DIVISION S-2-SOIL CHEMISTRY

Calcium-Induced Sulfate Adsorption by Soils

N. S. Bolan, J. K. Syers,* and M. E. Sumner
ABSTRACT the anion and Ca2+ ions (Helyar et al., 1976). This
Gypsum (CaSO4-2H2O) is used in agriculture both as a source of involves coordination of one Ca2+ to two adsorbed
calcium (Ca2+) and sulfate (SO2; ) and as an amendment to improve anion groups, reducing the repulsive force between
soil structure. We examined the effect of Ca2+ on the adsorption of two adjacent anion groups and thereby enhancing fur-
SO2; in variable-charge soils. Sulfate adsorption measurements from ther adsorption. This mechanism, however, does not
batch and column experiments showed that SO2; adsorption increased account for the changes in surface charge with anion
with increasing adsorption of Ca2+. The increase in SO2; adsorption adsorption. Increased retention of SO4~ and r^PO.f
per unit increase in Ca2+ adsorption was 12 times more in soils con- at higher levels of Ca2+ addition has been attributed
taining Fe and Al hydrous oxides as the major variable-charge com- to precipitation reactions occuring at pH > 7.0 (Ad-
ponent than in soils dominated by organic matter. In soils containing ams and Rawajfih, 1977; Freeman and Rowell, 1982).
Fe and Al hydrous oxides, specific adsorption of Ca2+ increased the Specific adsorption of Ca2+ by hydrous oxides has
positive charge and thereby induced further adsorption of SO2; . At been shown to increase the positive charge on the
low levels of solution Ca2+ (<0.003 mol L-1), most of the increase in surface (Kinniburgh et al., 1975) and thereby increase
SO2; adsorption (85-98%) due to Ca2+ adsorption could be attributed the adsorption of anions. Recently, Marcano-Martinez
to the increase in positive charge. At higher Ca2+ concentration (0.003- and McBride (1989) and Alva et al. (1991) discounted
0.015 mol L'1), the increase in positive charge accounted for up to all these mechanisms and proposed a mechanism in-
only 75% of the increase in SO2; adsorption. The remaining increase
in SO2; adsorption is attributed to the coadsorption of Ca2+ and
volving CaSO4 ion-pair adsorption on mineral surface
SO2; as a CaSO° ion pair. In soils with organic matter as the major
in which the presence of one ion in high concentration
variable-charge component, Ca2+ is complexed by organic ligands. facilitates the formation of an ion pair and increases
Calcium complex formation through electrostatic attraction does not the adsorption of the other ion.
create positive sites and this may be the reason for the absence of In this study, the adsorption of SO4~ when added
Ca2+-induced SO2; adsorption in these soils. as either K2SO4 or CaSO4 was examined in four var-
iable-charge soils. The effect of Ca2+ on SO2.- ad-
sorption was examined in the context of the mechanisms
discussed above for the simultaneous retention of these
two ions.
MATERIALS AND METHODS
HE ADSORPTION of Ca2+ and SO|~ ions by varia-
T ble-charge soils is higher when added in combi-
nation as a salt, such as CaSO4-2H2O (gypsum),
Soil Characteristics
The soils used in this study were surface samples (0-150
mm) representing Healaugh (Udollic Albaqualf) and David-
than when added separately as CaCl2 and K2SO4, stow (Lithic Dystrochrept) series (typical brown podzolic soils)
(Marcano-Martinez and McBride, 1989). The in- from the United Kingdom and Seqaqa series (Typic Haplustox)
creased adsorption of Ca2+ in the presence of SO2." from Fiji, and subsoil sample (150-300 mm) representing the
could be explained by the increase in negative charge Fagaga series (Typic Paleustult) from Western Samoa (Table
2 1). The soils had varying mineralogy, clay content (Jackson,
created by the adsorption of SO ." and the increase in
pH due to the exchange of SO2" for OH~ ions (Rajan, 1956), organic C (Tabatabai and Bremner, 1970), oxalate-ex-
1978; Zhang et al., 1987; Curtin and Syers, 1990). tractable Fe and Al (McKeague and Day, 1966), and CEC
Ryden and Syers (1976) have concluded that the re- (Blakemore et al., 1987, p. 103), but were of predominantly
low pH.
tention of Ca2+ in response to H^PO^r sorption by soils
results from the increase in negative charge induced Sulfate Adsorption Isotherms
by H2PO4~ sorption. Various mechanisms 2have been Sulfate adsorption was measured at various concentrations
suggested, however, for the increase in SO ~ adsorp- of SO2- (0-0.015 mol L-1) using CaSO4 (pH 5.7) or K2SO4
tion in the presence of Ca2+. (pH 5.9) solutions. The concentrations of Ca2* and K> in these
The increase in the adsorption of anions, such as two solutions was maintained at 0.015 and 0.03 mol L-1 using
H2PO^- and SO2.- 4 in the presence of Ca2+ has been CaCl2 and KC1 solutions, respectively. The adsorption of
related to the formation of a surface complex between SOJ~ from K2SO4 solution was also measured in the presence
of 0.015 mol L-' CaCl2 as a supporting electrolyte. Soil sam-
M.S. Bolan, Fertilizer and Lime Research Centre, Dep. of Soil ples (4 g) were equilibrated with 20 mL of SO2- solutions by
Science, Massey Univ., Palmerston North, New Zealand; J.K. shaking on an end-over shaker for 16 h at 23 °C. At the end
Syers, Dep. of Agricultural and Environmental Science, The Uni- of the shaking period the supernatant solutions were separated
versity, Newcastle upon Tyne, NE1 7RU, England; M.E. Sum- from the soil by centrifugation (133 Hz) and2filtration (What-
ner, Dep. of Agronomy, Univ. of Georgia, Athens, GA 30602.
Contribution from the Dep. of Agricultural and Environmental man no. 41) for measurement of pH and SO .- . The concen-
Science, Univ. of Newcastle upon Tyne. Received 26 June 1991. tration of SQ- 4 in solution was measured by ion
'Corresponding author.
Abbreviations: CEC, cation-exchange capacity; AA, atomic ab-
Published in Soil Sci. Soc. Am. J. 57:691-696 (1993). sorption; UV, ultraviolet; BTC, breakthrough curve.

691
692 SOIL SCI. SOC. AM. J., VOL. 57, MAY-JUNE 1993

Table 1. Selected properties of the soils used in the study.

Oxalate-
Cation- extractable
exchange Organic
Soil Soil subgroup PH capacity C Fe Al Clay Clay minerals!
cmolc/kg g/kg
Fagaga Typic Paleustult 4.8 3.2 5.2 8.17 2.56 213 GI, GE, KK
Seqaqa Typic Haplustox 5.7 29.7 22.8 5.26 1.84 280 GE, GI, KK
Healaugh Udollic Albaqualf 6.9 17.6 32.5 0.50 0.52 322 IL, CL, VM
Davidstow Lithic Dystochrept 4.3 32.5 75.0 0.26 0.35 110 IL, CL, VM, MM
t Gibbsite (GI), goethite (GE), kaolinite (KK), illite (II), vermiculite (VM), chlorite (CL), and montmorillonite (MM).
chromotography (Tabatabai and Dick, 1983). The amount of and Fagaga soils from the breakthrough curves obtained with
SO2- adsorbed was calculated from the difference between the a step input of SO2.- using 0.015 mol L"1 CaSO4 or K2SO4
amount added and that remaining in solution after equil- solutions. The columns consisted of 200-mm-long glass tubes
ibration. In a separate 2experiment, the adsorption of SOJ- at with an internal diameter of 38 mm. The columns were packed
increasing levels of Ca * was measured. Sulfate concentration to a depth of 140 mm at a bulk density of 1.32 Mg m"3. The
was maintained at2+
0.015 mol L-1 using K2SO4 and the con-1 columns contained 210 g soil with a pore volume of 86 mL.
centration of Ca was varied between 0 and 0.015 mol Lr The columns were saturated from the bottom with distilled
using CaCl2 solution. The amount of SO2."2 adsorbed was mea- water. Solutions of 0.015 mol L"1 CaSO4 or K2SO4 were passed
sured as described above. Adsorbed Ca * was displaced by through the columns from the top at a flow rate of 0.42 mL
washing the soil with five aliquots of 20 mL of 0.25 mol Lr' min- 1 . The columns were leached with these solutions until
Mg(NO3)2 and the supernatant solution collected in a 100-mL the concentration of SO2- in the effluent reached that of the
volumetric flask (Mackay et al., 1986). Calcium in solution input solutions. The effluent solutions were collected using a
was measured by AA spectrophotometry. 2
The absorbed Ca2+ fraction collector and analyzed for Ca 2+
and K> using AA
was calculated after correcting for Ca * in the entrained so- spectrophotometry and SO2- by ion chromotography as de-
lution. scribed above.
The effect of Ca2+ adsorption on the positive charge of the
soils was also examined by measuring the amount of Q- ad-
sorbed at various concentrations of Ca2+. A range of Ca2* RESULTS AND DISCUSSION
concentration (0-0.015 mol L~') was maintained using CaCl12
and the concentration of Cl~ was maintained at 0.030 mol L- Sulfate Adsorption
2
using KC1. The adsorbed Ca * and Cl~ were extracted with Sulfate adsorption isotherms for the soils are shown
0.25 mol L-12 Mg(NO3)2 as described above and the concen- in Fig. 1. Adsorption decreased in the order: Seqaqa >
tration of Ca * and Cl- was measured using AA spectropho- Fagaga > Davidstow > Healaugh. Whereas Seaqaq and
tometry and UV Spectrophotometry (Florence and Farrer, 1971),
respectively. Fagaga soils contained significant amount of Fe and Al
hydrous oxides, chlorite, and vermiculite were the dom-
Column Experiments inant clay minerals in the Davidstow and Healaugh soils
Adsorption of SO2." was measured in columns of Healaugh (Table 1). It has often been observed that SO|- adsorp-
tion is higher in soils dominated by Fe oxides than by
layer silicate minerals (Parfit, 1978; Couto et al., 1979;
Singh, 1984).
In the case of Seqaqa and Fagaga soils, SOl~ adsorp-
tion was higher in the presence of CaSO4 than of K2SO4.
When adsorption was measured using K2SO4 with 0.015
mol L-1 CaCl2 as the background electrolyte, however,
SO4~ adsorption increased and approached that mea-
sured using CaSO4. For Healaugh and Davidstow soils,
the SO^~ adsorption isotherms obtained with various sys-
tems were not significantly different (plots not shown).
When the adsorption data were fitted to various ad-
sorption equations, which included Freundlich, Lang-
muir, Tempkin, and Gunary, the best fit was obtained
with the Freundlich equation, which gave the highest
values of coefficient of determination (R values). The
Freundlich equation has often been found to best de-
scribe the adsorption of SO^~ by soils (Singh, 1984) and
the adsorption parameter obtained from Freundlich equa-
tion has been used successfully in predicting the leaching
of SO|- in soils (Weaver et al., 1985; Bolan et al., 1988;
0 3 6 9 12 15
Curtin and Syers, 1990). There was only a slight differ-
Equilibrium sulfate concentration (mM) ence in the value of the exponential coefficient (ft) be-
Fig. 1. Sulfate adsorption isotherms for Seqaqa (A), Fagaga tween the different systems (Fig. 1). However, the linear
(•), Davidstow, (A) and Healaugh (O) soils measured with coefficient (A), which is related to the adsorption capac-
CaSO4 (————), K2SO4 in CaCl2 (-».--) and K2SO4 alone
(—...) solutions. For clarity, adsorption from CaSO4 and ity (coefficient A is the adsorption at unit concentration),
KjSO4 in CaCI2 for Davidstow and Healaugh soils are not was larger for Seqaqa and Fagaga soils than for Healaugh
shown. and Davidstow soils; also the values were higher for
 BOLAN ET AL.: CALCIUM-INDUCED SULFATE ADSORPTION BY SOILS 693
40
Y - 36.5 + 0.42 X (R - 0.94)

32
Y = 25.3 + 0.73 X (R = 0.91)

o
£24
T3
I Y = 18.6 + 0.44 X (R = 0.95)
6
•8 16

Y = 9.4 + 0.72 X (R - 0.94)

co
O 8

3 6 9 12 15 Y = 5.2 + 0.02 X (R - 0.95)

Equilibrium calcium concentration (mM)
Fig. 2. Calcium adsorption isotherm for Seqaqa (4), Fagaga
Y = 0.65 + 0.09 X (R = 0.97)
(•), Davidstow (A), and Healaugh (O) soils.
10 20 30 40
Calcium adsorbed (mmol kg"1)
adsorption measured with CaSO4 and K2SO4 in CaCl2
than for adsorption measured in K2SO4 alone. This sug- Fig. 3. Relationship between the adsorption of SO; (-
or Cl~ (-——) and adsorption of Ca2+ for Seqaqa (A), Fagaga
gests that more sites are created for the adsorption of (•), Davidstow (A) and Healaugh (O) soils.
SO2." in the presence of Ca2+ than in the presence of
10
Calcium Adsorption
Adsorption
2
isotherms for Ca2+ are shown in Fig. 2.
As for SO ^ , the Freundlich equation described the ad- o>
sorption data better than did the other equations as evi-
dent2+ from the higher R2 value. The /3 coefficients for o
Ca were much higher than those for SO4~ and were £
Y = 0.117 + 0.353 X ( R •= 0.93)
close to one for Healaugh and Fagaga soils. This indi-
cates that the isotherms for these two soils2+are approach- T3
CD
ing a linear form. The adsorption of Ca decreased in
the order: Davidstow > Seqaqa > Fagaga > Healaugh. O

The high adsorption of Ca2+ by Davidstow soil is prob- 3

O
ably related to its high organic matter content (7.5%).
There was only a small effect of increasing Ca2+ ad-
sorption on SO2" adsorption by Healaugh and Davidstow
soils (Fig. 3). SO2.- adsorption by Seqaqa and Fagaga e/i
soils, however, increased substantially with increasing <
Ca2+ adsorption. The slope of the linear relationship be-
tween SO4- and Ca22+ adsorption (Fig. 4) suggests that
the increase in SO ~ adsorption, per unit increase in 0 5 10 15 20 25 30 35 40
Ca2+ adsorption, was almost 12 times higher for Seqaqa Calcium adsorbed (mmol/kg)
and Fagaga soils than for Healaugh and Davidstow soils. Fig. 4. Relationship between the increase in SOJ adsorption
and Ca2+ adsorption for Seqaqa (A)» Fagaga (•), Davidstow
Chloride Adsorption (A), and Healaugh (O) soils.

Chloride adsorption increased with increasing

2
Ca2+
adsorption (Fig. 3). As was the case for SO ,", the in- Cl~ adsorption at high levels of Ca2+ adsorption could
crease in Cl~ adsorption with increasing Ca2+ adsorption be attributed to an increase2+in positive charge due to the
was higher for Seqaqa and Fagaga soils than for Dav- specific adsorption of Ca (Huang and Stumm, 1973;
idstow and Healaugh soils. Marcano-Martinez and Kinniburgh et al., 1975). Similarly, Ilton Morais et al.
McBride (1989) also observed that Cl~ adsorption in- (1976) observed that the distribution of charge on vari-
creased with increasing Ca2+ concentration. There was able-charge surfaces, as measured by ion retention, was
no evidence for the formation of CaCl+ ion-pair under influenced by the valence of the ion. In one of their soils
the present experimental conditions and the increased (Oxic Tropudalf B), which contained the largest amount
694 SOIL SCI. SOC. AM. J., VOL. 57, MAY-JUNE 1993

10

c
_o
1 08 2
1 = 0.51 X ' (R = 0.95) "g 0.8
"c
o
c
o ° 0.6
E •£
D

•a *. 0.4
a> CL
.Q C

0.2
D
ID
"5

3 2 1 i 3 4 5 6 8 9 10
<
Pore volume
Fig. 6. Breakthrough curves for SO^~ in Healaugh (————)
and Fagaga (——) soils with CaSO4 (A, •) and K2SO4 (A,
O) as the input solution.
4 S 12 16 20
A Chloride adsorbed (mmol/kg) gaga soil, which adsorbed considerably larger amounts
of SO2~ than did the Healaugh soil. The effect was more
Fig. 5. Relationship between the increases in Cl~ and SOj- pronounced when SO2.- was added as CaSO4 than as
adsorption due to Ca2+ adsorption in Seqaqa (A)» Fagaga K2S04. There was no difference, however, in the BTCs
(•), Davidstow (A) and Healaugh (O) soils. Also shown are obtained for Healaugh soil with CaSO 4 or K2SO4 addi-
2+
fitted line (————) and the line (•• — ) for equivalent increases
in Cl- and SOJ- adsorption (1/2 SO^~ /Cl~).
tion. In both soils the ETC
2
for Ca and K+ were dif-
ferent from those of SO " (data not shown).
The amount of SO2" adsorbed was calculated from
of Fe oxides, positive charges was higher in MgSO4 than the area above the BTC after correcting for the amount
in K2SO4. of SO2" retained in one pore volume of solution in the
Chloride adsorption at various levels of Ca22+ adsorp- column. Accordingly, the amounts of SO2- adsorbed1
tion, was comparatively lower than that of SO .- adsorp- from CaS04 and K2SO4 were 4.7 and 4.3 mmol 1kg-
tion and the difference was more pronounced for Seqaqa for Healaugh soil and 29.8 and 19.3 mmol2 kg- for
and Fagaga soils than for Healaugh and Davidstow soils Fagaga soil, respectively. The amount of SO - adsorbed
(Fig. 3). Similary, Kingston et al. (1972) and Rajan (1978) in the column experiment at an equilibrium concentration
observed that between 1.2 and 1.6 times more SO4- than of 0.015 mol L,-1 was slightly less than that measured
Cl" was absorbed by hydrous alumina and goethite. 2This in the batch experiment. It has generally been observed
suggests that, in addition to the positive sites, SO " is that adsorption is less in column experiments than in
also adsorbed at neutral and negatively charged sites batch experiments (Hodges and Johnson, 1987; Bolan et
(Marsh et al., 1987). al., 1988) and this has been related to the better contact
There was a good relationship between the increase in of solute and soil due to vigorous shaking in the batch
Cl~ and2+ the increase in SO4~ adsorption resulting from experiments.
the Ca adsorption (Fig. 5). The relationship was de- Effect of Calcium on Sulfate Adsorption
scribed by an exponential function that2 deviates from the
1:1 relationship at high levels of SO - adsorption. The Sulfate adsorption measurements in batch and column
function indicates that the initial increase in positive charge experiments showed that adsorption was higher in the
(as measured by Cl~ adsorption) due to Ca2 2+ adsorption presence of Ca2+ than of K + and it increased as Ca2+
accounted for most 2of the change in SO " adsorption. adsorption increased. The effect was more pronounced,
With increasing SO .- 4 adsorption, the increase in pos- however, in Seqaqa and Fagaga soils than in Healaugh
itive charge accounted for only a part of the SO2.-2 ad- and Davidstow soils. The net surface charge of these
sorption. In other words, the increases in SO " 4 soils was found to be affected by pH, which indicates
adsorption per unit increase in Cl~ adsorption decreased that they contained significant amounts of variable-charge
with increasing Ca2+ concentration. For example, the components (Bache, 1976; Bolan et al., 1991 unpub-
regression line shows that at CaSO 4 concentrations of
lished data). Whereas Fe and Al hydrous oxides formed
0.002, 0.006, and 0.015 mol Lr1, the increase in pos- the major component of variable-charge surfaces in Seqaqa
itive charge accounted for 97, 85, and 75% of the in- and Fagaga soils, in Healaugh and Davidstow soils or-
crease in SO2.- adsorption. ganic matter contributed the major variable-charge com-
ponent.
Several mechanisms have been proposed to explain
Breakthrough Curves the positive effect of Ca2+ on SO2," adsorption. These
Sulfate BTCs for Healaugh and Fagaga soils are shown include: (i) precipitation of gypsum and basic Al sul-
in Fig. 6. As expected from the batch experiments, phates; (ii) coadsorption
2+
of Ca2+2 and SO4" as an ion
804" appeared much later in the leachate from the Fa- pair; and (iii) Ca -induced SO - 4 adsorption. These
 BOLAN ET AL.: CALCIUM-INDUCED SULFATE ADSORPTION BY SOILS 695

three possibilities are discussed below in relation to the however, there was 2+ a two-fold increase in Cl~ adsorption
results obtained in the present experiments. with increasing Ca adsorption. Unfortunately, there
Precipitation of gypsum was ruled out because the were no SO2" adsorption data for these two soils using
concentrations of Ca2+ and SO4~ were deliberately kept CaSO 4. In our study, the increase in positive charge with
below 0.015 mol L~l and the soil systems are undersa- Ca2+ adsorption accounted for most of the increase in
turated with respect to the solubility of gypsum. Barrow SO2.- adsorption at low levels of Ca2+ (< 0.002 mol L-1)
(1972) observed that S0^~ adsorption by three Austra- in solution. The role of positive charge in SO4~ adsorp-
lian soils increased with increasing Ca2+ concentration tion by soils has been well documented (Marsh et al.,
but obtained no evidence for precipitation of CaSO4, 1987, 1988; Curtin and Syers, 1990). With increasing
even at 0.02 mol L-1 Ca2* concentration. Precipitation concentration of Ca2+ in solution, however, the increase
of SO]- as basic Al sulphates, basaluminite in positive charge was less than the increase in SO4~
[A14(OH)10(SO4) or alunite (A13(OH)6(S04)2] dependent adsorption and accounted for only a part of the SO4~ 4
on the presence of soluble A13+ (Adams and Rawajfih, adsorbed.
1977). The solubility product relationships for these two Equimolar adsorption of Ca2+ and SO4~ at high con-
precipitates indicated that the concentration of soluble centrations of CaSO4 has been considered as evidence
A13++ in the background electolyte 1
solutions used for for the coadsorption of Ca2+ and SO2- as an ion pair
SO^ adsorption (0.015 mol L,- CaCl 2 and 0.03 mol (Marcano-Martinez and McBride, 1989; Alva et al.,
L-1 KC1) was much less (< 1
0.001 mmol L-1) than that 1990). In one of the soils (Bladen) used by Alva et al.,
required (0.018 mmol L,- ) for the formation of these (1990) that contained few variable-charge materials,
precipitates at the highest level of SO|- (0.015 mol L-1) however, less SO2- was adsorbed than Ca2*. Similarly,
addition. This suggests that precipitation of basic Al sul- in our study the Davidstow stoil adsorbed the highest
phates would not have occurred under the experimental 2+
amount of Ca but this only had a small effect on
conditions. SO2' adsorption. These observations do not support the
Marcano-Martinez and McBride (1989) proposed a suggestion that the coadsorption of Ca2+ and SOj~ as an
mechanism based on the cooperative adsorption of Ca2+ ion pair is the only mechanism to explain the increase
and SO4- as a CaSO4 4 ion pair. Their conclusion was in SO2- adsorption with increasing Ca2+ adsorption. This
based on the following observations: (i) the increase in mechanism probably operates only at a high concentra-
pH following SO4~ adsorption was less when adsorption tion of CaSO4, at which the CaSO4 ion pair is likely to
was measured in CaSO4 than in K2SO4 solution, (ii) the be formed. Calculations using the GEOCHEM specia-
increase in positive charge created due to Ca2+ adsorp- tion model of Sposito and Mattigod (1980) indicated that,
tion was not sufficient to account for the increased under our experimental conditions, substantial amounts
SO4~ adsorption, and1
(iii) at high concentrations of 2+
CaSO4 of CaSO4 ion pairs formed only when the concentration
(> 0.001 mol L- ), equimolar quantities of Ca and of CaSO4 exceeded 0.012 mol L-1.
SO2" were adsorbed. Based on the observation made in our study and evi-
In this experiment, the increase in pH was less with dence obtained from the literature (Marcano-Martinez
CaSO4 than with K2SO4 addition (data not shown). Alva and McBride, 1989; Alva et al., 1990), the following
et al. (1990) have recently observed that, inspite of large mechanism is proposed to explain Ca2+2+-induced SO2"
amounts of SO2" being adsorbed from a saturated gyp- adsorption in soils at low levels of Ca in solution. For
sum solution, the pH of the soil solution was not af- Healaugh and Davidstow soils, in which organic matter
fected. They proposed that the pH of the soil when CaSO4 forms the major variable-charge component, Ca2+ is
is added is a net effect of the balance between the release probably adsorbed by complex formation through elec-
of protons (H + ) due to the specific adsorption of Ca2+, trostatic (Coulombic) attraction of Ca2+ onto the car-
and2 the release of OH- due to ligand3+exchange of boxyl and phenolic hydroxyl groups (van Dijk, 1971;
SO " with OH~ ions. Displacement of A1 by the added Bunzl et al.,1976):
cations could also cause pH changes. The OH~ released
as a result of a ligand exchange reaction was counter- rC-O- rC-O-...Ca2+
acted by the release of H + resulting from Ca2+ adsorp- Ca2+ ^
tion. This may be one of the reasons why the increase [1]
in pH with SO4~ adsorption is less in CaSO4 than in L
K2SO4. Similarly, Wada (1984) suggested that the simul- O-H
taneous adsorption of an anion and a cation by Ando
soils results in the neutralization of the H + and OH~ ions \ / COOH \/
COOCa^
released during adsorption and this process was termed
apparent salt absorption. Ca2+ ^ [2]
Marcano-Martinez and McBride (1989) measured pos-
itive charge by Cl~ adsorption and found that in two
soils there was2+ a slight increase in Cl~ adsorption with
\ OH \ OH
increasing Ca adsorption. Also the increase in SO4~ Calcium complex formation through electrostatic attrac-
adsorption due to Ca2+ adsorption was much higher than tion does not create any additional positive charges2+ and
that of Cl~ adsorption. This led to the conclusion that this is consistent with the small effect of Ca on
Ca2+-induced SO|~ adsorption was not solely due to an SO4~ adsorption in these two soils. Hence, in these soils,
increase in positive charge created by Ca2+ adsorption. a small amount of SO4~ is probably adsorbed by dis-
In two of their soils (Red Yellow Acrustox A and B placing OH- coordinated to a metal ion, m (Eq. 3) (Ra-
horizons) that contained large amounts of Fe oxides, jan, 1978; Zhang et al., 1987):
696 SOIL SCI. SOC. AM. J., VOL. 57, MAY-JUNE 1993

M-OH, M-OH,
+ SO2.- ^ + OH- [3]
M-OHJ M-S04J
In the case of Seqaqa and Fagaga soils, specific ad-
sorption of Ca2+ onto the oxide surface increases the
positive charge on the surface (Kinniburgh et al., 1975;
Huang and Stumm, 1975):
M-OH M-OCa
+ Ca2+ ^ + HH [4]
M-OHJ M-OHJ
In addition to the S04~ adsorbed at the original positive
sites (Eq. [3]), SO2+
4~ 4 is also adsorbed onto positive
sites created by Ca adsorption:
M-OCa M-OCa
+ SO^- = + OH- [5]
M-OHJ M-SO4J
The H* released during the specific adsorption of Ca2+
(Eq. [4]) may partly balance the OH~ released during
the SOl" adsorption (Eq. [5]). This2 is consistent with
the small increase in pH due to SO ^ adsorption from
CaSO4 compared with K2SO4.