Clay Minerals (1993) 28, 461-474 THE RECOGNITION OF AMORPHOUS SILICA IN INDURATED SOIL PROFILES BALBIR SINGH* AND R. J. GILKES Soil Science and Plant Nutrition, School of Agriculture, The University of Western Australia, Nedlands, WA, 6009, Australia (Received 31 March 1992; revised 6 November 1992) ABSTRACT: Silica-indurated subsurface horizons of an in situ lateritic profile in semi-arid western Australia were investigated using a range of electron-optical and X-ray diffraction (XRD) techniques. These indurated materials were compared with underlying non-indurated pallid zone material. The secondary silica content of the indurated horizons, as determined by electron microprobe analysis, varied from 8 to 33%. Quantitative digital images for secondary silica, generated by mathematical manipulation of digital Si and AI-K0:images, showed that kaolinite pseudomorphs after micacontained the lowestamountsof secondarysilica, with the highest amounts being present in the inter-pseudomorph clay matrix. Variations in the amount of silica in the matrix are consideredto reflect variationsin the initial porosityof the clay matrix. Suchvariationsmay arise from differences in the Al/Si ratio of parent minerals. Transmission electron microscopy(TEM) showed that amorphous silica adhered to the (001) face of kaolinite crystals. The secondary silica could not be detected by either standard or differential XRD procedures. Amorphous or poorly ordered forms of silica are common constituents of soils. Opal of biological origin occurs as discrete grains recognizable by their distinct morphology, and rarely contains extraneous materials. It can, therefore, be easily identified by its characteristic morphology, chemical composition, specific gravity and optical properties (Wilding & Drees, 1971, 1974; Hurd & Theyer, 1977; Simpson & Volcani, 1981). The recognition and characterization of opal-A and opal-CT of pedogenic origin is, however, not as simple for a number of reasons (Drees et al., 1989). The foremost of these is that pedogenic opal has a nondescript morphology and often occurs as a thoroughly diffused cementing agent in a clay-rich matrix so that it can not be isolated and analysed by the techniques commonly used for biogenic opal. Selective dissolution and analysis of opal present in an indurated clay matrix is unreliable as dissolution is greatly affected by particle size and the extractant used. Larger particles of opal may not fully dissolve and contributions from the dissolution of clay minerals and fine quartz are unavoidable (Chadwick et al., 1987). Inorganic silica in soils mostly co-exists with clay minerals at a scale smaller than the resolution of the electron microprobe (5/~m). The XRD band of opal is very broad and is overlapped by hk reflections of clay minerals in such mixtures making opal undetectable by normal XRD procedures, even when present in moderate proportions. For these reasons, optical microscopy of thin-sections using reflected and * Present address: Department of Geology and Geophysics, University of California, Berkeley, CA 94720, USA. 1993 The Mineralogical Society
Transcript
Clay Minerals (1993) 28, 461-474
T H E R E C O G N I T I O N OF A M O R P H O U S S I L I C A IN I N D U R A T E D SOIL P R O F I L E S
B A L B I R S I N G H * AND R. J. G I L K E S
Soil Science and Plant Nutrition, School of Agriculture, The University of Western Australia, Nedlands, WA, 6009, Australia
(Received 31 March 1992; revised 6 November 1992)
ABSTRACT: Silica-indurated subsurface horizons of an in situ lateritic profile in semi-arid western Australia were investigated using a range of electron-optical and X-ray diffraction (XRD) techniques. These indurated materials were compared with underlying non-indurated pallid zone material. The secondary silica content of the indurated horizons, as determined by electron microprobe analysis, varied from 8 to 33%. Quantitative digital images for secondary silica, generated by mathematical manipulation of digital Si and AI-K0: images, showed that kaolinite pseudomorphs after mica contained the lowest amounts of secondary silica, with the highest amounts being present in the inter-pseudomorph clay matrix. Variations in the amount of silica in the matrix are considered to reflect variations in the initial porosity of the clay matrix. Such variations may arise from differences in the Al/Si ratio of parent minerals. Transmission electron microscopy (TEM) showed that amorphous silica adhered to the (001) face of kaolinite crystals. The secondary silica could not be detected by either standard or differential XRD procedures.
Amorphous or poorly ordered forms of silica are common constituents of soils. Opal of biological origin occurs as discrete grains recognizable by their distinct morphology, and rarely contains extraneous materials. It can, therefore, be easily identified by its characteristic morphology, chemical composition, specific gravity and optical properties (Wilding & Drees, 1971, 1974; Hurd & Theyer, 1977; Simpson & Volcani, 1981). The recognition and characterization of opal-A and opal-CT of pedogenic origin is, however, not as simple for a number of reasons (Drees et al., 1989). The foremost of these is that pedogenic opal has a nondescript morphology and often occurs as a thoroughly diffused cementing agent in a clay-rich matrix so that it can not be isolated and analysed by the techniques commonly used for biogenic opal. Selective dissolution and analysis of opal present in an indurated clay matrix is unreliable as dissolution is greatly affected by particle size and the extractant used. Larger particles of opal may not fully dissolve and contributions from the dissolution of clay minerals and fine quartz are unavoidable (Chadwick et al., 1987). Inorganic silica in soils mostly co-exists with clay minerals at a scale smaller than the resolution of the electron microprobe (5/~m). The X R D band of opal is very broad and is overlapped by hk reflections of clay minerals in such mixtures making opal undetectable by normal X R D procedures, even when present in moderate proportions. For these reasons, optical microscopy of thin-sections using reflected and
* Present address: Department of Geology and Geophysics, University of California, Berkeley, CA 94720, USA.
�9 1993 The Mineralogical Society
462 Balbir Singh and R. J. Gilkes
transmitted light has been the major technique for the recognition of discrete secondary silica bodies such as silans and glaebules in soil material (Flach et al., 1969; Brewer et al., 1972, 1976).
Optical microscopy cannot detect silica which is homogeneously diffused in a clay matrix at a submicron scale, and many fine-grained (<1-0/ tm) materials exhibit similar isotropic optical properties (Brewer, 1976; Butt, 1983, 1985). This problem can be overcome by the use of high-resolution electron-optical techniques which enable the direct observation and analysis of materials with a spatial resolution of nanometers.
This paper describes the use of XRD and microanalytical techniques to investigate the induration of soil matrix by pedogenic silica. Horizons indurated by secondary silica are common in hard pans developed in subsurface horizons of ancient lateritic profiles in semi- arid regions of Australia. Such hard pans have sometimes developed by silicification of the laterite pallid zone, which is a clay-rich zone produced by intense in situ chemical weathering (McCrea et al., 1990). Properties such as hardness, fabric and colour exhibit a diffuse boundary between siliceous hard pan and unaffected pallid zone suggesting that the hard pan was produced by induration of the pallid zone by illuviation of silica from other soil horizons. To test this hypothesis samples of pallid zone were compared with associated indurated hard pan.
M A T E R I A L S A N D M E T H O D S
Five profiles exposed in a railway cutting at Merredin, 265 km east of Perth, Western Australia were sampled at three depths i.e. hard pan, diffuse boundary between hard pan and pallid zone, and unaffected pallid zone. Blocks of soil, about 30 x 30 • 30 cm 3 in size, were excavated from the face of the profile with minimum disturbance. Freshly exposed fracture and ped surfaces of the samples were investigated by scanning electron microscopy (SEM) using a Philips 505 instrument equipped with a back scattered electron (BSE) detector and energy dispersive X-ray analyzer (EDX). Specimens of --1 cm 3 volume were mounted on A1 stubs and coated with carbon and subsequently with a 50 A-thick layer of Au in a vacuum evaporator before SEM analysis.
A portion of the sample was impregnated with epoxy resin and polished thin-sections mounted on glass slides were made for optical microscopy, SEM using BSE image, and electron probe microanalysis (EPMA). Digitized X-ray images of the distribution of elements in the fabric were obtained and processed using EDAX PV9900, an X-ray analytical software package available with the EDX system. A dilute suspension of sample was prepared in distilled water by shaking a crushed bulk sample. The < 10 ktm fraction was separated from the bulk suspension and drops were placed on holey carbon films with a micropipette to prepare samples for TEM. The drops were dried under cover at room temperature. Samples prepared for TEM were investigated using a Philips EM 430 instrument operated at 300 kV.
X-ray diffraction patterns of the whole material and the clay (<2 /~m) fraction were obtained using Cu-Ko~radiation with a computer-controlled Philips vertical goniometer and a graphite curved crystal monochromator. The digital XRD patterns were interpreted with the aid of XPAS analytical software (Balbir Singh & Gilkes, 1992) which enabled rapid calculation of crystallographic parameters and the generation of differential XRD (DXRD) patterns.
Recognition of pedogenic silica 463
R E S U L T S AND D I S C U S S I O N
Macro properties
The soil profiles exhibited a sequence of zones typical of deeply weathered laterite profiles in south-western Australia (Gilkes et al., 1973). Abundant quartz veins were preserved within the mottled and pallid zone (Fig. 1) demonstrating that the profiles developed by in situ isovolumetric weathering (Millot, 1970). The pallid zone materials exhibit a highly microporous structure, white colour and preservation of rock fabric (McCrea et al., 1990). The upper part of the hard pan in these profiles occurred in the lower part of the mottled zone and hard pans extended down into the pallid zone with diffuse boundaries with the unaffected pallid zone. In the field, hard pan can be easily distinguished by its high mechanical strength and lack of porosity. It can only be broken with difficulty using a geological hammer under both wet and dry conditions.
Mineralogy
The major crystalline phases in the pallid zone material, as determined by XRD of the bulk material (Fig. 2A), are kaolinite/halloysite and quartz. This mineralogy is consistent with past work on the lateritic weathering of primary minerals in the region, which has demonstrated that feldspar, biotite and muscovite alter directly to kaolinite and/or halloysite without any intermediate smectite or amorphous phases (Gilkes & Suddhiprak- arn, 1979; Anand et al., 1985; Singh & Gilkes, 1991). The major crystalline components of the hard pan material (Fig. 2B) are also kaolinite/halloysite and quartz. Amorphous silica
E
el.
Pallid zone I Hard pan
[ ] Mottled zone d-~Ouartz vein
F[6.1. A sketch showing the position of hard pan in a lateritic profile.
464 Balbir Singh and R. J. Gilkes
K
0
B
Q
0
C~
4.8 18.8 16.8 22.8
K K K O
2 8 . 8 34.1] 4 8 . 0 4 6 . 0 5 2 . 8 5 8 . 0 64.el 20 A N G L E
FIG. 2. Random powder XRD patterns of pallid zone (A) and hard pan (B) materials. Kaolinite/ halloysite (7/k) (K) and quartz (Q) are the major minerals in both materials.
was present in large voids (2-4 mm) between the peds. It may also have diffused into the clay matrix, although its presence is not indicated by the XRD patterns. Powder XRD patterns of pure amorphous silica extracted from the voids and an artificial mixture of pallid zone material and 20% amorphous silica are shown in Fig. 3A, B. Amorphous silica gives a broad band centred at 4.1/~ (Jones & Segnit, 1971) characteristic of opal-A. The diffraction band for amorphous silica is not present in the XRD pattern for the mixture of 20% opal-A and pallid zone material (Fig. 3B). Thus, soil materials may contain significant amounts of amorphous silica which would remain undetected by XRD.
Differential XRD has been shown to improve the detection of poorly ordered materials in mixtures with crystalline materials (Schulze, 1981; Schwertmann et al., 1982). A DXRD pattern obtained by subtracting a digital XRD pattern for pallid zone material (Fig. 3C) from the pattern for a mixture containing 20% amorphous silica (Fig. 3B) is shown in Fig. 3D. The broad band for the amorphous silica which might be expected to be present in the DXRD pattern is not evident. This is mainly due to inexact matching of the 11, 02 reflections of kaolinite in the two patterns and to the diffraction band for amorphous silica being relatively very weak. Thus, in this instance, DXRD does not provide a means of recognizing small to moderate amounts of amorphous silica in clay-rich soils.
F[6.3. XRD patterns of amorphous silica from voids in the hard pan (A), an artificial mixture of kaolinite, quartz and 20% amorphous silica (B), the artificial mixture containing no amorphous silica (C), and a DXRD pattern (D) obtained by subtracting pattern C from pattern B. The broad
diffraction band of amorphous silica is not resolved in the original (B) or DXRD (D) patterns.
Micro fabric
A typical region of the pallid zone material exhibits the excellent preservation of rock fabric with vermicular kaolinite pseudomorphs after primary grains of mica, an isotropic groundmass of kaolinite after feldspar, coarse quartz and abundant large voids (Fig. 4A). Thus, the pallid zone has a much lower bulk density and higher porosity than the parent rock. Kaolin pseudomorphs after mica and feldspar, and the coarse size of quartz grains indicate that the parent material was a micaceous granite which is of common occurrence in the region (Williams, 1975). The thin-section of the associated indurated hard pan material (Fig. 4B) exhibits a much more compact fabric. The pseudomorphs of primary minerals are still evident as in the pallid zone material but some isotropic and translucent material is also present in voids between the mica pseudomorphs and kaolinite groundmass.
The typical SEM images of the pallid zone (Fig. 5A, B) show features consistent with those observed in optical thin-sections. The fabric of the parent material is preserved with high fidelity in the pallid zone material. Pseudomorphs after primary minerals, including stacks of plates representing kaolinized exfoliated mica, fine-grained groundmass and large voids are present. The fabric of hard pan (Fig. 5C, D) is denser than that of the pallid zone material. The large pores and open exfoliated pseudomorphs after mica, which are abundant in the pallid zone, are rare in the hard pan. Ped surfaces exhibit a honeycomb-like morphology (Fig. 5E) which is characteristic of deposited clay-size material. The BSE image of a polished section of hard pan shows a much denser fabric than for pallid zone. The
466 Balbir Singh and R. J. Gilkes
Fl6. 4. (A) Optical micrograph of pallid zone material. Vermicular kaolinite pseudomorphs after mica (1) and feldspar grains (2), quartz grains (3) and voids (4) are present. (B) Optical micrograph of hard pan material showing pseudomorphs (1) similar to those in the pallid zone. There is
translucent material (5) present filling many of the voids between and within pseudomorphs.
regions of relatively bright contrast ( indicated by a white arrow in Fig. 5F) are mainly due to the high density of these regions although they also had a relat ively lower A1/Si ratio. These materials probably had the same porous fabric as in the pall id zone but have been modified
Recognition of pedogenic silica 467
F16. 5. (A and B) Scanning electron micrographs of a polished section of pallid zone material. Vermicular pseudomorphs (1) after mica, large voids (2) and pores (3) are present. Scanning electron micrographs of a fracture surface (C, D), ped surface (E) and polished section (F) of hard pan material. The fabric of hard pan is denser than that of pallid zone with few voids. The honeycomb morphology of ped surfaces (E) indicates deposition of clay illuviated from upper soil horizons (F).
subsequen t ly by depos i t ion wi thin pores of mater ia l r emoved f rom o ther soil hor izons ei ther as part icles or in solut ion.
Chemical composition Var ia t ion in the A1/Si a tomic rat io, de t e rmined by E P M A , for a n u m b e r of kaol in ized
regions for pall id zone and hard pan mater ia ls are shown in Fig. 6. The A1/Si rat io for pallid
468 Balbir Singh and R. J. Gilkes
zone material is typical of that for pure kaolinite (i.e. --1) and shows little variation. The equivalent areas for hard pan have a variable A1/Si ratio which is much lower than that for the pure kaolinite groundmass of pallid zone. The regions of bright contrast in the BSE image (Fig. 5F) invariably had low AI/Si ratios indicative of induration with Si. Thus the groundmass of the hard pan consists of a mixture of kaolinite and amorphous silica. However, no discrete regions of pure amorphous silica (i.e. AI/Si = 0) were detected within the matrix material. The amorphous silica may have been derived by illuviation of particles or may have precipitated from groundwater as water removed by evaporation from soil solution (Langford-Smith, 1978).
In order to define more precisely the spatial distribution of silica in the matrix, a novel technique was employed. Digital X-ray images for AI-Ko~ and Si-Ko: for a matrix of 6 • 6 #m 2 areas were obtained from 770 • 570 #m 2 regions of polished sections and stored in the computer memory. The spatial distribution of silica not associated with kaolinite (i.e. amorphous silica) was obtained by processing the digital X-ray images. The component of Si-Ko: arising from amorphous silica for each pixel was obtained by subtracting the Si-Ko: counts associated with kaolinite from the total Si-Ko: counts. The intensity of Si-Ko~ contributed by kaolinite was calculated from the AI-Ktr counts, assuming that all AI is associated with kaolinite and that the kaolinite had the average A1/Si ratio determined for pallid zone kaolinite. The optical, BSE and X-ray images for hard pan material are shown in Fig. 7. In the optical image, pseudomorphs (indicated as 1 in Fig. 7A) of mica grains which consist of oriented plates of kaolinite are easily distinguished from the fine-grained kaolinite groundmass (2). The pseudomorphs are, however, not distinct in the BSE image (Fig. 7B), possibly due to the partial infilling of pores in the pseudomorphs with silica. Quantitative elemental analysis, obtained by electron microprobe, of points marked 1 and 2 in the BSE image are given in Table 1 so that the image for amorphous silica could be interpreted semi- quantitatively. These analyses indicate that the mica pseudomorph (point 2) is only moderately enriched in Si (i.e. AI/Si is --0.87) relative to the fine-grained groundmass (i.e.
30
24 [] Pallid z o n e ?,',
-~ �9 Hardpan ,~,~
. i. m I I. _n iiii 0
0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1 AI/Si A t o m i c rat io
]FIG. 6. Histogram showing the distribution of values of the AI/Si atomic ratio for micron-size regions of the groundmass of pallid zone and hard pan materials.
Recognition of pedogenic silica 469
FIG. 7. Optical (A), back scattered electron (B) and X-ray images (C, D) of a polished thin-section of hard pan material. Chemical composition and amount of secondary silica, assuming all Al is in kaolinite, for points 1 and 2 in B are given in Table 1. The spatial distribution of amorphous silica is illustrated by the calculated Si-A1 (E) and Al/Si (F) images which emphasize the regions that contain Si in excess of that present in kaolinite. These regions are brighter and darker in E and F,
respectively.
A1/Si = 0-52 for point 1). The fine-grained regions in the BSE image, corresponding to the regions of bright contrast in the Si image (Fig. 7C), have a much lower A1/Si ratio than for mica pseudomorphs . Variat ion in the Si-Ko~intensity is in part due to meso-scale variat ions in the density of the kaolini te groundmass as also indicated by the A1 image (Fig. 7D). The distr ibution of amorphous silica is therefore best i l lustrated by Si-A1 and AI/Si images which
470 Balbir Singh and R. J. Gilkes
TABLE 1. Electron microprobe analyses of points 1 and 2 marked in Fig. 7B. The amounts of amorphous silica have been calculated from the analyses assuming that all AI is associated with
kaolinite and that no other phases except kaolinite and amorphous silica are present.
Point 1 Point 2 Fine grained Vermicular groundmass kaolinite Ideal kaolinite
are less sensitive to variations in density (Figs. 7E,F) . These images demons t ra te that amorphous silica is most abundant in the fine-grained groundmass be tween the pseudomorphs . The corresponding digital analysis of spectra for pall id zone mater ia l (Fig. 8) does not indicate the present of excess silica in ei ther mica pseudomorphs or the fine-grained groundmass with all f ine-grained materials having the composi t ion of kaolini te.
No discrete regions of pure amorphous silica were detected in these X-ray images of groundmass. The spatial resolution of E P M A is 5 ~m (the diffused beam diameter) and X-rays are thus derived from a volume of 25/~m 3. Kaol ini te crystals in the groundmass of
FIG. 8. Back scattered electron (A) and X-ray images (B, C) of a polished thin-section of pallid zone material. The digitized computer generated Si-A1 image of the groundmass is of uniform contrast as
would be expected for a non-silicified, solely kaolinite composition.
Recognition of pedogenic silica 471
these lateritic materials are mostly 0.2-2 pan in diameter and <0.1 ~m thick (the c axis) (Singh Balwant & Gilkes, 1992). Kaolinite crystals in pseudomorphs after mica can be larger, reaching sizes greater than ~10/~m (Singh Balbir & Gilkes, 1991). The mica pseudomorph shown in Fig. 7A contained 8% amorphous silica (Table 1) compared with 33% amorphous silica in the fine-grained groundmass which has formed from feldspar. The difference may be a consequence of different initial porosities and pore geometries of the two pseudomorph types which have affected the extent of deposition of amorphous silica.
Transmission electron microscopy
As is evident from the above discussion, EPMA is unable to resolve the amorphous silica present between clay crystals in the kaolinite groundmass and pseudomorphs after mica. The association of silica and kaolinite crystals required further investigation using TEM and electron diffraction to identify the materials and their spatial association at a submicron scale. A TEM micrograph and corresponding electron diffraction pattern of amorphous silica from the void-filling are shown in Fig. 9. The amorphous silica grain has curved boundaries, and voids of irregular size and shape are present within the grain. The electron diffraction pattern has a broad band at -4 .1 A which is consistent with the observed XRD pattern of amorphous silica (i.e. opal-A Fig. 3). However, no such grains of pure silica were found by TEM analysis of the hard pan groundmass. Kaolinite crystals had indistinct outlines due to cementation by amorphous silica. The TEM samples were prepared from very dilute suspensions and only isolated particles in a clean background were examined.
Fro. 9. Transmission electron micrograph and electron diffraction pattern of a fragment of amorphous silica (opal-A) from the hard pan horizon showing the characteristic conchoidal
morphology, voids of irregular shape (1) and diffuse diffraction pattern.
472 Balbir Singh and R. J. Gilkes
Amorphous silica adhering to the (001) face of a large kaolini te crystal and corresponding selected area electron diffraction ( S A E D ) pat tern are shown in Fig. 10. The S A E D pat tern consists of a hexagonal net of hk reflections due to kaolinite super imposed on the broad band at 4.1 /k due to amorphous silica. Submicron-size amorphous silica can thus be identified by its diffraction pat tern, chemical analysis and distinctive anhedral morphology. The A1/Si ratio for 30 nm-size regions of these particles consisting of mixtures of amorphous silica and kaolinite was de te rmined by TEM. The A1/Si rat io ranged from 0.6 to 0.7; whereas for regions of the particles that appeared to be free of silica the ratio ranged from 0.90 to 0.97, as was the case for kaolini te crystals in the pallid zone material . Thus, even at this submicron scale, there is an int imate association between kaolinite and amorphous silica in the hard pan.
G E N E R A L D I S C U S S I O N
Hard pans developed in clay-rich lateri te, pallid and mot t led zone materials in the Merredin soil profiles have been hardened by impregnat ion with amorphous silica. In all the profiles, high amounts of silica (30-40%) were present in the fine-grained groundmass and voids
F16. 10. (A) Transmission electron micrograph of amorphous silica adhering to the (001) face of a large kaolinite crystal from the hard pan horizon. (B) A magnified view of a part of A (indicated by white arrow) showing amorphous silica as a dense material with curved edges (1) on the surface of kaolinite (2) which exhibits straight edges and moire fringes characteristic (3) of crystalline materials. (C) An electron diffraction pattern of the region shown in B, the hexagonal net of sharp reflections
due to kaolinite is superimposed on a diffuse ring due to amorphous silica.
Recogni t ion o f pedogenic silica 473
between mica pseudomorphs, with comparatively less silica (7%) being present in the mica pseudomorphs. The variation in the Si content in the groundmass reflects the initial porosity of the matrix that had been produced by chemical weathering of the parent rock. Assuming that all AI in the parent mineral is conserved during isovolumetric lateritic weathering, kaolinite pseudomorphs after plagioclase, orthoclase, muscovite and biotite are expected to have about 5, 54, 15 and 53% porosity, respectively, so that the resulting AI/Si ratio of the material would vary considerably.
Silica could be deposited from soil solution within these voids during drying of the soil. Silica polymers may be adsorbed on the hydroxyl planes exposed on the (001) face of kaolinite crystals in these pseudomorphs to provide a template for further deposition of silica (Chadwick et al., 1987). With each wetting and drying cycle the layer of silica adsorbed on the kaolinite surface grows and eventually "welds" the adjacent kaolinite crystals at the point of contact. Therefore, even a small content of silica ( < 5 % ) forming a thin layer on clay mineral surfaces can bind the whole matrix without filling the pore-space to a significant extent. In sandy soils, much silica may be deposited as isolated flocs which simply occupy the pore-space. Presumably some contact points between quartz grains will also become welded but such contact points are likely to be much less abundant in sandy soils relative to clay-rich soils so that the degree of cementation would be less. A sandy soil would therefore not be hardened until most of the pore-space had been filled. Larger amounts of secondary silica might be required to harden a sandy soil in comparison to a clayey soil.
This study has shown that the nature and distribution of amorphous silica in an indurated material cannot be determined by any single technique. The X R D band of amorphous silica is very weak and is overlapped by hk reflections of clay minerals, preventing the detection of amorphous silica even when it is present in signficant amounts. A combination of optical, XRD and electron optical techniques is essential for determining the quantity, composition and distribution of amorphous silica in the soil matrix.
BREWER R., BETTENAY E. • CHURCHWARD H.M. (1972) Some aspects of the origin and development of the red and brown hard pan soils of Bulloo Downs, Western Australia. Division of Soils Technical Paper No. 13. CSIRO, Australia.
BREWER R. (1976) Fabric and Mineral Analysis of Soils, p. 482. R. E. Krieger Publishing Company, Huntington, New York.
BuTt C.R.M. (1983) Aluminosilicate cementation of saprolite, grits and silcretes in Western Australia. J. Geol. Soc. Aust. 30, 179-186.
BoTr C.R.M. (1985) Granite weathering and silcrete formation on the Yilgarn Block, Western Australia. Aust. J. Earth Sci. 32,415-432.
CHADWICK O.A., HENDRICKS D.M. & NETI~LETON W.D. (1987) Silica in Duric soils: I. A depositional model. Soil Sci. Soc. Am. J. 51, 975-982.
DREES L.R., WILDING L.P., SMECK N.E. & SENKAYI A.L. (1989) Silica in soils: Quartz and disordered silica polymorphs. Pp. 913-974 in: Minerals in Soils Environments (J.B. Dixon & S.B. Weed, editors). Soil Sci. Soc. Am., Madison, Wisconsin.
FLACH K.W., NETFLETON W.D., GlEE L.H. & CADY J.C. (1969) Pedocementation: induration by silica, carbonates, and sesquioxides in the Quaternary. Soil Sci. 107, 442-453.
GILKES R.J., SCHOLZ A. & D1MMOCK G.M. (1973) Lateritic deep weathering of granite. J. Soil Sci. 24, 523-536. G1LKES R.J. & SUDDHIPRAKARN A. (1979) Biotite al'Leration in deeply weathered granite. II. The oriented growth of
HURD D.C. & THEYER F. (1977) Changes in the physical and chemical properties of biogenic opal from the central equatorial Pacific: Part II. Refractive index, density and water content of acid-cleaned samples. Am. J. Sci. 277, 1168-1202.
JONES J.B. & SEGNIT E.R. (1971) The nature of opal. I. Nomenclature and constituent phases. J. Geol. Soc. Aust. 18, 57-68.
LAN~FORD-SMITH T. (1978) Silcrete in Australia. Mongr. Ser., Dept. Geography, Univ. New England, Armidale. MCCREA A.F., ANAND R.R. & GILg~S R.J. (1990) Mineralogical and physical properties of lateritic pallid zone
materials developed from granite and dolerite. Geoderrna 47, 33-57. MILLOT G. (1970) Geology of Clays. Springer-Verlag, Berlin. SCnULZE D.G. (1981) Identification of soil iron oxides minerals by differential X-ray diffraction. Soil Sci. Soc. Am.
J. 45, 437~140. SCHWERTMANN U,, SCHULZE D.G. & MURAD E. (1982) Identification of ferrihydrite in soils by dissolution kinetics,
differential X-ray diffraction, and Mrssbauer spectroscopy. Soil Sci. Soc. Am. J. 46, 869-875. SIMPSON T.L. & VOLCAN1 B.E. (1981) Silicon and Siliceous Structures in Biological Systems. Springer-Vedag, New
York, Inc. Secaucus, NJ. SINGH BALalR & GILKES R.J. (1991) Weathering of chromian muscovite to kaolinite. Clays Clay Miner. 39, 571-579. SINGrt BALBIR & GILKES R.J. (1992) XPAS: An interactive computer program for analysis of powder X-ray
diffraction patterns. Powder Diffraction 7, 6--10. S1NGIt BALWANT & GILKES R.J. (1992) Properties of soil kaolinites from south-western Australia. J. Soil Sci. 43,
645-667. WILDING L.P. & DREES L.R. (1971) Biogenic opal in Ohio soils. Soil Sci. Soc. Am. Proc. 35, 1004-1010. WILOIN6 L.P. & DREES UR. (1974) Contributions of forest opal and associated crystalline phases to fine silt and clay
fractions of soils. Clay Clay Miner. 22, 295-306. WILLIAMS I.R. (1975) South Western province. Pp. 65-69 in: Geology of Western Australia. West. Australian Geol.
