EFFECT OF I R O N OXIDES ON POSITIVE A N D
NEGATIVE CHARGES I N C L A Y S A N D S O I L S
By M. E. SUMNER*
Soil Science Laboratory, University o f Oxford.
[Received 20th March, 1963]
ABSTRACT
The contribution of free iron oxides to positive and negative charge
distribution has been studied on soils and synthetic kaolinite-iron oxide
complexes over a wide range ofpH values. The charge on the iron oxides
is pH-dependent, being positive at low and negative at high pH values,
respectively, and accounts for a considerable proportion of the pHdependent negative charge in soils. Iron oxides increase the buffer
capacities of all soils and kaolinite-iron oxide complexes. A considerable proportion of the surface of the clay fraction is covered by iron
oxides, resulting in a decreased negative charge on the clay. The isoelectric points of iron oxides in soils are considerably higher than those
reported for pure iron oxides. A number of the soils studied were isoelectric at their field pH values, so that under field conditions they would
be expected to be extremely infertile. Evidence is presented to show
that kaolinite has a permanent negative charge, confirming the view that
isomorphous substitutions occur.
INTRODUCTION
Until recently the investigation o f the role played by free iron oxides
in modifying the surface properties o f soils and clays has received
little attention. There is, however, considerable fragmentary evidence in the literature to suggest that iron oxides have p r o f o u n d
effects on the physical properties o f soils. The importance o f iron
oxides in stabilizing the structure o f tropical soils has often been
mentioned but little direct evidence has been presented to support
the views expounded. Rather more evidence is available to show that
properties such as anion adsorption, swelling, and surface area are
modified by the presence o f iron oxides but few thorough and soundlybased investigations have been undertaken.
The aim o f this work was to investigate the extent to which free
iron oxides contribute to the variations in positive and negative
charges with pH. Two avenues o f approach are possible: (a)
the role o f the iron oxides m a y be studied on artificial soils synthesized
by precipitating the iron oxides most frequently encountered in
nature on kaolinite, or (b) natural soils m a y be studied before and
after the removal o f free iron oxides.
~Now Senior Lecturer in Soil Science, University of Natal, Pietermaritzburg,
South Africa.
218
EFFECT OF I R O N OXIDES ON SURFACE C H A R G E
219
EXPERIMENTAL
Materials. Details of the samples used in this investigation are
presented in Table 1,
Methods. Free iron oxides were removed by the sodium dithionite
method (pH 5-6) proposed by Mitchell and Mackenzie (1954) and
Mackenzie (1954). This treatment did not cause any alteration in
the surface properties of kaolinitic minerals (Sumner, 1962). Synthetic iron oxides were precipitated on the kaolinite surfaces by the
method proposed by Fripiat and Gastuche (1952). Positive and
negative charges were measured by a method in which alcohol
washing is eliminated (Schofield, 1949).
RESULTS
Synthetic kaolimte-iron oxide complexes. Three types of complex
were prepared--namely, those with hematite, goethite and lepidocrocite precipitated on kaolinite (sample 1, Table 1). The curves in
TABLE l--Properties of samples used.
~1
Type and locality
2
3
4
5
China clay, Cornwall I
Subsoil from lava
and ash, Kenya
Tropical red clay,
Nigeria.
Subsoil from dolerite, Natal
Subsoil from shale,
Malaya.
Clay 1 Fe~Oz*
(%) (4o)
pH
93 ]
Mineralogical
C
(%) composition t
0,0l
0,00 K, q
0.6
78
11-1
5-2
0.24 Gb, K, m, v,H !
0.2,
66
52
11.6
15-0
4.4
5-3
0.42 K,q,H
0.4~
0-01
0-00 K, v, q, Gb,
H,G
48
90
4.1
0-0~
O-OO K,G
* Free iron oxide expressed as Fe203.
"~ KEY: m--mica; K--kaolinite; v--vermiculite; q--quartz; Gb--gibbsite; H--hematite;
G--goethite; capital letters denote dominant minerals.
Fig. 1 show the variation in charge with pH for sample 1 on which
varying amounts of different iron oxides had been precipitated. As
the charge patterns for complexes containing hematite and goethite
were very similar, only the curves for the goethite complexes have
been presented.
In all instances, the positive charge is greater the higher the ironoxide content and is roughly proportional to the amount of iron oxide
added. With increasing pH the positive charge decreases until a
point is reached where negative adsorption of anions takes place;
this occurs in the region pH 6-7. The positive charges at low pH
values probably arise from the acceptance of protons by hydroxyl
groups on the iron oxides and edges of the clay crystals (Schofield,
1949; Schofield and Samson, 1954; Wiklander, 1955; Taylor, 1959).
220
M . E . SUMNER
Below about pH 6, the negative charge on the kaolinite is reduced
by the addition of the iron oxides and decreases with increasing iron
oxide content. Above this value the iron oxides increase the negative
charge. The negative charge on the kaolinite is constant between
pH 2.4 and 5, but above this there is an increase with increasing pH.
This indicates that the kaolinite has a permanent negative charge
6
I/
/I
5
ne
25~
Fe OON
b~ ~ - F-eOOH
10~ d~ - Fe OCj~
4
3
j 1I
pos
+5
O
-5
O
-b
CHARGE me'/,
j
=-=Plus 2 5 ~ g - F e O O H
9- , Plus 5% ~'- FeOOH
r~g
pos
+~
~
-g
CHARGE me%
FIG. l--Effect of synthetic iron oxides on the positive, negative and net charges on
kaolinite.
below pH 5 due to isomorphous replacement within the lattice as
suggested by Schofield (1949) and demonstrated by Robertson,
Brindley and Mackenzie (1954) and Holdridge (1959).
The reduction in negative charge caused by the iron oxides at low
pH values is probably due to (a) the iron oxides being bound to the
22I
EFFECT OF IRON OXIDES ON SURFACE CHARGE
kaolinite by the mutual neutralization of the positive charges on the
iron oxide and the negative charges on the clay, and/or (b) the negative
charges on the clay being physically blocked by an iron-oxide covering. This reduction in negative charge at low pH values suggests
that a considerable proportion of the basal plane surfaces of the
kaolinite is covered with iron oxides. The greater proportion of the
pH-dependent negative charge is accounted for by the iron oxides
(Table 2) and increases with increasing iron oxide content.
The net charge curves (Fig. 1) are essentially buffer curves for the
complexes. Iron oxides greatly increase buffer capacity, the greatest
//
pH
D,
J;
~
.
-.Untreoted
"
L
,
CHz~RGE
met
/
CH~GE
~%
FIG. 2--Effect o f free iron oxides o n the positive, negative and net charges on,
soils.
increases being shown by the samples containing most iron oxide
(Table 2).
It is clear from the curves in Fig. 1 that the precipitated iron oxides
behave amphoterically, having iso-electric points in the region pH
6-7. This result is similar to that of van Schuylenborgh (1950) for
pure synthetic oxides.
The kaolinite-iron oxide complexes containing 10 per cent. Fe20~
and 10 per cent. ~-FeOOH are amphoteric with iso-electric points
222
M.E. SUMNER
between pH 3.5 and 4.0. Presumably greater additions of these
oxides would yield complexes which are iso-electric at higher pH
values.
Soils. The charge distributions for the various soils studied are
presented in Fig. 2.
In general, the behaviour of the soils was very
similar to that of the kaolinite-iron oxide complexes.
Soils 2 and 4 develop high positive charges at low pH values,
whereas in other soils low (soil 3) or intermediate (soil 5) positive
charges are found. As distinct from the synthetic complexes, soils
do not show a direct relationship between positive charge and ironoxide content. For example, soils 2 and 3 have approximately the
same free iron-oxide content but the iron oxide in soil 2 develops 7.2
times as much positive charge as that in soil 3. This is to be expected
as the type of iron oxide, its surface area, and its degree of crystallinity
TABLE 2--Proportion of pH-dependent negative charge (pH 5.0-8.5) and buffer
capacity (pH 2.5-8.5) arising from iron oxides and clay.
Sample
1 +2"5 ~ a-FeOOH
1 + 5 % a-FeOOH
1 4-10 ~ a-FeOOH
! +2"5 % 7-FeOOH
1+ 5 ~ 7-FeOOH
14-10~ 7-FeOOH
pH-dependent negative
charge
Buffer capacity
Iron oxide
(%)
Clay
(%)
Iron oxide
(%)
Clay
(%)
41
53
77
30
53
66
59
47
23
70
47
34
52
70
82
45
62
73
48
30
18
55
38
27
52
23
50
31
48
77
50
69
59
23
65
58
41
77
35
42
are likely to be of greater importance in determining charge than the
actual free iron-oxide content. Soil 4 is unique in that it exhibits
positive charges at high pH values--in agreement with the views of
Quirk (1960). On deferrification, this positive charge disappears
entirely, indicating that it arises from the iron-oxide fraction. It is
possible that this positive charge at high pH values could arise from
isomorphous substitution within the iron-oxide structure (Norrish
and Taylor, 1961).
All deferrified samples exhibit a positive charge at low pH values.
In soils 3 and 5 the positive charge at pH 2-5 is 0.8 m-eq/100g and 0.6
m-eq/100g, respectively. This could be accounted for entirely by
the edges of the clay crystals. However, in soils 2 and 4 this charge
(2 m-eq/100g) is too high to be accounted for by the edges and it is
likely that aluminium and titanium oxides and hydroxides play a part.
EFFECT OF IRON OXIDES ON SURFACE CHARGE
223
In every soil the removal of iron oxides leads to an increase in
negative charge at low pH values and a decrease at high pH values.
This supports the view that a considerable proportion of the surfaces
of the clays are covered with iron oxides.
Previously the pH-dependent negative charge in soils (i.e. above
pH 5) was attributed to the edges of the clay crystals which become
negatively charged in this region (Schofield, 1939, 1949). However,
as can be seen from Table 2, a large proportion of this charge in fact
arises from the iron oxides.
The deferrified soils fall into two categories with respect to the
behaviour of the negative charge at low pH values. Soils 4 and 5
have a constant negative charge at low pH values, which supports the
view that isomorphous replacements can take place in kaolinite.
The remaining soils fall into the second category in which the negative
charge is not constant at low pH values. Assuming that most of the
negative charge at tow pH values on these deferrified soils is due to
isomorphous replacement,* it is necessary to postulate some mechanism to explain the observed variation in negative charge with pH
value. There are several possible causes, for example:
(a) During the determination of negative charge by Schofield's
method, it is essential to reduce the exchangeable AP + to an irreducible minimum, especially at low pH values. In soil 3 far greater
amounts of AP+ were liberated at equivalent pH values than in
other samples and great difficulty was experienced in reducing the
exchangeable AP + to a minimum. If more AI 3+ were held by the
exchange complex the lower the pH value, a variation in negative
charge with pH would be observed.
(b) The presence of small amounts of allophane-like material in
soil 2 (derived from lava and ash) could account for the pH-dependent
charge at low pH values.
The buffer capacities of all soils are increased by the presence of
iron oxides (Fig. 2, TaMe 2). In contrast to the synthetic kaoliniteiron oxides complexes, this increase is not proportional to the quantity
of iron oxide present.
It has been shown by van Schuylenborgh and Sanger (1950) and van
Schuylenborgh (1950) that the iso-electric points for highly crystalline
hematite and goethite were 2.1 and 3-2, respectively; the more highly
crystalline the iron oxide the lower the iso-electric point. The iron
oxides in soils (Fig. 2), however, appear to have higher iso-electric
points than the pure crystalline varieties. This indicates that the
iron oxides in soil are probably only poorly crystalline.
Soils 4 and 5 (untreated) are iso-electric at or near their field pH
values of 5.3 and 4.1, respectively. The net negative charge on the
remaining soils is much lower than on the clay minerals themselves.
Some of the soils studied contained small amounts of organic
matter (Table 1). The possible effects of this organic matter on the
*This assumption is not unreasonable in view of the shape of the curves and the
magnitude of the negative charge at low pH values.
224
M.E.
SUMNER
charge distribution were studied by comparing values before and
after treatment with hydrogen peroxide. The distribution was
unaltered by this treatment in soils 4 and 5. Soils 2 and 3 showed no
change at low pH values, but a slight decrease in negative charge at
high pH values.
Assuming uniform surface density of charge on the clay, the reduction in negative charge at low pH by the iron oxide, expressed as a
percentage of the negative charge on the clay at the same pH value,
should give a rough estimate of the proportion of the surface of the
clay covered by the iron oxide. Values for this together with the
maximum positive charge developed by the iron oxide at pH 2.4 are
presented in Table 3.
In most instances about 20-40 per cent. of the clay surfaces are
covered by the iron oxides, but soil 3 is an exception since only 9 per
TABLE 3--Proportion of clay surface covered and maximum p~sitive charge
developed by iron oxide.
Proportion of clay
surface covered by
iron oxide*
(%)
Maximum positive
charge on iron
oxide at pFI 2"4t
(m-eq/!0Og)
1+1o% Fe2Oa
1 + 10 % a-FeOOH
1+ 10 % "y-FeOOH
38
35
26
45
40
22
2
3
42
9
34
18
80
Sample
4
5
* Calculated as:
t Calculated as:
17
48
32
reduction in negative charge at low pH • 100..
negative charge on clay
net charge ( u n t r e a t e d ) - net charge (deferr)
percentage freo iron oxide
cent. of the surface is covered. The iron oxides in this soil appear to
be highly crystalline as reflected by the weak surface properties of the
iron-oxide fraction (Fig. l, Tables 2 and 3). Values for the proportion of the surface of two kaolinites covered by iron oxide,
calculated from the data presented by Fripiat and Gastuche (1952),
are in general agreement with the figures in Table 3.
The maximum positive charge developed by the iron oxides varies
from 20 m-eq/100g to 80 m-eq/100g. These fgures agree with those
o f 20 m-eq/100g for a pure sample of goethite (Sumner, unpublished)
and 60 m-eq/100g for the iron oxides in a Rothamsted subsoil
(Schofield, 1949).
DISCUSSION AND CONCLUSIONS
Part of the effects attributed to the iron oxides in this paper may
be due to aluminium oxides and hydroxides as some aluminium was
EFFECT OF IRON OXIDES ON SURFACE CHARGE
22~
removed by dithionite treatment. However, the contribution of the
aluminium compounds removed is likely to be of minor importance,
as only relatively small amounts of aluminium were extracted.
Iron oxides behave amphoterically in soil and contribute significantly to the buffer capacity of tropical soils rich in iron oxides.
The buffering exhibited by the iron oxides is probably due to proton
transfer and may be visualised as follows:
[Fe] H20
+ 89
+1
ACID
[Fe]~
OH
[Fe
1
0
I-EP
_.~
--1
ALKALINE
Previously it was thought that iron- and aluminium-rich compounds in the soil could develop positive charges in acid solution but
were uncharged in alkaline solution (Schofield, 1939, 1949). The
results of this investigation clearly demonstrate that iron oxides
develop negative charges under alkaline conditions and contribute
significantly to the pH-dependent negative charge in soils.
The amphoteric behaviour of iron oxides in soils has important
consequences in agriculture:
(a) A number of the soils investigated were found to be iso-electrie
at or near their field pH values. Under field conditions (low
electrolyte concentration), the positive and negative charges at
the iso-electric point will largely neutralize each other as the distances
separating them are small in comparison with the thickness of the
double layer. As a result, the inorganic colloidal fraction has little
power to retain nutrients that are essential to plant life. Such soils
should be extremely infertile--a fact which has been observed by
Orchard and Darby (1956; Darby and Orchard, 1956). Even in soils
which are not iso-electric, the positive charges developed by the iron
oxides serve to reduce the effective negative charge. Hence, such
soils would have a decreased capacity for holding cations. The use
of organic compost and manures on iso-electric soils will probably be
of greater value than inorganic fertilizers, as the humus formed by
decomposition will increase the cation-holding powers of the soil.
(b) On the basis of the present investigation, most conventional
methods for the determination of exchange capacity would appear
to be invalid on tropical soils, as these methods usually employ
alcohol-water mixtures for the removal of the excess salt solution
entrained in the soil. This washing leads to a decrease in the electrolyte concentration with a resultant expansion of the diffuse double
layer. Under such conditions positive and negative charges will
neutralize each other with the resultant loss of the cation and anion
previously held. This aspect is receiving further attention at the
moment.
226
M.E. SUMNER
Cation exchange in kaolinite has usually been attributed to unsatisfied valencies produced by 'broken bonds' at the edges of the
particles, chiefly because much of the information available seemed to
indicate an increase in cation-exchange capacity with decreased
particle size. Recent work by Worrall, Grimshaw and Roberts
(1958) has, on the contrary, shown that the exchange capacity in
kaolinites is virtually independent of particle sizes. Most of the
kaolinites studied in this investigation have permanent negative
charges of considerable magnitude at low pH values. This supports
the view that isomorphous substitutions take place within the lattice.
It would appear, therefore, that broken bonds only contribute to the
negative charge at high pH values and that a large part of the negative
charge in kaolinite is as in the other clay minerals due to isomorphous
replacements.
Acknowledgements.--The author desires to thank Mr T. W. Parker, English
Clays Lovering Pochin and Co. Ltd., St Austell, for the sample of Cornish kaolinite
and Mr R. Scott, EAAFRO, Kenya, and Dr A. R. McWalter, Department of
Agriculture, Malaya, for providing the soils from Kenya and Malaya, respectively.
REFERENCES
DARBY, G. D., and ORCI~ARD, E. R., 1956. Trans. 6th Int. Congr. Soil Sci.,
5, 263.
FRIPIAT, J. J., and GASTUCHE,M. C., 1952. Publ. I N E A C Ser. sci., No. 54.
HOLDRtDGE, D. A., 1959. The A. T. Green Book (N. F. Astbury, editor). British Ceramic Research Association, Stoke-on-Trent.
MACKENZIE, R. C., 1954. 3". SoilSci., 5, 167.
MITCr~ELL, B. D., and MACKENZIE,R. C., 1954. SoilSci., 77, 173.
NORRISH, K., and TAYLOR,R. M., 1961. J. SoilSci., 12, 294.
ORCHARD, E. R., and DARBY, G. D., 1956. Trans. 6th Int. Congr. Soil Sci.,
4, 263.
QtnRK, J. P., 1960. Nature, Lond., 188, 253.
ROBERTSON, R. H. S., ]3RINDLEY, G. W., and MACKENZIE,R. C., 1954. Amer.
Min., 39, 118.
SCHOFIELD,R. K., 1939. Soils andFert., 2, 1.
SCHOFIELD, R. K., 1949. J. Soil Sci., 1, 1.
SCHOFIELD,R. K., and SAMSON,H. R., 1954. Disc. Faraday Soc., No. 3, 51.
VAN SCHUYLENBORGH,J., 1950. Trans. 4th Int. Congr. Soil Sci., 1, 89.
VAN SCHUYLENBORGH,J., and SANGER, A., 1950. Rec. Tray. chim. Pays-Bas,
68, 999.
SUMNER, M. E., 1962. Agrochimica, 6, I83.
TAYLOR,A. W., 1959. J. Soil Mech. andFound. Div., 85, 19.
WIKLANDER, L., 1955. Chemistry of the Soil (F. E. Bear, editor). American
Chemical Society Monograph No. 126. Reinhold, New York.
WORRALL, W. E., GRIMSHAW,R. W., and ROBERTS, A. L., 1958. Res. Pap.
Brit. Ceram. Res. Ass., No. 405.