Clay Minerals (1988) 23, 81-90
THE E F F E C T OF A N F E ( I I ) - S I L I C A T E ON S E L E C T E D
P R O P E R T I E S OF A M O N T M O R I L L O N I T I C C L A Y
D. W . O S C A R S O N
AND R. B. H E I M A N N *
Atomic Energy of Canada Limited, Whiteshell Nuclear Research Establishment, Pinawa, Manitoba ROE 1LO
Canada
(Received 3 July 1987; revised 7 December 1987)
A B S T R A C T : A montmorillonite clay and an X-ray amorphous Fe(II)-silicatewere suspended
in a synthetic groundwater solution under oxic and anoxic conditions at 23 and 70 ~ for 200
days. The clay samples were then analysed for total Fe, Fe(II), dithionite- and oxalateextractable Fe (Fed and Feo), cation exchange capacity (CEC), and examined by X-ray
diffraction (XRD) and diffuse reflectance spectrometry (DRS). The total iron content and
Fe(II)/Fe(III)ratio of the clay increased with an increase in temperature and both were greater
under anoxic than oxic conditions, as were the amounts of Fed and Feo. The CEC of the clay was
lower in the presence of the Fe-silicate, suggesting that Fe-hydroxide had precipitated at the
edges of the clay blocking some cation-exchange sites. No changes in the clay were detected by
XRD or DRS after contact with the Fe-silicate, nor was there any evidence for Fe-hydroxide
material in the interlayer region of the clay. However, it is shown that the amount and nature of
iron associated with clay can be significantly altered and certain properties of the clay, such as
CEC, can be affected when the clay is in contact with Fe(II)-containing materials.
The C a n a d i a n concept for the disposal of nuclear fuel waste involves its emplacement in
corrosion-resistant containers and subsequent disposal in a vault excavated 500 to 1000 m
deep in plutonic rock in the C a n a d i a n Shield (Hancox, 1986). A montmorillonitic clay may be
used as a component of a barrier material to surround the containers in a disposal vault.
Montmorillonite has many desirable properties including a high sorption capacity for many
radionuclides, a high swelling potential, and a low hydraulic conductivity, thus making it a
potentially effective barrier component.
There are likely to be significant quantities of iron in a disposal vault. The sources of iron
may include some structural components of the waste containers composed of carbon steel:
supporting structures e.g. steel bolts, anchors, and flanges, required to stabilize the
surrounding rock mass; Fe(II)-containing material mixed with the barrier material to ensure
that reducing conditions are maintained; and Fe-minerals associated with granite.
Iron(II) is relatively soluble, and if the redox potential in a disposal vault is uniform and
sufficiently reducing, Fe(II) species can diffuse throughout the barrier material. This should
have no detrimental effect on the desired properties of the clay component of the barrier. If,
however, there are zones in the barrier where the redox potential is high enough to oxidize
Fe(II) to Fe(III), the Fe(III) formed will precipitate as an Fe(III)-hydroxide 1. This material
could block sorption sites on the surface of the clay (Clark & Nichol, 1968; Carstea, 1968),
* Present address: Alberta Research Council, Edmonton, Alberta T6H 5X2 Canada.
1 In this paper the term Fe-hydroxide is used to denote all Fe-hydroxides, oxides and oxyhydroxides.
9 1988 The Mineralogical Society
82
D. IV. Oscarson and R. B. Heimann
hence decreasing its potential to sorb some radionuclides that will eventually be released from
the nuclear fuel waste after the waste containers are breached. Iron hydroxide could also act
as a cementing agent to bond clay particles together, thus potentially adversely affecting the
swelling capacity and permeability of the clay (Alperovitch et al., 1985; EI-Swaify &
Emerson, 1975; Blackmore, 1973; E1-Rayah & Rowell, 1973).
It is apparent that iron materials can significantly affect the properties of soils and clays
under some conditions. The purpose of this study is to examine the effect of an amorphous
Fe(II)-silicate on selected properties of a montmorillonitic clay to understand better the
behaviour of this clay when in contact with Fe-containing materials in a nuclear fuel waste
disposal vault.
MATERIALS AND METHODS
Solids
The clay used was a bentonite mined by Avonlea Mineral Industries Limited, Regina,
Saskatchewan, Canada, from the Bearpaw Formation of Upper Cretaceous age, in southern
Saskatchewan. The mineralogical composition of this clay was determined by Quigley (1984)
and is approximately 80 wt~o montmorillonite, 10~o illite, and minor amounts of quartz,
feldspar, gypsum, carbonate, and organic matter. The clay has a CEC of 82 mEq/100 g, a
specific surface area of 630 m2/g, and Na § is the predominant exchangeable cation.
A homoionic, Na-saturated clay was prepared by suspending the clay in a 2 mol/l solution
of NaCI for one day. The clay was then dialysed against deionized, distilled water until a
negative test for CI- was obtained using AgNO3, and finally freeze-dried.
The X-ray amorphous Fe-silicate (obtained from Inco Limited, Thompson, Manitoba,
Canada) was analysed by X-ray fluorescence spectrometry and the Fe(II) content was
determined by the method given by Hillebrand et al. (1953); the results are given in Table 1.
The material was crushed and sieved, and the < 74 #m particle size fraction was used in this
study.
TABLE 1. Elemental composition of the Fe-silicate
Element
wt~
Fe (total)
Fe(II)
Si
AI
Mg
Ca
K
Na
Ti
Mn
32.9
30.9
18.1
3.58
3.34
1'75
1.41
0.80
0-13
0.05
Effect of Fe(I1)-silicate on montmorillonitic clay
83
Solution
The suspending medium for the solids was a synthetic groundwater solution (SGW) (Table
2) whose composition was chosen to represent groundwaters found deep ( > 500 m) in the
Canadian Shield (Frape et al., 1984).
TABLE 2. Composition of
the synthetic groundwater
solution
Species
mg/1
Ca2+
Na+
Mg2+
K+
CIso~NO~
HCO~*
2140
1800
63
17
6460
1040
27
68
* Actual HCO~ concentration depends on the pH.
Experimental
Five grams each of the Fe-silicate and clay were suspended in 800 ml of the SGW in a 1 1
Erlenmeyer flask and reacted under oxic and anoxic conditions at 23 and 70 ~ for 200 days.
To separate the solids at the end of the experiment, the clay was contained in a semipermeable membrane tubing. (The membrane--obtained from Spectrum Medical Industries
Inc., Los Angeles, California--has a molecular weight cut-off of 12 000 to 14 000.) Control
experiments, without the Fe-silicate, were conducted for all systems under similar conditions,
and all the tests were performed in duplicate.
Oxic and anoxic conditions were maintained by continuously bubbling air or argon,
respectively, through the suspensions. For the anoxic systems, the SGW was deoxygenated
by bubbling argon through it for 4 h prior to adding the solids. One set of flasks (with both
oxic and anoxic conditions) was placed in a water bath at 70 + 0.2 ~ and another set was
placed on a laboratory bench where the temperature was 23 + 2 ~ over the reaction period.
Hereafter, the systems will be referred to as 23/ox (meaning 23 ~ and oxic conditions),
23/anox (23 ~
70/ox (70 ~
and 70/anox (70 ~
The air and argon were pre-saturated with moisture in a gas bubbler to minimize
evaporation loss from the samples. For the 70 ~ experiments both the sample flask and the
bubbler were placed in the water bath; a second bubbler, placed outside the water bath, was
found to be necessary to limit evaporation at 70 ~
The solutions were sampled by opening a swage-lok valve on a sampling tube. Because the
pressure inside the flask was slightly greater than atmospheric, solution was expelled from the
vessel. A 10 ml aliquot was carefully withdrawn (to minimize disturbance to the solid phases)
from each flask at the following times: 5, 10, and 14 days, then weekly to day 49, and then
biweekly to day 200. After filtering the solutions (0.45 #m Millipore filter), the pH was
measured with a glass electrode. The solutions were then acidified using HC1 and analyzed
for Fe, Si, AI, and Mg by inductively coupled plasma spectrometry (ICP), and for Ca, Na,
and K by atomic absorption spectrophotometry (AA).
84
D. W. Oscarson and R. B. Heimann
After 200 days, the clay was removed from the reaction flasks, dialysed against deionized,
distilled water for 24 h, and then freeze-dried. The clay was analyzed for total Fe and Fe(II)
content (Stucki, 1981), dithionite-extractable Fe (Fed) (Jackson, 1975), and oxalateextractable Fe (Feo) (McKeague & Day, 1966). The CEC of the clay was determined using
Ca 2+ and Mg 2§ as the saturating and displacing cations, respectively (Jackson, 1975). The
exchangeable cations (Ca 2§ Mg 2+, Na § K § Fe) on the clay were displaced using
ammonium acetate (Thomas, 1982), Mg 2§ and Fe were determined by ICP, and Ca 2+, Na +,
and K § by AA.
Two types of mounts were made from each clay sample for analysis by XRD. A random
powder mount was prepared by packing the clay into a holder, and an oriented aggregate was
prepared by using the porous tile technique (Kinter & Diamond, 1956). An XRD trace was
obtained from the powder, and three traces were obtained of the oriented aggregate after the
following treatments: (i) Mg-saturated and dried at room temperature, (ii) Mg-saturated and
solvated with glycerol (Srodon, 1980), and (iii) Mg-saturated and heated at 350 ~ for one
day. The patterns were obtained using Ni-filtered, Cu-K~ radiation generated at 50 kV and
150 mA. Rehydration of the clays after heating at 350 ~ was minimized by cooling the
samples in a desiccator, and by passing dry air through the specimen chamber of the
diffractometer during the X-ray examination (Oscarson & Miller, 1988).
Samples of the clays were also mixed with KBr powder (2 to 5 wt~ clay) and diffuse
reflectance spectra were obtained using a Bomem model DA 3-02 spectrometer.
RESULTS
AND
DISCUSSION
Solution chemistry
The final pH and concentration of various species in solution at the end of the 200 day
reaction period are given in Table 3. Examination of the data indicated that equilibrium was
established at about 150 days or earlier. Therefore, the data obtained from the last four
sampling periods (158, 172, 186, and 200 days) were averaged to arrive at the "final" values,
given in Table 3.
The pH in all systems was slightly alkaline, and did not vary by more than __ 0.3 pH units
throughout the entire 200 day reaction period in any of the systems. This suggests that the pH
of the suspensions was effectively buffered by the SGW and/or the clay.
TABLE 3. Final pH values and selected ion concentrations (mg/l) in solution in contact with the clay and Fe-silicate.
System
pH
Fe
Si
Ca
Mg
Na
K
23/ox
23/anox
7.4_+0.1.1 (7.3_+0.1) *2
<0.01 *3 (<0.01)
19.1_+1-3(18-2_+0.1)
2090 _+4 (2350 _+69)
66.8_+0.8(71.8_+2.0)
1880 _+8 (2180 _+56)
16.6_+0.2(23.0_+0.5)
7.6_+0-1 (7.6+0.1)
0-13_+0.03 (0.16_+0.02)
9.71_+0,05(13.5_+0-2)
1800 _+68 (2210 _+ 5)
62.3_+1-6(58-7_+0.9)
1620 _+60 (2060 _+ 5)
15.7_+0.1 (21,8_+0.3)
70/ox
7.6_+0-3 (7.4_+0.1)
<0-01 (<0.01)
55.8_+4.5(71.0_+1.5)
2290 + 110 (2510 -+ 52)
72.7_+2.2(53.9_+1,7)
2100 _+ 100 (2390 _+ 35)
19.7_+0.5(24.4_+1.3)
*~ Average _+one-half the difference between the duplicate determinations.
,2 Values in parentheses are for the control samples in which there was no Fe-silicate present.
*3 Detection limit,
70/anox
7.2_+0-1 (7.3_+0-1)
0-11 _+0.03(0.15+0.02)
42.6_+1.6(86.8_+2-8)
1650 _+60 (2070 _+ 154)
35.8_+3.0(19.2_+3.8)
1620 -+ 60 (1980 + 200)
52.3_+2.3(27.4_+0.2)
Effect of Fe(I1)-silicate on montmorillonitic clay
85
Under oxic conditions and at both temperatures the concentration of iron in solution was
below the detection limit of 0.01 mg/l. There were, however, detectable amounts of iron in
solution in the anoxic systems at both temperatures (Table 3). This is direct evidence that the
redox potential in the anoxic systems was significantly lower than that in the oxic systems-the lower the redox potential, the greater the solubility of Fe phases.
The predominant iron species in solution under the conditions of this study are Fe 2+,
FeOH § and FeSO o (Lindsay, 1979). Activity coefficients for these species were obtained
from the modification of the Debye-Hiickel equation proposed by Davies (1962). From the
activity coefficients and the total concentration of iron in solution, the activity of Fe 2§ in
solution in the 23/anox system was calculated to be ~ 10-7.3 mol/1; this suggests that ferrosic
oxide, Fe3(OH)s, (also referred to as fresh Fe precipitate) may be the solid phase controlling
the iron activity in solution in this system (Lindsay, 1979). Ferrosic oxide is an amorphous Fe
phase of mixed valence state. Its solubility product has been examined by Arden (1950), and
it is slightly more soluble than magnetite. It has been hypothesized that this Fe phase forms in
reduced soils (Ponnamperuma et al., 1967), but further work is necessary to verify this.
The final Si concentrations of 19.1 mg/l (23/ox) and 9-71 mg/l (23/anox) (Table 3)
correspond to H4SiO4 concentrations of 10-3'17 and 10-3.46 tool/l, respectively. These
systems are, therefore, supersaturated with respect to quartz (log K ~ -4.00) and
cristobalite (log K ~ -3.94), and undersaturated with respect to amorphous silica (log
K ~ = - 2.74) (Lindsay, 1979). The concentrations of Si in solution are significantly greater at
70~ than at 23 ~ under both redox conditions (Table 3), consistent with the fact that the
solubility of silica phases generally increases with temperature (Marshall, 1980; Fournier &
Potter, 1982). The concentration of AI in solution was below the detection limit of 0.1 mg/1 in
all systems.
The decrease in the level of Mg z§ in solution in the 70/anox system relative to the initial
concentration (Tables 2 and 3) suggests the precipitation of a Mg-rich phase. The significant
increase in the K § concentration in solution in this system (Tables 2 and 3) indicates that a
large fraction of the structural K (up to 27%) was released from the Fe-silicate during the
reaction period, probably as a result of the partial dissolution of the Fe-silicate. An increase in
the K § concentration in the pore solution of a smectite-based barrier material is undesirable
because K § accelerates the alteration of smectite to illite (Roberson & Lahann, 1981).
However, given the very low K+/(Ca z+ + Mg z+ + Na § ratio in deep groundwaters in the
Canadian Shield (Frape et al., 1984), the increase in K § originating from a barrier additive
such as the Fe(II)-silicate would likely not have a significant effect on smectite stability.
Nevertheless, any material added to the barrier material should have as low a K content as
possible.
Clay chemistry
The total iron content of the clays after the 200 day reaction period in the presence of the
Fe-silicate increased relative to the control samples by ~ 3% for both the 23/ox and 70/ox
systems (this small change is likely within the precision of the method of measurement) and
~ 7 and 32% for the 23/anox and 70/anox systems, respectively (Table 4). The greater amount
of iron associated with the clays under anoxic conditions reflects the greater mobility of iron
in the more reducing environments: under anoxic conditions, more Fe(II) will be in solution
where it can be sorbed on to the clay (which in this case could act as a sink for Fe). Also, after
the reaction period, when the clay was washed and dried, some of the sorbed Fe(II) may have
D. I41. Oscarson and R. B. Heimann
86
TABLE4. Content and nature of iron, CEC, and exchangeable cation composition
in the clay after the reaction period.
System
23/ox
Total Fe (wt~)
3.1.1 (3-0)*2
Fe(II) (wt%)
0.22 (0.20)
Fe(II)/Fe(III)
0-07 (0-07)
Feo (wt%)
0-36 (0.36)
Fed (wt~)
0.6l (0-71)
CEC mEq/100 g
73 (84)
Exchangeable cations mEq/100 g
Ca 2+
73 (89)
Mg 2§
3.4 (4.1)
Na §
3.2 (1.8)
K§
0.40 (1.4)
Total
80.0 (96-3)
23/anox
70/ox
70/anox
3.2 (3-0)
0-35 (0.25)
0.12 (0.09)
0.49 (0.36)
0.83 (0.66)
70 (88)
3-2 (3-1)
0.24 (0-20)
0.08 (0.07)
0.28 (0.30)
0.57 (0.60)
68 (72)
3.7 (2.8)
0-42 (0.28)
0.13 (0.11)
0.35 (0.26)
1.1 (0.58)
68 (72)
63 (90)
4.8 (7.7)
4.7 (1.7)
0-53 (1.3)
73.0 (101)
70 (66)
6.9 (11)
3.3 (1-5)
0.75 (1-3)
81.0 (79.8)
76 (68)
4-6 (7.8)
1-9 (1.6)
1.4 (1.1)
83-4 (77.9)
91 Average of the duplicate determinations; one-half the difference between the
duplicates is < 10~ of the average values in all cases.
92 Values in parentheses are for the control samples in which there was no Fesilicate present.
been oxidized to Fe(III), which would then hydrolyse and precipitate as Fe-hydroxide on the
clay surface.
The Fe(II) content of the clays was also greater in the anoxic systems (Table 4). This could
be due to the sorption of Fe(II) originating from the Fe-silicate (as discussed above) and/or to
the partial reduction of the Fe(III) originally present in the clay to Fe(II).
There was little difference in either the Feo or Fed contents of the clays in the oxic systems
relative to the control samples at both temperatures. On the other hand, in the anoxic systems
the amounts of both Feo and Fed increased significantly over those of the control samples of
clay (Table 4). This again is the result of the greater mobility of iron in anoxic environments.
No crystalline iron phase associated with the clay was detected by X R D in any of the
systems. The absence of the formation of crystalline iron phases is not surprising: various
solution species, such as silica, are known to inhibit the crystallization of Fe-hydroxide
(Schwertmann & Taylor, 1971).
Cation exchange capacity
Relative to the control samples, the C E C of the clay decreased by ~ 13 and 20% in the
presence of the Fe-silicate for the 23/ox and 23/anox systems, respectively, (Table 4). One
explanation for the decrease in the C E C is that Fe-hydroxide precipitated at the clay particle
edges, thereby blocking some of the cation-exchange sites. Generally, the edges of smectite
clays contribute ~ 10 to 15% of the total C E C of the clay, the remaining exchange sites being
on the basal, interlayer region (Grim, 1968). It is unlikely that Fe-hydroxide has precipitated
in the interlayer region, as discussed below. At 70 ~ the C E C of the clay decreased in both
redox environments by ~ 5% relative to the control sample. The reason for the greater
decrease, relative to the control samples, in the C E C at 23 ~ than at 70 ~ is not clearly
understood. One possible explanation is the formation of different iron phases at the two
Effect of Fe(I1)-silicate on montmorillonitic clay
87
temperatures--the formation of various iron phases depends on the temperature of the
system (see, for example, Lewis & Schwertmann, 1979). These phases have different particle
morphologies and surface properties, which, in turn, could affect the clay/Fe-hydroxide
interaction.
The CEC of the control samples of clay are also significantly less at 70~ than 23 ~ (Table
4). Some of the processes that could affect the clay at the higher temperature and in the
presence of the SGW, not related to the presence of the Fe-silicate, are increased dissolution
of the clay resulting in the destruction of exchange sites, formation of a precipitate on the clay
surface blocking some exchange sites, or alteration of the mineralogical composition of the
clay. However, no mineralogical changes were detected by XRD.
The decrease in the CEC of the clay in contact with Fe-containing materials may have
important consequences with respect to the long-term disposal of nuclear fuel waste. An
important function of the barrier material is to inhibit the migration of radionuclides from
the waste containers to the surrounding rock mass; one mechanism by which this may be
accomplished is the sorption of radionuclides on the clay component of the barrier material.
A decrease in the CEC of clays results in a concomitant decrease in their capacity to sorb
certain cationic radionuclides via ion-exchange reactions. This may be offset, however, by
the high sorption affinity for many species of the Fe-hydroxide associated with the clay
(Ames et aL, 1983; Jenne, 1968). It is also possible that mass transport processes in the barrier
material, such as the diffusion rates of radionuclides, could be affected by a change in the
CEC of the clay.
The exchangeable cation composition of the clays after the reaction period is also given in
Table 4. The fact that the sum of the exchangeable cations was greater than the measured
C EC indicates that there was some dissolution of soluble salts during the determination of the
exchangeable cations. On contact with the SGW, the clay was converted from a Na-saturated
to a predominantly Ca-saturated state (Table 4), and this is consistent with the greater
affinity of clay for divalent than monovalent cations. In addition to the cations listed in Table
4, the extracting solution was also analysed for iron, but none was detected; therefore, if Fe
species, such as Fe z+ and FeOH +, were present as exchangeable cations on the clay, the
amounts were <0-5 mEq/100g. No significant difference in the exchangeable cation
composition among the different systems is apparent.
Mineralogy
The d(001) XRD peaks of the clays from the 23/anox system are shown in Fig. 1. The clay
shows the typical expansion to 17-6 to 17.8A after saturating with Mg z+ and solvating with
glycerol, and a collapse to 9-6 to 10.0A after heating to 350 ~ (Brown & Brindley, 1984).
The patterns were identical for the clays from the 23/ox system and at 70 ~ under both redox
conditions (data not shown). These results indicate that no material was present in sufficient
quantity to cement the clay particles together to inhibit their expansion by glycerol. This
suggests that the swelling potential of the clay upon contact with water would not be
significantly affected by the iron material, although care must be exercised when
extrapolating from a glycerol to a water system, and further studies are necessary to confirm
this. The data also show there was no interlayer material in the clays to prevent their collapse.
Although the formation of Al-hydroxy material in the interlayer region of smectites is a well
known phenomenon (Rich, 1968), the evidence in the literature suggests that Fe-hydroxy
interlayers are only poorly developed, if at all (Carstea, 1968; Carstea et al., 1970).
D. W. Oscarson and R. B. Heimann
88
d(O01), .~.
25
I
16
i
12
!
I0 9
1 i
25
i
16
i
12
i
I0 9
i
i
control
b
a
5
I
i
i
i
5
I0
5
5
l
I
I0
degrees 20 (Cu-Ka)
FIG. I. X-raydiffraction peaks (d(001)) of the clays from the 23 ~
system after the 200
day reaction period: (a) Mg2+-saturated and glycerated; (h) Mg2+-saturated and dried at room
temperature (c) Mg2+-saturated and heated at 350 ~ for 24 h; "control" indicates the control
samples in which there was no Fe-silicate in the system.
From the XRD powder patterns, no new phases or changes in the original mineralogical
composition of the clays were identified in any of the experiments.
The diffuse reflectance spectra of the clays in contact with the Fe-silicate under both redox
conditions and at both temperatures were essentially identical to those of the clays from the
control experiments (data not shown), and no new bonding modes involving iron species
were evident.
CONCLUSIONS
The results of this study show that the amount and nature of Fe associated with clay can be
significantly altered when the clay is in contact with Fe-containing materials. This is
particularly true under anoxic conditions and at higher temperatures where iron is relatively
mobile.
The CEC of the clay decreased by up to 20~ when the clay was in contact with the Fesilicate, suggesting precipitation of Fe-hydroxide at the edge region of the clay, and the
blocking of some sorption sites.
The expansion of the clay upon contact with glycerol was not inhibited by Fe-hydroxide,
nor was there any evidence of the formation of Fe-hydroxy material in the interlayer region of
the clay.
In the light of the data presented, particularly the results of the CEC of the clay, further
studies are warranted, and these should focus on the swelling capacity and permeability of
Effect o f Fe(II)-silicate on montmorillonitic clay
89
c o m p a c t e d smectitic clays as affected by i r o n - c o n t a i n i n g materials. I f these properties are not
significantly altered by iron materials, and mass t r a n s p o r t processes such as the diffusivity
and the hydraulic c o n d u c t i v i t y o f the clay-based barrier are not affected, a decrease in the
C E C of the clay in a disposal vault m a y not be a significant factor.
ACKNOWLEDGMENT
We thank Mr. R. L. Watson and Ms. M.A.T. Stanchel for technical assistance.
REFERENCES
ALPEROVlTCHN., SHAINBERGI., KEREN, R., & SINGERM.'I. (1985) Effect of clay mineralogy and aluminum
and iron oxides on the hydraulic conductivity of clay-sand mixtures. Clays Clay Miner. 33, 443450.
AMES, L.L., MCGARRAn .I.E., WALKER B.A., & SALTER P.F. (1983) Uranium and radium sorption on
amorphous ferric oxyhydroxide. Chem. Geol. 40, 135-148.
ARDEN T.V. (1950) The solubility products of ferrous and ferrosic hydroxides. J. Chem. Soc. 882-885.
BLACK,lOREA.V. (1973) Aggregation of clay by the products of iron (III) hydrolysis. Aust. J. Soil Res. 11, 7582.
BROWNG. & BRINDLEYG.W. (1984) X-ray diffraction procedures for clay mineral identification. Pp. 305-360
in: Crystal Structures of Clay Minerals and their X-ray Identification (G.W. Brindley & G. Brown, editors).
Mineralogical Society, London.
CARSTEAD.D. (!968) Formation of hydroxy-Al and -Fe interlayers in montmorillonite and vermiculite:
influence of particle size and temperature. Clays Clay Miner. 16, 231-238.
CA~STEAD.D., HARWAROM.E. & KNOX E.G. (1970) Comparison of iron and aluminium hydroxy interlayers
in montmorillonite and vermiculite: I. Formation. Soil Sci. Soc. Am. Proc. 34, 517-521.
CLARK'I.S. & NICHOLW.E. (1968) Reactions of iron hydrous oxide with Wyoming bentonite. Can. J. Soil Sci.
48, 173-183.
DAVIESC.W. (1962) Ion Association. Butterworths, London.
EL-RAYAnH.M.E. & ROWELLD.L. (1973) The influence of iron and the permeability of a Na-soil. J. Soil Sci.
24, 137-144.
EL-SWA1FYS.A. & EMERSONW.W. (1975) Changes in the physical properties of soil clays due to precipitated
aluminum and iron hydroxides: I. Swelling and aggregate stability after drying. Soil Sci. Soc. Am. Proc. 39,
1056-1063.
FOURNIERR.O. & POTTERR.W. (1982) An equation correlating the solubility of quartz in water from 25 ~ to
900 ~ at pressures up to 10 000 bars. Geochim. Cosmochim. Acta 46, 1969-1973.
FRAPES.K., FRIVZP. & McNu'rr R.N. (1984) Water-rock interaction and chemistry of groundwaters from the
Canadian Shield. Geoehim. Cosmochim. Acta 48, 1617-1627.
GRIM R.E. (1968) Clay Mineralogy, 2nd ed., McGraw-Hill, New York.
HANCOXW.T. (1986) Progress in the Canadian nuclear fuel waste management program. Proc. 2nd Int. Conf.
Radioactive Waste Management, Winnipeg, Manitoba, 1-9.
HILLEBRANDW.F., LUNDELLG.E.F., BRIGHTH.A. & HOFFMANJ.I. (1953) Applied Inorganic Analysis, 2nd ed.,
John Wiley & Sons, New York.
JACKSONM.L. (1975) Soil Chemical Analysis--Advanced Course, 2nd ed. Published by the author, University
of Wisconsin, Madison, Wisconsin.
J E ~ E E.A. (1968) Controls on Mn, Fe, Co, Ni, Cu, and Zn concentrations in soils and water: the significant
role of hydrous Mn and Fe oxides. Am. Chem. Soc., Adv. Chem. Ser. 73, 337-387.
KINrER E.B. & DIAMONDS. (t956) A new method for preparation and treatment of oriented-aggregate
specimens of soil clays for X-ray diffraction analysis. Soil Sci. 81, 111-120.
LEwls D.G. & SCHWERTMANNU. (1979) The influence of aluminum on the formation of iron oxides. I~. The
influence of (A1), (OH), and temperature. Clays Clay Miner. 27, 195-200.
LINDSAYW.L. (1979) Chemical Equilibrium in Soils. John Wiley, New York.
90
D. W. Oscarson and R. B. Heimann
MARSHALLW.L. (1980) Amorphous silica solubilities--III. Activity coefficient relations and predictions of
solubility behaviour in salt solutions, 0-350 ~ Geochim. Cosmochim. Acta 44, 925-931.
MCKEAGUEJ.A. & DAY J.H. (1966) Dithionite- and oxalate-extractable Fe and AI as aids in differentiating
various classes of soils. Can. J. Soil Sci. 46, 13-22.
OSCARSOND.W. & MILLERH.G. (1988) A method for preventing the rehydration of heated smectites during Xray diffraction analysis. Atomic Energy of Canada Limited Technical Record TR-446.
PONNAMPERUMAF.N., TIANCOE.M. & LOY T. (1967) Redox equilibria in flooded soils: I. The iron hydroxide
systems. Soil Sci. 103, 374-382.
QUIGLEYR.M. (1984) Quantitative mineralogy and preliminary pore-water chemistry of candidate buffer and
backfill materials for a nuclear fuel waste disposal vault. Atomic Energy of Canada Limited Report AECL7827.
RlCri C.I. (1968) Hydroxy interlayers in expansible layer silicates. Clays Clay Miner. 16, 15-30.
ROBERSONH.E. & LAHANNR.W. (1981) Smectite to illite conversion rates: effects of solution chemistry. Clays
Clay Miner. 29, 129-135.
SCHWERTMANNU. & TAYLORR.M. (1977) Iron oxides. Pp. 145-180 in: Minerals in Soil Environments. (J.B.
Dixon & S.B. Weed, editors). Soil Science Society of America, Madison, Wisconsin.
SRODONJ. (1980) Precise identification of illite/smectite interstratifications by X-ray powder diffraction. Clays
Clay Miner. 28, 401-4tl.
STUCKIJ.W. (1981) The quantitative assay of minerals for Fe 2§ and Fe 3§ using 1,10-phenanthroline. II. A
photochemical method. Soil Sci. Soc. Am. J. 45, 638-641.
THOMASG.W. (1982) Exchangeable Cations. Pp. 159-165 in: Methods of Soil Analysis, Part 2, Chemical and
Microbiological Properties, 2nd ed., American Society of Agronomy, Inc., Madison, Wisconsin.