerals of the montmorillonite group. U. S. Geol. Survey Profess. Paper 2058:23-77. 103 Ross, C. S. 1959. Review of the relationships in the montmorillonite group of clay minerals. Clays and Clay Minerals. Pergamon Press, London:225. Shannon, E. V. 1924. Notes on the mineralogy of three gouge clays from precious metal veins. Proc. U.S. Nat. Museum, 62, art. 15. Shawhney, B. L. 1960. Aluminum interlayers in clay minerals. Trans. Intern. Cong. Soil Sci. 7th Cong. Madison. 4:476-481. - Soil Survey Staff. 1951. Soil Survey Manual. U.S.D.A. Handbook No- 18. U.S. Govt. Ptg. Off. Washington. Strunz, H. 1957. Mineralogische Tabellen, 3rd ed., Akad. Verlagsgesellschaft, Leipzig. Sudo, T. and 0ta, S. 1952. Iron-rich montmorillonite from 0ya-ishi. J. Geol. Soc. Japan. 58:487-490. Venkata Raman, K. V” and Jackson, M. L. 1964. Vermiculite Surface Morpholo y. Clays and Clay Minerals. Pergamon Press, New York: 23-429. Weir, A. H. and Greene-Kelly, R. 1962. Beidellite. Am. Min. 47:137-146. White, J. L. 1950. Transformations of illite into montmorillonite. Soil Sci. Soc. Amer. Proc. 15:129-133.