Published online 5 April 2007
Published in Soil Sci Soc Am J 71:703-710 (2007)
DOI: 10.2136/sssaj2006.0206
© 2007 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
SOIL PHYSICS
Simultaneous Adsorption of Calcium and Sulfate and Its Effect on Their Movement
Rogerio Cichotaa,*,
Iris Vogelerb,
Nanthi S. Bolanc and
Brent E. Clothierd
a Massey Univ., P.O. Box 11222, Palmerston North, Manawatu 4442, New Zealand
b HortResearch, Tennent Dr., Private Bag 11-030, Palmerston North 4442, New Zealand
c Massey Univ., P.O. Box 11222, Palmerston North, Manawatu 4442, New Zealand
d HortResearch, Tennent Dr., Private Bag 11-030, Palmerston North 4442, New Zealand
* Corresponding author (rcichota{at}hortresearch.co.nz).
 |
ABSTRACT
|
|---|
Ion retention in variable-charge soils can be enhanced by the presence of certain ions with opposite charge, thereby influencing the movement of these ions through the soil profile. Studies examining these interactions are still incipient, however, especially regarding its modeling. We present results from batch and miscible displacement experiments describing Ca2+ and SO42–movement in a variable-charge soil from New Zealand. Evidence was found for ion-pair adsorption (IPA) of both Ca2+and SO42– The results were modeled using the convection–dispersion equation (CDE), coupled with two different mathematical approaches proposed to account for IPA. The first approach related IPA to the single soil adsorption capacity, which is governed by particle-surface phenomena. For the second approach, IPA was related solely to the soil solution concentration. Both these approaches described the adsorption data from the batch experiment reasonably well, as well as the breakthrough curves from the miscible displacement experiments. The first approach showed better overall agreement. Significant differences were found, however, when the adsorption parameters were identified by fitting models to data from either batch or miscible displacement experiments. Although more studies are needed to better understand IPA, our results showed that the extent of IPA can be large and it should not be ignored when predicting SO42– and Ca2+movement in variable-charge soils.
Abbreviations: BTC, breakthrough curve • CDE, convection dispersion equation • IPA, ion-pair adsorption • PGA, particle governed approach • SGA, solution governed approach
|
INTRODUCTION
|
|---|
Ion adsorption onto soil particles can be influenced by the presence of other ions in the soil solution. Same-charge ions compete for adsorption sites, whereas opposite-charged ions may cooperate to increase adsorption. Specific adsorption of ions, which leads to an increase of the net soil charge, has been used to explain the results of several studies (Alva, 1993; Curtin and Syers, 1990; Cajuste et al., 1998; Bolan et al., 1999). Either anions or cations can drive this process, and subsequently increase the adsorption of opposite-charged ions. Also, interactions other than or additional to specific adsorption have been suggested (Ajwa and Tabatabai, 1995; Pearce and Sumner, 1997; Ishiguro et al., 2006). It has also been observed, especially in variable-charge soils, that cooperative adsorption can occur without changing the net soil charge (Marcano-Martinez and McBride, 1989; Fahrenhorst et al., 1999; Qafoku and Sumner, 2002). In this case, cation adsorption increases because of the presence of an anion, as well as vice versa. Because the amount of additional adsorption resulting from this co-adsorption process is equivalent to the ratio of the molar mass of the anion and cation, this process has also been called salt adsorption or ion-pair adsorption (IPA) (Marcano-Martinez and McBride, 1989; Qafoku and Sumner, 2002).
Ion-pair adsorption seems to be a common phenomenon in variable-charge soils (Pearce and Sumner, 1997; Qafoku and Sumner, 2001). Although it can occur with various ion combinations, it is more likely to happen with multivalent ions (Ajwa and Tabatabai, 1995; Pearce and Sumner, 1997). Ion-pair adsorption has been identified in particular for the adsorption of Ca2+ and SO42– in variable-charge soils (Marcano-Martinez and McBride, 1989; Bolan et al., 1993; Davis and Burgoa, 1995; Mora et al., 2005).
Various reasons have been suggested for the occurrence of IPA involving SO42– and Ca2+, and some physicochemical models attempting to describe IPA have been presented. Marcano-Martinez and McBride (1989) and Bolan et al. (1993) invoked charge-balance models to describe the mechanisms of specific adsorption of SO42–, or Ca2+, which is favored by the presence of the counter ion. The release of potential-determining ions such as H+ or OH– is balanced by the subsequent adsorption of the counter ion. Although they present differences, both models use surface-charge reactions to explain IPA, and therefore will be regarded here as one common model. In contrast, Qafoku and Sumner (2002) presented an ingenious physical model to explain IPA. According to their model, in highly weathered soils composed mainly of two different soil components with opposite net charge (kaolinite and Fe or Al oxides, for example), the double layer of each particle type can expand and overlap the zone of influence of the other. In these intersections, the electrical field of one particle balances the other. According to these researchers, ions can be trapped as ion pairs in these intersections, because there is no need for electrical charge balancing. This model is in agreement with several observations when IPA occurs, such as no change in the net soil charge and pH; the additional adsorption is in equivalent amounts of counter ions; and the fact that IPA is observed when the soil contains opposite-charged particles (Wada, 1984; Qafoku and Sumner, 2002). None of these approaches, however, provide a means to describe the amount of solute adsorbed as ion pairs and its subsequent equilibrium with the soil solution concentration.
It has been shown that the SO42– adsorption isotherms are significantly distorted when Ca2+ is present in variable-charge soils (Marcano-Martinez and McBride, 1989; Bolan et al., 1993; Mora et al., 2005). Higher adsorption in the presence of the counter ion leads to biased estimation of the adsorption parameters. The use of these parameters could result in large errors for predictions made under different conditions, such as with different concentrations of the counter ion. Some studies have already shown that IPA can significantly influence the leaching of ions from the soil (Bolan et al., 1993; Davis and Burgoa, 1995; Qafoku and Sumner, 2001). In miscible displacement experiments, the classical breakthrough curve (BTC) is delayed when IPA is present, because the retardation factor is increased. The normal sigmoidal shape of the BTC can also be distorted depending on the extent of IPA. Coupled SO42– and Ca2+ transport in the soil with IPA has not been well explored yet, and the modeling of the leaching of these ions under such conditions is still incipient, especially when attempting to quantify the extent of IPA.
Contrasting batch data with miscible displacement experiments may provide insights for further understanding of this phenomenon. It has already been shown that adsorption parameters estimated by batch experiments can differ significantly from those found by methods with smaller water/soil ratios, as in the unsaturated transient-flow method (Katou et al., 2001) or miscible displacement experiments (Altfelder et al., 2001; Maraqa, 2001). Such divergences seem not to have been investigated when IPA is present.
Allophane is one of the most important constituents with variable charge in volcanic soils, thus being related to its pH-dependent net charge and adsorption of ion pairs (Pigna and Violante, 2003; Mora et al., 2005; Ishiguro et al., 2006). Gypsum application is often recommended in these soils for supplying both Ca2+ and SO42– resulting from reduced leaching due to the enhanced adsorption (Mora et al., 2005). Using an allophanic variable-charge soil from New Zealand, a series of batch and miscible displacement experiments was conducted to investigate the effect of cations on SO42– leaching. Evidence of IPA for Ca2+ and SO42– in this soil are illustrated. We also present two approaches for modeling the partition of the solutes between the adsorbed and solution phases to achieve thermodynamic equilibrium. Thus, we expect to contribute toward a better understanding of this phenomenon and its relevance for solute leaching and retention in the soil.
|
MATERIALS AND METHODS
|
|---|
Soil
The soil used for this study was the Egmont loam, an allophanic soil (Typic Dystrandept) from South Taranaki, New Zealand. It is a weathered soil of volcanic origin, containing allophane as the main clay mineral, but also considerable amounts of Fe and Al oxides. It has a well-developed structure and the bulk density is usually <900 kg m–3 (Molloy, 1988). Soil samples were collected from the top 0.20-m depth. Some characteristics are presented in Table 1.
Batch Experiments
Three series of solute adsorption isotherms were determined: SO42– adsorption isotherms at different levels of Ca2+; Ca2+ isotherms with different amounts of SO42–; and the simultaneous adsorption isotherms for Ca2+ and SO42– added at equimolar concentrations. Several adsorption values of Ca2+ and SO42– were measured at various concentrations of these solutes (0, 1, 2, 4, 6, 8, 10, 12, and 15 mmol L–1), with duplicate samples for each equilibration. Sulfate solutions were prepared with K2SO4 and Ca2+ solutions with CaCl2. Equimolar solutions were prepared using CaSO4, and KCl was used to maintain the ionic strength constant.
Because this soil already had considerable indigenous amounts of SO42– and Ca2+, the soil was first leached with a solution of MgCl2 (10 mmol L–1), followed by a solution of KCl (10 mmol L–1), and finally distilled water to reduce the influence of the counter ions. This procedure lowered the indigenous (Table 1) Ca2+ and SO42– levels to 6.5 and 0.8 mmol kg–1, respectively. Adsorption isotherms were determined using the leached soil, and checked against samples of the original soil.
Four grams of air-dried soil was equilibrated with 25 mL of solution in centrifuge tubes by shaking for 24 h in an end-over-end shaker at room temperature. After this equilibration period, the samples were centrifuged for 5 min at 8000 rpm, and filtered (Whatman no. 42) to separate the solution from the soil. The solution was then analyzed; the adsorbed amount was computed as the difference between the amount in the equilibrium solution before and after equilibration. Sulfate was determined by the methylene blue method (Johnson and Nishita, 1952), and Ca2+ using atomic absorbance spectroscopy. The detection limits were approximate 0.005 mmol L–1 for S and 0.025 mmol L–1 for Ca.
Miscible Displacement Experiments
Four miscible displacement experiments were performed with repacked columns using Egmont soil. Details of the experiments are presented in Table 2. Two columns (E1 and E2) were leached with different concentrations of CaSO4 following a preleaching with CaCl2, thus allowing examination of the influence of IPA on the BTCs at different solute concentrations. Another column (E3) was leached with K2SO4 in a Ca-impoverished soil due to preleaching with KCl. This produced a BTC for SO42– with minimum influence of Ca2+. Finally another column was leached with solutions containing Ca2+ only, using CaCl2, but with the same step change in Ca2+ concentration as column E2. The columns E3 and E4 therefore present the situation where SO42– and Ca2+, respectively, are leached with the same concentrations as in columns E2, but with minimum IPA.
View this table:
[in this window]
[in a new window]
|
Table 2. Concentration of the leaching solutions and some physical characteristics of the soil columns used in the miscible displacement experiments on Egmont soil.
|
|
All leaching experiments were conducted under unsaturated conditions, at a constant pressure potential of –0.1 kPa (at both column ends). The columns were 0.15 m high and had a diameter of 0.10 m. The experiments were performed following a procedure similar to that described by Magesan et al. (2003). The solution was applied using a disk permeameter placed in the top of the column. The leachate solution was collected in small aliquots of between 50 and 100 mL; the collecting containers were in a pressure-controlled chamber (Fig. 1
). These aliquots were analyzed for SO42– and Ca2+ concentrations as described for the batch experiment. Chloride concentration was also analyzed using the Hg thiocyanate–Fe method, with a detection limit of 0.085 (Florence and Farrar, 1971). Similar analyses were performed for the soils before and after leaching, with SO42– being extracted with a KH2PO4 solution, and Ca2+ using NH4OAc.
Partition of the Soil Solute Concentrations Considering Ion-Pair Adsorption
We used the simple approach where the total solute in the soil (Q, mol m–3) can be divided into adsorbed (S, mol kg–1) and solution (C, mol m–3) phases. The mass balance for a given ion in the soil can therefore be described by
 | [1] |
where 
is the soil water content (m3 m–3) and 
is the soil bulk density (kg m–3). The quantities S and C should be in a thermodynamic equilibrium:
 | [2] |
Considering that the distribution coefficient, Kd (m3 kg–1), is invariant for both adsorption and desorption processes, different functions have been presented to relate the concentration in the soil solution and the amount of solute adsorbed onto the soil particles. This adsorption component we call single adsorption (SS, mol kg–1), and it is described by the nonlinear Freundlich isotherm:
 | [3] |
where N is an empirical fitting parameter.
As shown in batch experiments described elsewhere (Marcano-Martinez and McBride, 1989; Bolan et al., 1993; Ajwa and Tabatabai, 1997), if IPA is present, the amount of adsorbed ion increases in the presence of its counter ion. Therefore, IPA is regarded as an additional adsorption over the single one (SIP, mol kg–1). The total amount of adsorption will then be the sum of these two quantities:
 | [4] |
The partition between the different solute components in the soil (C, SS, and SIP) that leads to thermodynamic equilibrium has not yet been explored. In an attempt to identify this partition and to describe the results of the column experiments, two approaches are presented here. The first one is designated a particle governed approach (PGA). In this approach, it is assumed that IPA is driven by particle surface interactions similar to those of the single adsorption. It is therefore implicit that the higher the single adsorption (SS), the higher the additional one (SIP) should be. For this we follow observations that the amount of additional adsorption of SO42– in the presence of Ca2+ can be reasonably well described using a linear regression relationship between SO42– and Ca2+ adsorbed concentrations (Bolan et al., 1993). Thus, assuming that such relationship is also linear for Ca2+ additional adsorption, the following equation can be derived:
 | [5] |
where Ks (kg mol–1) is the IPA factor, and considered to be independent of the solute concentration.
Alternatively, the value of SIP can be determined by assuming an equilibrium between the additional amount adsorbed and the soil solution concentration, similar to that of Eq. [2]. An equilibrium coefficient, or ion-pair factor (Kc, L mol–1), is used to relate these amounts. With this approach also, it is predicted that Kc does not depend on the concentration. This approach does not imply any direct interaction with the soil particle surfaces. Therefore this is designated a solution governed approach (SGA), and can be written as
 | [6] |
where the ratio between soil water content and soil density is used to adjust the units. This is the water/soil ratio.
Equation [6] is similar to the stability equation of ion pairs dissolved in an aqueous solution, where the amount of the ion pairs (CaSO40) is proportional to the product of the concentration of the dissociated ions (Ca2+ and SO42–). It seems reasonable to assume that the higher the concentration of CaSO40, the greater the likelihood of this ion pair to be adsorbed. For example, these ion pairs could be trapped in the neutral overlaps of the particles' double layer following the model presented by Qafoku and Sumner (2002). On the other hand, there has been no detailed investigation of this.
Prediction of the partitioning of the total concentration of each solute into C, SS, and SIP was performed using a Newton–Raphson procedure to solve the equation
 | [7] |
where SIP is equivalent for both solutes. The equilibrium was assumed instantaneous, following Eq. [3] for SS and Eq. [5] or [6] for SIP. Because the values of S were found to be more sensitive to the changes in the adsorption parameters, they were used to evaluate the model fittings by contrasting predicted with observed values.
Solute Movement in the Soil
The concentration of each solute in the soil solution, at time t (s) and depth z (m), was predicted using the CDE (Eq. [8]), assuming the water flux density (qw, m s–1) to be steady. The dispersion coefficient (D, m2 s–1) was estimated using data from both solutes and also the chloride BTCs.
 | [8] |
To solve Eq. [8], a finite difference model with central differences was written based on Eq. [9]. The soil column was divided into layers of thickness 
z (m), where the solute concentration was considered homogeneous and in equilibrium with the soil matrix during each time step t (s). After computing the movement at each t, variations in Q were partitioned into C, SS, and SIP following the partitioning procedures described above. Both solute movement and adsorption were simulated using a computer procedure written in Visual Basic language.
 | [9] |
|
RESULTS AND DISCUSSION
|
|---|
Batch Experiment
The adsorption of both Ca2+ and SO42– was clearly affected by the presence of the respective counter ion. Figure 2
shows examples of SO42– and Ca2+ isotherms for the leached soil at three different levels of the counter ion. Also shown, for comparisons between these sets and with those from other studies, are the adjusted Freundlich isotherms (without considering IPA). The parameters are in the range reported elsewhere (Marcano-Martinez and McBride, 1989; Bolan et al., 1993; Mora et al., 2005), although comparisons are difficult because of differences in soil mineralogy and especially due to the strong dependency on pH of the surface charge of the allophanic soils (Bolan et al., 1999; Mora et al., 2005). In our case, no significant changes in pH were found when varying the concentration of the equilibrium solution, except for those with a higher concentration of K2SO4, where a slight increase in pH occurred, indicating some specific adsorption of SO42–. Nonetheless strong evidence for IPA is provided by the enhanced adsorption of both ions when its counter ion is present in increasing amounts.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 2. Observed adsorption isotherms, where S is the adsorbed phase and C is the solution phase, for (a) SO42– and (b) Ca2+ using three different amounts of the counter ion in the added solution. Lines show adjusted Freundlich isotherms without considering ion-pair adsorption (IPA).
|
|
Table 3 shows the adjusted parameter values for the isotherms using the two approaches for describing IPA. Both approaches, PGA or SGA, allow solving the equations for the equilibrium between solution and sorbed concentrations quite easily, although the solution must be done simultaneously for both ions. Good agreement between predicted and measured values was found with both approaches. There are considerable differences, however, in the shape of the isotherms evaluated by the two approaches. In particular, the values for SIP were estimated to be much higher using PGA than with SGA, especially at lower concentrations. This causes the shape of the single adsorption isotherm to be significantly different between the two approaches. In Fig. 3
, isotherms for both SO42– and Ca2+ obtained equilibrating different amounts of CaSO4 are presented (therefore in approximate equimolar amounts), and modeled using both IPA approaches. This example shows how the estimates of SIP differ when increasing the solute concentration. Because both approaches use a product formula (of either SS or C), there is a tendency to exponentially increase the estimates of SIP when the amount of solute increases. As C is favored at higher concentrations (due to N < 1), SGA estimates increasingly higher values for SIP. This produces a rather unusual shape for the total adsorption (Fig. 3c). This feature is not as pronounced in the other sets, where the amount of the counter ion was fixed. Nevertheless, it indicates that this approach is more complex than in the simple approach presented here, demanding further investigation. Marcano-Martinez and McBride (1989) have also found the isotherm (for Ca2+) with N > 1 when equimolar amounts of Ca2+ and SO42– were added to the soil.
View this table:
[in this window]
[in a new window]
|
Table 3. Adjusted parameters of Freundlich isotherms (distribution coefficient Kd and fitting parameter N) and the ion-pair adsorption factor (Ks or Kc) for SO42– and Ca2+. The adjustment was made using batch experiment data. Also presented are the agreement measures (contrasting measured against estimated values of adsorbed component SSO4+SCa).
|
|
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 3. Adsorption isotherms, where S is the adsorbed phase and C is the solution phase, for SO42– and Ca2+, added in equimolar amounts. Dots show observed data, bold lines represent S, and thin lines show the single adsorption component, SS, as estimated using (a and b) the particle governed approach, PGA, or (c and d) solution governed approach, SGA.
|
|
It was not possible to determine the Freundlich parameters for the single adsorption simply by regression analysis of the data set in which no counter ion was added, because the soil still had some residual amount of Ca2+ and SO42–. The best fit using the three-phase equilibrium model (Eq. [7]) gave quite unusual values for the Freundlich isotherm parameters, especially when SGA was used to describe IPA. In particular, the value of the exponent N was relatively small compared with those reported in the literature (Marcano-Martinez and McBride, 1989; Bolan et al., 1993, 1999; Mora et al., 2005). In fact, the shape of the SO42– isotherms suggests that the maximum limit for the single adsorption of this ion had been reached.
It is also important to stress that large uncertainties or correlation between parameters may be expected because of the large number of parameters (five) that need to be adjusted simultaneously. For example, the value of N will be clearly restricted when the IPA factor is increased. The problem of adjusting models with a large number of parameters has been the focus of several studies (e.g., Abbaspour et al., 2004; Vrugt et al., 2005; Beulke and Brown, 2006; Mertens et al., 2006), and frequently involves extensive computer work. Although care must be taken when using such multiparameter models, the problem of parameterization does not invalidate their use. A better understanding of the sensitivity and correlation of the model parameters is important but beyond the scope of this study. The sensitivity analysis of various parameters will be examined in a subsequent study.
Miscible Displacement Experiments
Clear evidence for IPA was also found in the miscible displacement experiments. The adsorption of both Ca2+ and SO42– increased in the presence of the other. The breakthrough of the SO42– solution of Columns E1 and E2 occurred at a different time to that of column E3 (Fig. 4a
). This indicates that the presence of Ca2+ in the former has retarded the movement of SO42– in the soil due to increased adsorption. Furthermore, for Columns E1 and E2, the Ca2+ concentration in the leachate dropped to almost zero when the accompanying anion changed from Cl– in the preleaching to SO42– in the main leaching phase. This contrasts significantly with Column E4, which was leached without replacing Cl– by SO42– (Fig. 4b). Retardation of the BTCs and changes in their shape have been reported (Bolan et al., 1993; Qafoku and Sumner, 2001; Nakajima et al., 2002) and IPA was used to explain those results. In addition to this, the final solute concentration in Columns E2, E3, and E4 can be compared, as they can be regarded as replications that contained either both Ca2+ and SO42–, only SO42–, or only Ca2+, respectively. The difference in SO42– between Columns E2 and E3 represented an additional retention of 18.4 mmol kg–1. This is about the same amount as the difference for Ca2+ between Columns E2 and E4, which was 19.2 mmol kg–1.
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 4. Measured breakthrough curves for (a) SO42– and (b) Ca2+, for the four experimental columns (PV = pore volumes). Columns E1 and E2 leached with CaSO4, E3 leached with K2SO4, and E4 with CaCl2 (see Table 2).
|
|
Although we had no replications, these comparisons are relevant considering the good mass balance attained, with recovery deviations within a 7% limit. Similar to the batch experiment, the pH increased when K2SO4 was added to the soil (Column E3), and the change was equivalent to that experiment (from 4.7 to 5.2). For Columns E1 and E2, the pH before and after the introduction of SO42– solution was the same, although a peak occurred during the displacement. No changes in pH occurred in Column E4.
Problems of non-uniqueness can be found when evaluating parameters by fitting models to miscible displacement data, because of the number of unknown parameters and their related effects on BTCs' shape (Friedel, 2005; Kohne et al., 2006). Obtaining larger amounts of data and independent estimation of some parameters are used to reduce this problem. In our case, the value of the dispersion coefficient (D from Eq. [8]) was obtained using the Cl– data, and the adsorption parameters found with the batch experiment data were used as starting estimates. Frequently, however, adsorption parameters found in batch experiments cannot be used for describing miscible displacement results (Heng et al., 1999; Katou et al., 2001; Maraqa, 2001). The different hydraulic regime, especially the water/soil ratio and the time for equilibrium, is generally suggested as the reason for this disagreement. The Cl– data showed no significant adsorption in this soil (Fig. 5
). Fitting a CDE model to the Cl– data revealed consistent values for D. As this parameter has only a small effect, the average value of 2.0 m2 s–1 was assumed for all columns. Table 4 shows the adsorption parameters found by inverse modeling using Eq. [9], and the two IPA approaches to describe the data of Columns E1 and E2. Interestingly, while the values of the distribution factor were much smaller than those of the batch experiment, as expected, the IPA factor was found to be much larger, especially for PGA. This suggests that IPA is more likely to occur at smaller water/soil ratios, contrary to the single adsorption. Frequently the restricted access to adsorption sites is used to explain the decrease in the single adsorption found in miscible displacement experiments. This would also apply for IPA, unless the phenomenon responsible for IPA is different from that of the single adsorption.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 5. Measured Cl– breakthrough curves for the four experimental columns; also shown is the adjusted model. Concentration is relative to the input solution (PV = pore volumes). Columns E1 and E2 leached with CaSO4, E3 leached with K2SO4, and E4 with CaCl2 (see Table 2).
|
|
View this table:
[in this window]
[in a new window]
|
Table 4. Adjusted parameters of Freundlich isotherms (distribution coefficient Kd and fitting parameter N) and the ion-pair adsorption factor (Ks or Kc) for SO42– and Ca2+. The adjustment was made using data from columns E1 and E2 of the miscible displacement experiment (using CaSO4 solution). Also presented are the agreement measures (measured vs. estimated values of the leachate concentration).
|
|
The BTCs modeled with IPA predicted the results from the column experiments reasonably well (Fig. 6
). The description using the PGA resulted in better agreement, however, as we were able to better simulate the drop of Ca2+ in the presence of SO42–. This difference between the two approaches is a result of the fact that, at low concentrations, PGA estimates higher amounts of IPA (see isotherms on Fig. 3). Equilibrium may, therefore, occur with a very low value of C and a larger value of S (SS + SIP). For SGA, only a small fraction of the solute is adsorbed by IPA in relation to the single adsorption, and thus equilibrium occurs with smaller values of S at low solution concentrations. We speculate that the discrepancies between the measured and the simulated BTC using SGA are due to the adsorption–desorption processes of the solutes not being fast enough to be considered instantaneous. If the desorption of Ca2+ was not as fast as adsorption, as has been shown elsewhere (Verburg and Baveye, 1994; Undabeytia et al., 1998), it would result in a higher concentration of Ca2+ in the adsorbed phase. This would, therefore, produce a more pronounced reduction in the concentration of this ion in the solution phase.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 6. Measured (dots) and modeled (lines) breakthrough curves for SO42– and Ca2+ from (a and b) columns E1 and (c and d) E2 (leached with CaSO4). Ion-pair adsorption was described using (a and c) the particle governed approach, PGA, and (b and d) solution governed approach, SGA (PV = pore volumes).
|
|
Clearly, more studies are required to better understand the IPA phenomenon and thus describe with greater precision the movement of solutes in soils displaying it. The extent of IPA can be considerably large, and its influence on the solute leaching is evident. Our approaches for describing IPA do allow predictions of the general patterns of the adsorption isotherms and BTCs, and this provides a starting point for further investigations.
|
CONCLUSIONS
|
|---|
We present evidence of IPA in the Egmont soil, which is dominated by variable-charge components. Both SO42– and Ca2+ adsorption increased significantly when applied simultaneously. Results from both batch and miscible displacement experiments showed features that supported the theory of cooperative adsorption for these two solutes as ion pairs. Two simple approaches were used to describe IPA, in which the adsorbed solute is partitioned into different components. These two approaches were used for modeling our experimental data.
Both the SO42– and Ca2+ adsorption isotherms were well described by a Freundlich model, plus IPA. Using the CDE model coupled with either of the two approaches for describing ion adsorption, we achieved a reasonable description of the shape of the breakthrough curves. However, the PGA model, which relates IPA to the soil's single adsorption capacity, resulted in better overall agreement.
Our approaches for describing IPA can be used to estimate the simultaneous movement of SO42– and Ca2+ in soils. More work must be done, however, to obtain a better understanding of the IPA phenomenon and for further model development. The extent of IPA should not be ignored when describing S or Ca dynamics in variable-charge soils.
|
NOTES
|
|---|
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
Received for publication June 1, 2006.
|
REFERENCES
|
|---|
- Abbaspour, K.C., C.A. Johnson, and M.Th. van Genuchten. 2004. Estimating uncertain flow and transport parameters using a sequential uncertainty fitting procedure. Vadose Zone J. 3:1340–1352.[Abstract/Free Full Text]
- Ajwa, H.A., and M.A. Tabatabai. 1995. Metal-induced sulfate adsorption by soils. 2. Effects of metal type, valence, and concentration. Soil Sci. 160:281–290.
- Ajwa, H.A., and M.A. Tabatabai. 1997. Metal-induced sulfate adsorption by soils. 3. Application of Langmuir equations. Soil Sci. 162:169–180.[CrossRef]
- Altfelder, S., T. Streck, M.A. Maraqa, and T.C. Voice. 2001. Nonequilibrium sorption of dimethylphthalate—Compatibility of batch and column techniques. Soil Sci. Soc. Am. J. 65:102–111.[Abstract/Free Full Text]
- Alva, A.K. 1993. Sorption of calcium and sulfate by sandy soils from flue-gas desulfurization gypsum and phosphogypsum. Commun. Soil Sci. Plant Anal. 24:2701–2713.[ISI]
- Beulke, S., and C.D. Brown. 2006. Impact of correlation between pesticide parameters on estimates of environmental exposure. Pest Manage. Sci. 62:603–609.[CrossRef]
- Bolan, N.S., R. Naidu, M.A.R. Khan, R.W. Tillman, and J.K. Syers. 1999. The effects of anion sorption on sorption and leaching of cadmium. Aust. J. Soil Res. 37:445–460.[CrossRef]
- Bolan, N.S., J.K. Syers, and M.E. Sumner. 1993. Calcium-induced sulfate adsorption by soils. Soil Sci. Soc. Am. J. 57:691–696.[ISI]
- Cajuste, L.J., R.J. Laird, B. Cuevas, and J. Alvarado. 1998. Phosphorus retention by tropical soils as influenced by sulfate application. Commun. Soil Sci. Plant Anal. 29:1823–1831.[ISI]
- Curtin, D., and J.K. Syers. 1990. Mechanism of sulfate adsorption by two tropical soils. J. Soil Sci. 41:295–304.[CrossRef]
- Davis, J.G., and B. Burgoa. 1995. Interactive mechanisms of anion adsorption with calcium leaching and exchange. Soil Sci. 160:256–264.
- Fahrenhorst, C., J. Botschek, A. Skowronek, and J. Ferraz. 1999. Application of gypsum and lime to increase cation adsorption of a Geric Ferralsol in the Brazilian Amazon region. J. Plant Nutr. Soil Sci. 162:41–47.[CrossRef]
- Florence, T.M., and Y.J. Farrar. 1971. Spectrophotometric determination of chloride at parts-per-billion level by mercury(II) thiocyanate method. Anal. Chim. Acta 54:373–377.[CrossRef][ISI]
- Friedel, M.J. 2005. Coupled inverse modeling of vadose zone water, heat, and solute transport: Calibration constraints, parameter nonuniqueness, and predictive uncertainty. J. Hydrol. 312:148–175.[CrossRef]
- Heng, L.K., R.W. Tillman, and R.E. White. 1999. Anion and cation leaching through large undisturbed soil cores under different flow regimes. 1. Experimental results. Aust. J. Soil Res. 37:711–726.
- Ishiguro, M., T. Makino, and Y. Hattori. 2006. Sulfate adsorption and surface precipitation on a volcanic ash soil (allophanic Andisol). J. Colloid Interface Sci. 300:504–510.[ISI][Medline]
- Johnson, C.M., and H. Nishita. 1952. Microestimation of sulfur in plant materials, soils, and irrigation waters. Anal. Chem. 24:736–742.
- Katou, H., K. Uchimura, and B.E. Clothier. 2001. An unsaturated transient flow method for determining solute adsorption by variable-charge soils. Soil Sci. Soc. Am. J. 65:283–290.[Abstract/Free Full Text]
- Kohne, J.M., B.P. Mohanty, and J. Simunek. 2006. Inverse dual-permeability modeling of preferential water flow in a soil column and implications for field-scale solute transport. Vadose Zone J. 5:59–76.[ISI]
- Magesan, G.N., I. Vogeler, B.E. Clothier, S.R. Green, and R. Lee. 2003. Solute movement through an allophanic soil. J. Environ. Qual. 32:2325–2333.[ISI][Medline]
- Maraqa, M.A. 2001. Effects of fundamental differences between batch and miscible displacement techniques on sorption distribution coefficient. Environ. Geol. 41:219–228.[CrossRef]
- Marcano-Martinez, E., and M.B. McBride. 1989. Calcium and sulfate retention by 2 Oxisols of the Brazilian cerrado. Soil Sci. Soc. Am. J. 53:63–69.[ISI]
- Mertens, J., R. Stenger, and G.F. Barkle. 2006. Multiobjective inverse modeling for soil parameter estimation and model verification. Vadose Zone J. 5:917–933.[Abstract/Free Full Text]
- Molloy, L. 1988. Soils in the New Zealand landscape. N.Z. Soc. Soil Sci., Lincoln.
- Mora, M.D.L., C. Shene, A. Violante, R. Demanet, and N.S. Bolan. 2005. The effects of organic matter and soil chemical properties on sulfate sorption in Chilean volcanic soils. p. 223–244. In P.M. Huang et al. (ed.) Soil abiotic and biotic interactions and the impact on the ecosystem and human welfare. Science Publ., Enfield, NH.
- Nakajima, M., T. Nishimura, M. Kato, and S. Suzuki. 2002. Percolation and reactions of Ca solution through acid Kunigami Mahji soil. In Trans. World Congress of Soil Science, 17th, Bangkok, Thailand. 14–21 Aug. 2002. Available at http://www.ldd.go.th/Wcss2002/papers/0790.pdf (verified 30 Dec. 2006). Land Development Dep., Bangkok.
- Parfitt, R.L., and A.D. Wilson. 1985. Estimation of allophane and halloysite in three sequences of volcanic soils, New Zealand. Catena Suppl. 7:1–8.
- Pearce, R.C., and M.E. Sumner. 1997. Apparent salt sorption reactions in an unfertilized acid subsoil. Soil Sci. Soc. Am. J. 61:765–772.[Abstract/Free Full Text]
- Pigna, M., and A. Violante. 2003. Adsorption of sulfate and phosphate on Andisols. Commun. Soil Sci. Plant Anal. 34:2099–2113.[CrossRef][ISI]
- Qafoku, N.P., and M.E. Sumner. 2001. Retention and transport of calcium nitrate in variable charge subsoils. Soil Sci. 166:297–307.[CrossRef]
- Qafoku, N.P., and M.E. Sumner. 2002. Adsorption and desorption of indifferent ions in variable charge subsoils: The possible effect of particle interactions on the counter-ion charge density. Soil Sci. Soc. Am. J. 66:1231–1239.[Abstract/Free Full Text]
- Undabeytia, T., S. Nir, G. Rytwo, E. Morillo, and C. Maqueda. 1998. Modeling adsorption–desorption processes of Cd on montmorillonite. Clays Clay Miner. 46:423–428.[Abstract]
- Verburg, K., and P. Baveye. 1994. Hysteresis in the binary exchange of cations on 2:1 clay minerals: A critical review. Clays Clay Miner. 42:207–220.[Abstract]
- Vrugt, J.A., C.G.H. Diks, H.V. Gupta, W. Bouten, and J.M. Verstraten. 2005. Improved treatment of uncertainty in hydrologic modeling: Combining the strengths of global optimization and data assimilation. Water Resour. Res. 41(1):W01017, doi:10.1029/2004WR003059.
- Wada, S. 1984. Mechanism of apparent salt absorption in ando soils. Soil Sci. Plant Nutr. 30:77–83.
TOP
ABSTRACT
INTRODUCTION
