SELECTIVE DISSOLUTION AND DIFFERENTIAL INFRARED SPECTROSCOPY FOR CHARACTERIZATION OF 'AMORPHOUS' CONSTITUENTS IN SOIL CLAYS

Transcription

1 Clay Minerals (1970) 8, 241. SELECTIVE DISSOLUTION AND DIFFERENTIAL INFRARED SPECTROSCOPY FOR CHARACTERIZATION OF 'AMORPHOUS' CONSTITUENTS IN SOIL CLAYS K. WADA AND D. J. GREENLAND Faculty of Agriculture, Kyushu University, Fukuoka, Japan Department of Agricultural Biochemistry and Soil Science, Waite Agricultural Research Institute, University of Adelaide, South Australia (Received 17 October 1969) ABSTRACT : Poorly crystalline inorganic materials were removed from soil clays of different origin and mineral composition by successive treatments with sodium dithionite, 2 ~ Na2CO3 and 0"5 n NaOH. Techniques are described whereby difference spectra representing the infrared absorption of the materials removed by the treatments can be obtained. These spectra indicate that the materials dissolved are related to the mineral composition of the soil clay. Dissolution of allophanes of different composition was only proved for two volcanic ash soil clays in which these dominated. Layer silicates, probably including kaolin, were dissolved from the clay fractions of the red-brown earth and krasnozem studied. In addition, alumina-rich gel-like material and gibbsite, but no allophanes, were dissolved from another volcanic ash soil clay in which gibbsite and layer silicates are present in considerable amounts. INTRODUCTION 'Amorphous' clays are important as one of the most surface reactive materials in soils. Little direct evidence, however, has been obtained of the presence of such amorphous constituents in soils in general, except for allophanes in some volcanic ash soils. Recognition of the presence of such material is hampered by lack of appropriate methods of identification. Differential dissolution techniques (also termed 'selective dissolution analysis' or SDA) involving treatments with alkalis have been used to obtain some estimates of the 'amorphous' constituents in soil clays (Hashimoto & Jackson, 1960; Mitchell & Farmer, 1962; Follet et al., 1965a). Hashimoto & Jackson (1960) concluded frem their study on dissolution of clays with NaOH that substantial amounts of allophanes, free silica and alumina were brought into solution by boiling for 2-5 minutes with 0.5 n NaOH, whereas only small amounts of crystalline clay minerals were dissolved during the same digestion period. Follet et al. (1965a) studied alkali-soluble materials from a podsol and a noncalcareous humic gley soil in which crystalline layer silicates predominated. Effects of graded alkali dissolution were followed by various analyses. Their conclusion

2 242 K. Wada and D. J. Greenland was that the alkali-soluble material consisted of poorly ordered aluminosilicates of different composition, present primarily as coatings on the surface of the wellcrystallized clay particles, but not as separate, particulate phases. The mineral identity of these poorly ordered aluminosilicates, however, was not positively indicated. The association of 'amorphous' silicates with 'free' iron oxides has also been inferred from dissolution of silicon and aluminium in addition to iron during removal of extractable iron oxides from soil clays (Mitchell & Mackenzie, 1954; Mehra & Jackson, 1960; Follet et al., 1965b; Yoshinaga, 1966; Weaver, Syers & Jackson, 1968). Again, critical data concerning the nature of these 'amorphous' components have not been presented. Follet et al. (1965b) concluded that the dithionite-soluble materials from the two soil clays were amorphous, ferruginous complexes containing quantities of silica and alumina. They considered electrondense granules seen in the electron microscope to be the principal form of these ferruginous complexes. X-ray examination of the clays before and after the dithionite treatment gave no evidence of any attack by dithionite on the clay minerals. They failed to obtain electron diffraction patterns from the granules, which substantiated their conclusion that they were amorphous. In the present work removal of components of soil clays of different origin and mineral composition by dithionite and alkali treatment, has been studied by determining the weight loss and obtaining infrared spectra. The infrared spectra of the soluble fractions were obtained as difference spectra between those of the clays before and after the respective treatments. Infrared spectroscopy was selected for this particular characterization because unlike X-ray diffraction, the contributions of the soluble and insoluble components to any absorption band are additive, and the double-beam spectrophotometer enables the difference between the two samples to be recorded very simply. The spectra alone provide much information regarding the mineralogy and composition of a rather wide range of materials, from amorphous to well-crystallized silicates, aluminium and iron oxides and hydroxides. MATERIALS AND METHODS Brief descriptions of the soils used are given in Table 1. The clay fractions (<~2 t~) were dispersed by sonic wave treatment in distilled water without H202 treatment, and collected from the appropriate depths of the suspensions after sedimentation. Addition of HC1 and NaOH to assist the dispersion was made for the 1041 and W-106 soils, respectively. Sodium chloride was then added and the clays kept as the flocculated suspensions. The clay yields expressed as percentages of the whole air-dry soil were 12, 14, 44, 40 and 54% for the 1041, Mt Schank, W-106, Glencoe and Urrbrae soils, respectively. The low clay yield from sample 1041 which has a significant content of organic matter resulted from omission of predigestion with H202. The outline of the differential dissolution--infrared spectroscopic analysis is shown schematically in Fig. 1. The following experimental procedure was adopted.

3 "Amorphous' constituents in soil clays TABLE 1. Sample description 243 Number or Locality; soil type; abbreviation and horizon Major clay minerals* 1041 Choyo, Kumamoto, Japan; A>~Im Mt Schank volcanic ash soil; B (30-80 cm) Mt Schank, S.A., Australia; A W-106 volcanic ash soil; B (30-75 cm) Kuroishibaru, Kumamoto, Japan; G>I>V, C>K Glencoe volcanic ash soil; B ( cm) Glencoe, S.A., Australia; krasnozem from Pleistocene basalt; B (20-60 cm) K > M > I Urrbrae Urrbrae, S.A., Australia; red-brown earth from shale; subsoil K>I>V *A, allophane; C, chlorite; G, gibbsite; Im, 'imogolite'; I, illite; K, kaolinite; M, montmorillonite; V, vermiculite. Aliquots of the clay suspension containing exactly the same amount of clay, usually mg, were treated in 10 ml centrifuge tubes with fitted glass stoppers. In most cases, iron oxide was extracted twice by the method of Mehra & Jackson (1960); 4 ml of 0.3 M Na citrate, 0"5 ml of 1 M NaHCO3 and about 0"1 g of Na2S~O4 were added for each treatment, and the extraction was carried out at 80~ for 15 min. After the second extraction, the clays were washed once with a mixture of 5 ml of 0"3 M Na citrate and 1 ml of acetone. Par+ removed Pert Treafmen+ remainin~~ 0 O... O 9 IR Spec+ro ond weigh+ ross 0... I FIG. 1. A schematic representation of differential dissolution--infrared spectroscopic analysis.

4 244 K. Wada and D. J. Greenland The clays were then treated successively with 2% Na~CO~ (Jackson, 1956) and 0.5 N NaOH (Hashimoto & Jackson, 1960); 8 ml of the respective solutions were used for each treatment. The boiling for 5 and 2-5 rain in 2% Na2COs and 0-5 N NaOH solutions in the original procedures was replaced by heating at 90~ for 15 min in a water bath. The relationship between the weight loss and the time for the treatment with 0.5 N NaOH was studied for the 1041 and Urrbrae clays, and the results are shown in Table 2. The rate of dissolution decreased rather rapidly and the difference spectra showed no significant changes except in total absorbance, between those treated for 5 rain in the single and for 30 rain in the double treatments. Therefore, a single 15 min treatment was normally adopted for the present study, while two 15 rain treatments were applied for the W-106 clay, because gibbsite, which is only difficultly soluble, is a major component. TABLE 2. Effect of number and time of treatment on dissolution of soil clays in 0"5 s NaOH at 90~ Number of Total time for Soil clays treatment treatment Weight loss 1041 Urrbrae min % All the treated clays and the original clays without any dissolution treatment were then washed three times with 8 ml of 1 M NaCH~COO adjusted to ph 5 with acetic acid. This was done to avoid secondary effects of the dissolution treatments on the measurements due to changes in the status of exchangeable cations. The excess NaCI-I~COO was removed by successive washings with 5 ml of water-methanol (1 : 1), methanol-acetone (1 : 1) and acetone. The experiments were designed to determine the nature of the materials dissolved by the treatments applied, and it was therefore essential that no dispersed clay be lost during the washing and decanting procedures. Care was taken to ensure that the clay remained flocculated throughout the treatments and washings. At the end of the washing procedure the clays in the centrifuge tubes were dried at 105~ overnight and weighed. The tube weight was obtained for the tube cleaned after all the treatments and measurements. The infrared spectra were obtained from KBr discs using an IR-S spectrophotometer with double NaC1 prisms. In order to avoid errors in transfer, mixing

5 'Amorphous' constituents in soil clays 245 and weighing of the clays, two-step dilution was adopted for preparation of the KBr discs. The final concentration for measurement was 2-3 mg of the original clay and the weight of treated clay obtained from the same weight of original clay in mg of KBr. Difference spectra were obtained by placing the KBr discs prepared from the clays before and after the treatment at 'sample' and 'reference' positions of the spectrophotometer, respectively. The ordinary spectra were obtained by using a pure KBr disc as reference. All the clay-kbr mixtures were allowed to stand at least for 24 hr in the spectrophotometer room in which relative humidity was controlled at 30-40%, to equilibrate their moisture retention. Any imbalance between the KBr discs due to causes other than selective dissolution gives rise to absorption effects in the difference spectra. Very careful manipulation in the preparation of the discs, as in the preparation and treatment of the clay samples, is therefore essential. In the present study, differences in weight loss between duplicate samples were in the range mg for mg of original material. The observed variation between duplicate discs in the major Si-O stretching, OH stretching and OH(H20) bending absorption bands did not exceed 5 % of the absorbance. Dithionite extraction EXPERIMENTAL RESULTS The infrared spectra of the Na2S20~-NaHCO~-Na citrate soluble fractions are shown in Fig. 2, together with those of the original clays. The weight loss and extractable iron oxide content are also shown expressed as percentages of the weight of the original clay. All the clay samples were appreciably attacked by the treatment, and the weight loss exceeded the amount of the iron oxide removed. The difference was particularly marked for the 1041 and Mt Schank clays derived from volcanic ash, where the amount of iron oxide removed accounts for only half of the observed weight loss. The presence of goethite may be inferred in the Glencoe clay, and possibly in the others, from its strong and broad OH stretching band with a maximum at about 3200 cm -1 (van der Marel, 1961; Farmer & Russell, 1967). The spectra also suggest dissolution of allophane-like materials from the 1041 and Mt Schank clays. These allophanic components are characterized by their relatively low Si/AI ratio, possibly about 1:2, suggested by the location of the absorption maxima of Si--O stretching band at lower wave numbers (Mitchell, Farmer & McHardy, 1964; Wada, 1966; Kanno, Onikura & Higashi, 1968). Further, the stronger absorption in the region from cm -1 in comparison with the respective 2% Na~COn and 0-5 N NaOH soluble fractions (Figs 3 and 4) may suggest that additional aluminarich components were dissolved by the dithionite-citrate treatment. On the other hand, the components dissolved from the Glencoe and Urrbrae clays, in which crystalline layer silicates predominate gave Si-O stretching and (A1)-OH bending bands characteristic of layer silicates (Fig. 2). Their OH stretching

6 246 K. Wada and D. J. Greenland '.,=" ;..," ""'"'"'"... -'... ~ ~ i I i 1 i I I I [ f I I "T'"I I I ;,-~ Mf Schank I L I : i.o. ",...~ ~ ".~ '~ f ""- 30/14"8 '" " I I I I I i L I I I i I '.... ~..' ~ -. ~ "~ ,.., W 106 I I I I 19 /12"5 '-. -" i I 1 I I I I I "'] I I -=x 2!..... : ~.- Glencoe I I I I Urrbrae I I I } /20'2 ". :" I I I I 1 I [ L ""1 I I I o,...,, _ -~/ "! f%;.: 14/11'3 "-.. :" 9.," I I I I I I I',1 [ I I I0 8 xlo0 cm -I FIo. 2. Difference spectra of NazS204--NaHCO3--Na citrate soluble fractions. Figures indicate the weight losses/the contents of extractable iron oxide expressed as weight of the original clays. In this and Figs 3, 4 and 5, dotted lines show the infrared spectra of the original clays, and x 2 notes that the spectra are magnified twice on the transmittance scale. bands, however, are rather broad and strong, and gave maxima at wave numbers lower than those of well-crystallized kaolins, and the insoluble materials in these clays (Fig. 5). The results may indicate close association of the layer silicates dissolved with the iron oxide and/or incorporation of iron atoms into their peripheral structure. The dithionite soluble fraction from W-106 exhibits significant yet very weak absorption in both the Si-O and OH stretching regions. The fiat and

7 'Amorphous' constituents in soil clays 247 I ~..-': '. 9..," ' %.. :'"': '" "', 9.,.,. I I I I I I I I I I "I" I ] I I i:..:" 9 %,' "-% :; ~ 9,..- Mr Schank 18.5 "....": I I I T I I I I l I I I I I I 1 :'%.. 9 ""." ;...- "V';'.'::', ".... "~: W " : :' I I I I I I I I I I I I I "1 I I I......,..,...,... :..... x2 ~ _~_...~--... " - : : :- -.: '... :: Glencoe 6.5 ~ ':.." I I I I I I, I I I I I I I """1 I I I ~==1i... x2... : ~..."(":" 9,; '....j I I I I I Urrbrae 2" 5 "~ I I l I I I I l :'"1 I I I I : I0 8 6 X I00 cro -I FIG. 3. Difference spectra of 2 ~ Na2CO3 soluble fractions. Figures indicate the weight losses expressed as ~ weight of the original clays. relatively high level of absorption in the region from cm -1 suggests dissolution of poorly organized gel material rich in alumina (Mitchell et al., 1964; Leonard et al., 1964; Kanno et al., 1968). The dithionite-citrate-bicarbonate method (Mehra & Jackson, 1960) and the dithionite-hydrochloric acid method (Mitchell & Mackenzie, 1954) for removal of 'iron oxides' were compared for the two clays, 1041 and Urrbrae, to determine their relative effects on silicate dissolution. The dithionite-acid method increased the weight loss to 37% for the 1041 clay, and decreased it to 11% for the Urrbrae clay with incomplete disappearance of its brown chroma. However, major features

8 248 K. Wada and D. J. Greenland,! ;.,":'" "":"' 'Y'"""... I I I I I I I I I I I "'['" 1 I I ;, _.-' -... Mf Schank I I I I ~,,...,;,.* ~.- 35'5 :"...." I I I I I 1 I I I I I I I i i W : q,.... X2 -' Glencoe 44.5 ~.- I I I I I I I "1 I I I I 15"5 ':... II 1 I I I I I :..::.'"" x2 I I I I I I I I "'"i I I I I... ~... ~176 "",~r'. t :.. %] Urrbroe I I 1 t I l : "[. %~ :" l 1 1 t I I '-I I t l I0 8 x I00 cm -I FIG. 4. Difference spectra of 0-5 N NaOH-soluble fractions. Figures indicate the weight losses expressed as ~ weight of the original clays. of the Si-O and (A1)-OH bending bands of the fractions dissolved were almost the same by the two methods. A notable difference between the two methods was found in the region cm -1 for the 1041 clays. Treatment with dithionite and acid produced a very marked decrease in absorbance in this region, whereas after treatment with dithionite and citrate only a small decrease was observed. Since the clay had not been pretreated with HzO2, the results may be interpreted in terms of desorption and/or dissolution of humic matter present in the original clay and adsorption of

9 'Amorphous' constituents in soil clays 249 I I I I l I K [ I ""l... I I I i, '...:..'....-" x 2--'"'"", MI" Schank 16 ""... 9 [ [ ' ' ' L [ [ ' ' ' [ [ [ [ ' ~,: 9 W t :,o 9 I I I I I 1 I I I 1 I '1 I [ 1 i: Glencoe : I l I I I t I I I I l I'"1" 1 I I Urrbroe 72 ] 1 1 I I l I I I I".q I l I0 8 xlo0 cm -I F1o. 5. Infrared spectra of residues remaining after 0'5 N NaOH treatment9 Figures indicate the contents of the residues expressed as percentage of the original clays, organic anions, if any, during the treatment. Both reactions would occur during treatment with dithionite and citrate while only desorption and/or dissolution ol humic matter would occur for the dithionite-acid treatment9 In this particular run, the washing with NaCHsCOO (ph 5"0) was replaced by one with 0.5 N NaC1, in view of possible adsorption of acetate ion. The absorption maximum at 1400 cm -1 increased after the Mehra & Jackson treatment and decreased after the Mitchell & Mackenzie treatment. This observation suggests an organic origin of the 1400 cm -1 band often noted on the infrared spectra of allophanes. The slight negative absorption

10 250 K. Wada and D. J. Greenland at 1660 cm -~ and at 1400 cm -~ on the difference spectrum for the Mt Schank clay (Fig. 2) can also be explained by the effect of adsorption of citrate. Dissolution in 2% Na~COs solution Digestion with 2% Na2CO~ solution has been used to remove 'amorphous' cementing material which consists of silica and alumina, and might inhibit the dispersion of clays (Jackson, 1956). The 1041 and Mt Schank clays showed appreciable dissolution in 2% NazCO3 solution (Fig. 3). The difference spectra suggest again dissolution of allophane-like materials, which absorb less in the region cm -1 compared with the previous dithionite soluble fractions, but give their absorption maxima at slightly lower frequencies. Two fairly strong absorption bands occur around 1630 cm -1 and 1400 cm-l which arise from concurrent dissolution of humic matter adsorbed originally, and of citrate adsorbed in the preceding treatment for removal of iron oxide. Although the Glencoe and Urrbrae clays were resistant to the treatment, the difference spectra suggest that the material dissolved was more crystalline than the 1041 and Mt Schank allophanes (Fig. 3). From the W-106 clay, a hydrous component with very poorly defined spectral features was dissolved in small amounts. Dissolution in 0"5 N NaOH solution All the clays were dissolved in significant amounts by 0.5 N NaOH solution (Fig. 4). The pronounced dissolution of those derived from volcanic ashes is again illustrated. For the 1041 clay, the spectra of the 2% Na2CO~ and 0"5 N NaOH soluble fractions are similar, suggesting homogeneity of the allophanic components present. On the other hand, there is a significant difference between the corresponding fractions from the Mt Schank clay. The higher absorption in the region above 1000 cm -1 indicates dissolution of allophanic components with relatively high Si/A1 ratio in 0-5 N NaOH solution. A previous differential thermal analysis gave 33% as weight of gibbsite in the W-106 clay (Wada & Aomine, 1966) and correspondingly gibbsite constitutes a large part of the 0.5 N NaOH soluble fraction (Fig. 4). However, since there is some overlap of the infrared absorption bands of gibbsite and crystalline layer silicates, some dissolution of the latter might also have occurred. There is no positive indication of the presence of allophane in the W-106 clay, as there is for the 1041 and Mt Schank clays, irrespective of the fact that all three clays are from volcanic ash soils. This may be ascribed to the differences in the rock phase and age of the volcanic ashes. The 1041 and Mt Schank soils are derived from andesitic and basaltic ashes, which may not be older than years, whereas the W-106 soil is derived from much older, dacitic ash. The spectra obtained from the Glencoe and Urrbrae clays are relatively unaffected, although dissolution occurred in significant amounts (Fig. 4). Patterns of poorly crystalline layer silicates predominate in both difference spectra, and 'amorphous' silicates removed by the NaOH treatment are very minor in amount. The spectra of the layer silicates dissolved show relative enhancement in

11 PLAT~ 1. Electron micrographs of 1041 clay (a) after dithionite-citrate-bicarbonate extraction, x 30000; (b) dithionite treated clay after 2 ~ Na2CO3 extraction, x 30000; and (c) dithionite and Na2CO3 treated clay after 0-5 N NaOH extraction, x (Facing p. 250)

12 'Amorphous' constituents in soil clays 251 absorption at 3500 cm -1 and 910 cm -1 in comparison with the respective 0.5 N NaOH insoluble fractions (Fig. 5). The observed difference may signify selective dissolution of kaolin minerals, or possibly their disordered, peripheral portions. A very slight increase in the ratios of the X-ray reflection intensity (7 A/10 A and 3.5 A/3-33 A)was noted for the Urrbrae but not for the Glencoe clay. The spectra of the residues from the 1041 and Mt Schank clays remaining after the 0.5 N NaOH treatment (Fig. 5) showed marked differences from those found for the NaOH soluble fractions (Fig. 4). The principal components are essentially nonhydrous silicates with relatively high Si/A1 ratios and low hydroxyl contents. X-ray examination showed the residues to be almost amorphous. Volcanic glasses of different composition are therefore the most likely major constituents. The electron micrographs (Plate 1) show the changes occurring after the successive treatments for the 1041 clay. The shape of the materials seen in the micrographs accords with the dominance of glass fragments in the final residue. Layer silicate patterns predominate in the spectra of the residues from the W-106, Glencoe and Urrbrae clays. X-ray analyses proved that all the crystalline minerals present in the original clays except gibbsite were still present after the 0'5 N NaOH treatment. DISCUSSION The differential dissolution--infrared spectroscopic analysis technique shows that for the soft clays examined the material dissolved by any particular treatment is primarily dependent on the initial mineral composition. Dissolution of amorphous aluminium silicates, allophanes, was only found for the clays in which they are the major constituents. Layer silicates, probably including disordered kaolin, were dissolved from the clays in which layer silicates predominated. Identification of iron oxide and hydroxide by difference spectra alone is difficult, but the presence of goethite was inferred for the one soil which contains about 20% of extractable iron oxide. Differences in the materials removed from the volcanic ash soils by the successive treatments are also made apparent by the infrared spectra shown in Figs 2-5. Dithionite-citrate and sodium carbonate treatments removed material with an absorption maximum at frequencies lower than 1000 cm -1 and relatively rich in aluminium. The residues left after the 0.5 N NaOH treatment (Fig. 5) absorbed at frequencies greater than 1000 cm -1 and were relatively rich in silica. A similar trend was not observed for the clays in which layer silicates predominated, but is seen in the chemical composition data for the Glencoe sample (Table 3). Had the carbonate treatment been omitted the material removed by NaOH would have shown a broader absorption in the 1000 cm -1 region resembling that of allophane rather more. However the spectra due to the NaOH soluble material (Fig. 4) show clearly that this treatment removes material giving quite sharp absorption bands which in those clays containing crystalline minerals may be attributed to less-ordered surface phases. The carbonate-soluble materials appear

13 252 K. Wada and D. J. Greenland ~r~ O" I"-. eq r oo 0 ~ e~ ("4 o 6 0.~ e~ 0 oo e~ r ~ t~. o 6 t~ o o o "d 0.~_ E 0 [..., M [.,.4 r,,q "4

14 'Amorphous' constituents in soil clays 253 to be less well ordered than those removed by the NaOH treatment, but the true ailophanes of samples 1041 and Mt Schank were not entirely dissolved by the treatments, Hence the carbonate-soluble material should not be described as allophane either. The 1630 cm -1 band due to the OH bending vibration of adsorbed water appears on some of the difference spectra. Provided that the condition of the reference and sample discs is similar and that they have attained equilibrium with the external humidity, the intensity of the absorption at 1630 cm -1 should be directly related to the extent of the hydrophilic surface of the material removed by the treatments. When the absorbance is large, finely divided materials of high specific surface area must have been removed. This would appear to be true of the dithionite-soluble layer silicates and 'free' iron oxides from the Glencoe and Urrbrae clays (Fig. 2), the 2% NazCO~-soluble alumina-rich gel-like material from the W-106 clay (Fig. 3) and less certainly, the 2% Na2CO~ and 0-5 N NaOH-soluble allophanes from the 1041 and Mt Schank clays (Figs 3 and 4). For these last two the intensity of the 1630 cm -1 band is confounded by overlying CO absorption. The absence of the 1630 cm -~ band indicates that the components dissolved were initially present as either surface coatings or a surface phase, the removal of which does not affect the surface area of the clays significantly. This partial dissolution of layer silicates is seen in the treatments of the Glencoe and Urrbrae clays with 2% NazCO3 and 0"5 r~ NaOH solutions. Finally, the negative absorption at 1630 cm -1 on the difference spectra might result from improved dispersion by the treatments, but this, particularly the weak one, likely results from slight error in the transfer, mixing and weighing of the clays in preparation of the KBr discs, unless the dissolution occurs to a considerable extent. The surface structure of particles of the Urrbrae clays has also been studied by electron microscopy using replica techniques (Greenland & Wilkinson, 1969). The replicas demonstrate clearly the irregular surface structures of these clays, the negligible change induced by dithionite treatment, and the smooth surfaces prevailing after removal of the disordered surface coatings by sodium carbonate treatment. These results are in excellent agreement with the interpretation of the differential infrared spectra. Further characterization of the materials removed by the so-called selective dissolution analyses would seem necessary. Differential thermal analyses could be used in a manner similar to that described here. It is desirable that supplementary information be obtained by X-ray diffraction, chemical analysis and electron microscopy. The influence of the disordered materials on the charge characteristics and adsorption properties of the days also deserves attention. ACKNOWLEDGEMENT This study was commenced whilst K. Wada held a Leverhulme Fellowship at the Waite Agricultural Research Institute. He wishes to thank the Leverhulme Trust and the University of Adelaide for the Fellowship granted.

15 254 K. Wada and D. J. Greenland REFERENCES FARMER V.C. ~ RUSSELL J.D. (1967) Clays Clay Miner. 15, 121. FOLLETT E.A.C., McHARDY W.J., MITCHELL B.D. & SMITH E.F.L. (1965a) Clay Miner. 6, 23. FOLLETT E.A.C., McHARDY, W.J., MITCHELL B.D. & SMITH E.F.L. (1965b) Clay Miner. 6, 35. GREENLAND D.J. & WILKINSON G.K. (1969) Trans. 3rd Int. Clays Conf. 1, 861. HASHIMOTO I. ~ JACKSON M.L (1960) Clays Clay Miner. 7, 102. JACKSON M.L. (1956) Soil Chemical Analysis--Advanced Coarse, Published by the author, Madison, Wisconsin. KANNO I., ONIKURA Y. ~ HIGASm T. (1968) Trans. 9th Int. Congr. Soil Sci. 3, 111. LEONARD A., Suzurd S., FRIPIAT J.J. & DE KIMPE C. (1964) J. phys. Chem. 68, MEHRA O.P. & JACKSON M.L. (1960) Clays Clay Miner. 7, 317. MITCHELL B.D. & MACKENZIE R.C. (1954) Soil Sci. 77, 173. MITCHELL B.D. & FARMER V.C. (1962) Clay Miner. Bull. 5, 128. MITCHELL B.D., FARMER V.C. & MCHAgOY W.J. (1964) Adv. Agron. 16, 317. VAN DER MAREL H.W. (1961) Acta Univ. Carolina--GeoL SuppL 1, 23. W,~d)A K. (1966) SoilPl. Fd, Tokyo, 12, 176. WADA K. & AOMINE S. (1966) SoilPl. Fd, Tokyo, 12, 151. WEAVER R.M., SYERS J.K. & JACKSON M.L. (1968) Proc. Soil ScL Am. 32, 497. YOSI-RNAGA, N. (1966) Soil PI. Fd, Tokyo, 12, 47.