SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Evangelou, V. P.
Right arrow Articles by Lumbanraja, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Evangelou, V. P.
Right arrow Articles by Lumbanraja, J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Evangelou, V. P.
Right arrow Articles by Lumbanraja, J.
Related Collections
Right arrow Nutrient Cycling
Right arrow Soil Chemistry
Soil Science Society of America Journal 66:445-455 (2002)
© 2002 Soil Science Society of America

DIVISION S-2—SOIL CHEMISTRY

Ammonium–Potassium–Calcium Exchange on Vermiculite and Hydroxy-aluminum Vermiculite

V. P. Evangelou* and J. Lumbanraja

Iowa State University, Dep. of Agronomy, Crop, Soil, and Environmental Science, Ames, IA 50011-1010

* Corresponding author (vevangel{at}iastate.edu)


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Theoretical Considerations
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Binary and ternary cation-exchange reactions involving NH+4, K+, and Ca2+ on vermiculite and hydroxy-Al interlayered vermiculite (HIV) were investigated. In the case of vermiculite, NH4-Ca exchange was nearly indistinguishable from the nonpreference isotherm while HIV exhibited preference for NH+4. The presence of K+ had significant suppressing impact on the apparent NH+4 selectivity in NH4-Ca exchange in HIV most likely because of retardation of Ca2+ diffusion in the interlayer. For vermiculite, the presence of K+ in the NH4-Ca system induced apparent high affinity sites for NH+4 most likely because of interlayer collapse. The overall study demonstrated that cation-exchange selectivity coefficients in vermiculite and HIV are dependent on number and type of cations present.

Abbreviations: HIV, hydroxy-Al interlayered vermiculite • HAP, hydroxy-Al polymer • {sum}Exi, sum of exchangeable cations • CEC, cation-exchange capacity • ARNH4, ammonium activity ratio • CRNH4, ammonium concentration ratio • ARK, K activity ratio • CRK, K concentration ratio • ECa, equivalent fraction of Ca on the exchange phase • ENH4, equivalent fraction of ammonium on the exchange phase • EK, equivalent fraction of K on the exchange phase • d-H2O, distilled water


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Theoretical Considerations
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THIS PARTICULAR STUDY deals with NH4-K-Ca exchange, a ternary cation-exchange system. Ternary exchange data involving NH4-K-Ca to our knowledge are limited but this exchange mode is more representative of agricultural soils than the binary exchange mode, e.g., NH4-Ca or K-Ca. Farmers often use NH4 and K salts as fertilizer sources. Even though applied NH+4 has an apparent short life span in agricultural soils (1–4 wk or more depending on rates of nitrification) (Green et al., 1994; Sims and Atkinson, 1974), the NH4-K-Ca exchange interactions control distribution of these cations between the exchange and solution phases during that period. Thus, the availability of K+ and NH+4 in the solution phase would be affected by all ions present.

Much research has been reported on soil clay minerals and soils as exchanger surfaces at both theoretical and experimental levels (Vanselow, 1932; Gapon, 1933; Argersinger et al., 1950; Sposito, 1981a, b; Sposito et al., 1983; Sposito et al., 1981; Jardine and Sparks, 1984a, HREF="#BIB27">b). These studies and many others (Beckett, 1964; Beckett and Nafady, 1967; Evangelou et al., 1994) consider that soils are two ion-exchange systems. Field soils, however, are multi-ion exchange systems (Curtin and Smillie, 1983; Adams, 1971). It is assumed that data of binary exchange reactions can be employed to predict ternary exchange reactions. In order for this assumption to be valid, one has to accept that binary exchange-selectivity coefficients are independent of exchanger-phase composition. Therefore, there is need for ternary and binary cation-exchange data representing various minerals or soils to demonstrate that binary cation-exchange data can be used to predict ternary cation-exchange data. Chu and Sposito (1981) showed at a theoretical level that one couldn't predict exchange–phase solution–phase interactions of a ternary system solely from data obtained from binary exchange. These researchers argued that experimental data obtained from systems with three ions are necessary.

Experimental data of ternary exchange involving NH4-Ba-La on montmorillonite were reported by Elprince et al. (1980) who concluded that binary-exchange data described adequately the ternary exchange system. Studies on Na-Ca-Mg exchange reactions by Fletcher et al. (1984b), Sposito et al. (1983), and Sposito and Fletcher (1985) also showed that ternary-exchange data on smectites can be predicted from binary exchange data. Feigenbaum et al. (1991) showed that in the case of K-Mg-Ca exchange on montmorillonitic soils ternary exchange data could be predicted by binary data. However, Sposito et al. (1986) and Sposito and LeVesque (1985) reported that in the case of Na-Mg-Ca exchange on illite the presence of Na+ did affect the Ca-Mg selectivity coefficient.

It appears that much of the available ternary data deal mostly with freely expanding 2:1 clay minerals, e.g., montmorillonite and bentonite, and some limited ternary data are available for illite. However, to our knowledge no data are available for ternary-exchange reactions on vermiculite or HIV, two clay minerals with some cation-size limitation effects on apparent selectivity (Kozak and Huang, 1971). The objective of this study was to evaluate NH4-Ca exchange equilibrium in the presence or absence of added K+ in vermiculite and HIV.


   Theoretical Considerations
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Theoretical Considerations
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
A binary-exchange reaction at equilibrium involving NH+4 and Ca2+ in a soil system can be written as

[1a]

However, for reasons of consistency and direct comparison with our previous work (Evangelou and Phillips, 1987, 1989) the above exchange expression would be presented as

[1b]
where Ex denotes an exchanger phase taken to have a charge of negative one (-1) and NH+4 and Ca2+ denote solution species. Note that according to the Vanselow exchange expression (Vanselow, 1932), the difference between the exchange-selectivity coefficients of the above two equations is that the equilibrium expression for [1a] is equal to the equilibrium expression [1b] raised to the second power.

Based on the above, the thermodynamic exchange equilibrium constant Keq for reaction [1b] at room temperature and 1.01 MPa (one atmosphere) pressure can be represented by:

[2]
where {alpha}Ca, {alpha}NH4 equals the activity of solution phase Ca2+ or NH+4 and {alpha}ExNH4, {alpha}ExCa equals the activity of exchange phase NH+4 or Ca2+. Solution activity {alpha}i, is defined by the equation

[3]
where ci equals the concentration of species i and {gamma}i equals the activity coefficient of species i. To define solution phase {alpha}i, its value is set to 1, hence {gamma}i = 1, when solution ionic strength (I) approaches zero. For mixed electrolyte solutions when I > 0, the single ion-activity concept introduced by Davies (1962) is employed to estimate {alpha}i which is given below in the Materials and Methods section.

The activity component of the adsorbed or solid phase is defined by employing the mole fraction concept (Mi) introduced by Vanselow (1932). According to Vanselow, for a heterovalent binary-exchange reaction such as NH4-Ca, assuming that the system obeys ideal solid-solution theory (Evangelou and Phillips, 1987 and references therein), the activity term ({alpha}Exi) is defined by

[4]
and

[5]
where MNH4 or MCa denote mole fraction of NH+4 or Ca2+, respectively. In this study the cation exchange capacity (CEC) of the soil was taken to be CEC = ExNH4 + ExCa1/2 for the binary system and CEC = ExNH4 + ExK + ExCa1/2 for the ternary system where ExCa1/2 denotes cmolc kg-1. It was also assumed that any other cation, such as exchangeable H+, is present in negligible quantities and does not interfere with NH4-Ca exchange, or H+ is tightly bound to the solid surface giving rise only to pH-dependent charge (Sposito, 1981b).

Based on the above, the Vanselow (1932) exchange selectivity coefficient expression, Kv, for reaction [1b] can be given as

[6]
where {alpha}NH4/({alpha}Ca)1/2, the inverse of ({alpha}Ca)1/2/{alpha}NH4), is known as the ARNH4. The magnitude of Kv is taken to represent relative affinity of NH+4 with respect to Ca2+ by the clay surface (Shainberg et al., 1980; Sposito and LeVesque, 1985). When Kv equals 1 at a given level of exchangeable NH+4, the exchanger at that level of NH+4 load shows no preference for either NH+4 or Ca2+. On the other hand, a Kv >1 signifies exchanger preference for NH+4 and a Kv <1 signifies exchanger preference for Ca2+.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Theoretical Considerations
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The clay sample used in this study was Connecticut Vermiculite obtained from Ward's Natural Science Establishment, Rochester, NY. The <2-µm clay fraction of the vermiculite was fractionated by grinding and gravity separation (Carstea et al., 1970; Jackson, 1975) in distilled water (d-H2O) suspension. Total elemental analysis of clays was performed using a fusion method employing 1:5 Ca-saturated clay/LiBO2 flux ratio, fusion at 1050°C, and dissolution of the fused sample by 4% HNO3 (wt./wt.) (Johnson and Maxwell, 1981). The elements Si, Al, Fe, Mg, Ca, K, and Na were analyzed with atomic absorption spectrophotometry. Based on the total elemental analysis, structural formulae of the clay samples were reconstructed according to the procedures described by Jackson (1975).

To evaluate NH4-Ca exchange behavior as affected by hydroxy-Al polymer (HAP), HIV was prepared according to the procedure described by Brydon and Kodama (1966) and Carstea et al. (1970). Briefly, the suspended vermiculite clay in d-H2O was flocculated with 0.33 M AlCl3 and equilibrated twice for 2 d each time with 150 mL of 0.33 M AlCl3. The clay was then washed with d-H2O (six times) until Cl-free before it was vacuumed-dried at room temperature. A 5-g sample of Al-saturated vermiculite was placed in a 300-mL beaker and suspended in 250 mL of 0.11 M AlCl3. While being stirred vigorously, the suspension was titrated employing a Radiotitrimeter set (TTT 80, ABU 80, PHM 82, Radiometer, Copenhagen) with 0.021 M Ca(OH)2 solution in N2 atmosphere and at room temperature to a final pH of 5.0. The samples were equilibrated for 6 mo and twice a week they were shaken for 10 min. At the end of the 6-mo period the samples were washed eight times with d-H2O until Cl-free and then air-dried under vacuum at room temperature. Formation of HIV was verified through x-ray diffraction.

Exchange reactions of NH4-Ca and NH4-K-Ca were carried out employing a batch technique. Clay samples in triplicate of 0.20 g were weighed into preweighed 50-mL test tubes. To each of these test tubes was added 20 mL of a solution composed of various NH4-Ca, and NH4-K-Ca ratios with a constant Cl concentration at 1 M. After shaking for 11 d, the test tubes were centrifuged at 670 x g for 10 min and the clear supernatants were decanted. To each tube a 20-mL solution of the corresponding NH4-Ca, and NH4-K-Ca ratio was added at 10 mM Cl concentration. The test tubes were shaken for 3 h, centrifuged at 670 x g for 10 min and the clear supernatants were again decanted. This last treatment was repeated three times; the third time, the test tubes were shaken for 24 h and the supernatants were collected for Ca, K, and NH4 determinations in the solution phase. Each tube was again weighed to quantify the entrapped solution. The exchangeable NH+4, K+, and Ca2+ were displaced twice with 20 mL of 0.2 M CsCl. This displacement was derived by trial and error for the purpose of maximizing extractable cations. The soil samples were shaken for 3 h, centrifuged, and the clear supernatants were analyzed for K and Ca by atomic absorption and for NH+4 by a Technicon Autoanalyzer (Technion Instruments, Tarrytown, NY) with the phenolhypochlorite reaction (Knopp and McKee, 1978).

The Vanselow exchange selectivity coefficient (Kv) was calculated using Eq. [4], [5], and [6]. Single-ion activity ({alpha}) for NH+4, K+, and Ca2+ employed in Eq. [6] were estimated by the previously described equation:

[7]

The values for {gamma}i,j were estimated by employing the effective ionic strength (I) equation:

[8]
where Z denotes charge of the ionic species b, (b denotes all ionic species in solution), and the Davies equation (Davies, 1962)

[9]

Equations [7], [8] and [9] were solved by an ion-association computer model (Evangelou, 1986).


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Theoretical Considerations
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
The data in Fig. 1 and 2 show x-ray diffraction patterns of the two clay mineral samples saturated with Mg2+, Mg-glycerol or with K+ with and without heating. These diffractograms clearly demonstrate the d-spacing behavior of vermiculite. Interlayer expansion (~1.4 nm) occurred when clays were treated with Mg2+ with and without glycerol and collapsed to ~1.0 nm when K-saturated with and without heating. In the case of HIV, the peaks remained ~1.4 nm by Mg2+ or K+ saturation. Heating of K-saturated vermiculite clay sample did not produce any unique features in the diffractograms except in the case of HIV. In this clay the x-ray diffraction peak near 1.4 nm appeared collapsed and shifted to the left when heated to 100°C (Fig. 2). The instability of the 1.4-nm peak when the K-saturated samples were heated from 25 to 100, 300, and 550°C indicated incomplete filling of the vermiculite interlayer (Barnhisel and Bertsch, 1989).



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 1. X-ray diffractograms of vermiculite clay.

 


View larger version (35K):
[in this window]
[in a new window]
  		Fig. 2. X-ray diffractograms of hydroxy-Al interlayered vermiculite (HIV).

 
Based on complete elemental analysis and reconstitution (Table 1), the vermiculite used in this study was trioctahedral. Cation-extractability behavior of the two clay samples is reported in Table 2. These data show that estimated charge based on structural reconstitution was lower on HIV than that of vermiculite. While lacking absolute agreement between experimental and reconstituted charge, we observed there was good agreement in the trends between experimental and reconstituted charge. The poorest agreement was that between total charge determined experimentally by loading the clays with Ca2+ and the estimated charge based on reconstitution of HIV (6.51 vs. 39.27 cmolc kg-1). The data in Table 2 also show that vermiculite clay, upon introducing HAP, the total charge was reduced by half (79.73 vs. 39.27 cmolc kg-1). Furthermore, fixed K+ decreased by approximately half (65.66 vs. 31.16 cmolc kg-1) while fixed ammonium was dramatically suppressed (57.49 vs. 2.88 cmolc kg-1). As expected, these data support the conclusion that certain surface mechanisms (i.e., steric effect or specific adsorption) were responsible for changes in mineral's ion exchange capacity (Rich and Black, 1964).


View this table:
[in this window]
[in a new window]
  		Table 1. Formula of the two clay samples.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Exchangeable and non-exchangeable cations of the clay fraction after saturation with K, NH4, or Ca and formula cation-exchange capacity (CEC).

 
Tables 3 through 6 contain exchange data for the binary NH4-Ca, and ternary NH4-K-Ca exchange systems for both the vermiculite and HIV clays. At first glance, one can clearly see from these data that {sum}Exi represented only a small fraction of the vermiculite and HIV clays' total charge (Table 2), suggesting that the exchange reactions in the above systems occurred mainly on the clays' external clay surfaces, and possibly some of the interlayer clay surfaces as well (Shainberg et al., 1980). For the vermiculite clay, two important points can be made for the binary and ternary cation-exchange systems: First, we observed a shift in pH (Table 3) as the predominant cation was switched from Ca2+ (pH 7.7) to NH+4 (pH 6.9). Secondly, changes in selectivity over different NH+4/Ca2+ ratios were marked by changes in exchange capacity ({sum}Exi).


View this table:
[in this window]
[in a new window]
  		Table 3. Experimental data on NH4-Ca exchange at 297 K on vermiculite clay.

 

View this table:
[in this window]
[in a new window]
  		Table 6. Experimental data on NH4-K-Ca exchange at 297 K on hydroxy-Al interlayered vermiculite (HIV).

 

View this table:
[in this window]
[in a new window]
  		Table 4. Experimental data on NH4-K-Ca exchange at 297 K on vermiculite clay.

 

View this table:
[in this window]
[in a new window]
  		Table 5. Experimental data on NH4-Ca exchange at 297 K on hydroxy-Al interlayered vermiculite (HIV).

 
A close examination of pH and {sum}Exi data across the entire isotherm (Table 3) indicates that pH was not significantly influential on magnitude of {sum}Exi. First, the pH variation across the entire isotherm would be considered relatively small, ±0.4 (Table 3). Published data (Fletcher et al., 1984a) did not show large pH dependent charge in 2:1 expanding clay minerals. Secondly, while pH remained relatively constant for a portion of the isotherm (Table 3), {sum}Exi varied from 62.7 to 36.4 cmolckg-1. This large variation in {sum}Exi can only be because of variation in NH+4/Ca ratio, thus a cation blocking effect or layer collapse was induced. In the case of the ternary system (Table 4), variation in {sum}Exi across the entire isotherm was ~5 cmolckg-1. However, this variation came about when pH varied from 7.6 to 7.5 whereas solution NH+4 concentration doubled. Therefore, this difference in {sum}Exi cannot be because of a 0.1 unit change in pH but most likely because of change in the composition of the exchange phase. For these two cation-ratio treatments with pH 7.6 and 7.5 exchangeable Ca decreased from 23.6 cmolckg-1 to 13.1 cmolckg-1, exchangeable NH+4 increased from 0.9 to 3.0 cmolckg-1, and exchangeable K+ increased from 15.0 to 19.2 cmolckg-1. Thus, a blocking effect or partial interlayer collapse was the most likely cause for this decrease in {sum}Exi.

Ideally, pH and {sum}Exi or CEC should be constant across the entire isotherm to generate a cation-exchange isotherm or estimate cation-selectivity coefficients. However, pH or CEC constancy in soil clay minerals is rarely met. For example, in Na-Ca, Na-Mg, and Ca-Mg exchange studies using a montmorillonitic soil (Fletcher et al., 1984a,b), pH across the entire isotherm varied from 6.7 to 7.1 in one case and in another case from 4.4 to 5.0. Additionally, CEC in one case varied from 51 to 60 cmolckg-1 and in another case from 64 to 75 cmolckg-1. Furthermore, the above soil's CEC in the binary mode was 53 cmolckg-1 whereas in the ternary mode was 70 cmolckg-1. The latter CEC variation was not attributed to pH because the binary and ternary systems were studied under approximately similar pH. Furthermore, when the exchange data, representing the above soil under three different pH values (pH 4.7, 5.8, and 6.9), were used to calculate CEC ({sum}Exi) the latter varied very little with respect to the three pH values tested. In essence, the so-called pH dependent charge of a 2:1 shrink/swell clay in the ternary exchange mode, under certain exchangeable cation composition, was basically negligible.

In the present study, the highest {sum}Exi values (Table 3) were associated with the highest Ca2+ concentrations added, while the lowest {sum}Exi values corresponded with the highest NH+4 concentrations added. Accordingly, these trends were consistent with the order of exchangeable cations, in terms of magnitude of experimental charge (Table 2), as NH+4 < K+ < Ca2+. Also, note that this trend was consistent with the order of nonexchangeable cations in vermiculite, Ca2+ < NH+4 < K+ (Table 2). Based on this information, we propose that vermiculite controlled ion selectivity by acting as a molecular sieve, discriminating among ions based on cation size, charge and hydration energy (Black and Rich, 1964). System pH may also have some role in determining NH+4 selectivity, as we note the decrease in {sum}Exi with decrease in pH (Sposito, 1981b) but in the present study this role may have been small. Based on our data there was no clear consistency in the relationship between pH and CEC. For example, the first four NH4/Ca treatments in Table 3 exhibited similar pH (pH 7.7) but {sum}Exi varied from ~62.7 to 41.1 cmolckg-1. Furthermore, for pH treatments 7.1 and 6.9 {sum}Exi varied from 26.8 to 25.3 cmolckg-1. This suggested factors other than pH, e.g., steric factors may lave played a role. Note that similar inconsistencies in {sum}Exi were present in a montmorillonitic soil at pH 6 (Fletcher et al., 1984a, b).

The exchange-isotherm data for both the binary (NH4-Ca) and ternary (NH4-K-Ca) exchange systems for the HIV are presented in Tables 5 and 6. In general, {sum}Exi values were lower in HIV than in vermiculite because of Al-hydroxylation which may serve to shield the interlayer surface. Also solution pH was the lowest in the HIV systems, most likely because of hydrolysis of free Al3+. Only the data in Table 5 (NH4-Ca) showed a significant {sum}Exi gradient (~6.0 cmolckg-1 for the Ca-rich isotherm vs. ~11.7 cmolckg-1 for the NH+4-rich isotherm) in the opposite direction of the data in Table 3. This suggested (i) a potential Ca retardation effect (which would reduce {sum}Exi) because HAP present in the interlayer (Kozak and Huang, 1971) and (ii) perhaps a pH-dependent effect (which would increase {sum}Exi) because of a pH increase (Sposito, 1981b). One could interpret the increase in {sum}Exi at higher NH+4 concentrations as a phenomenon of Ca2+ exclusion by the solid phase. It is possible that certain interlayer-exchange sites are accessible to NH+4 and not to Ca2+. Unlike the vermiculite system, HIV system is less likely to trap NH+4 via interlayer collapse at higher ENH4, thus making it accessible to Cs+ exchange. In the case of ternary exchange (Table 6), changes in pH and {sum}Exi were negligible. Here, the relatively small pH fluctuation of the HIV system minimized the potential influence of pH-dependent external sites on exchange, thus, helping elucidate the role of wedge surfaces on binary and ternary exchange.

Variations in CEC ({sum}Exi) and pH do not forbid us from estimating Kv as long as such variations are substantiated by appropriate measurements (Fletcher et al., 1984ab; Zhang and Sparks, 1996). The estimated Kv values represent individual data sets for a particular monovalent cation load and signify the behavior of the clay at that load. In our particular study the data showed that {sum}Exi within a clay mineral was not necessarily related to pH and varied greatly with changes in cation load. The findings suggested that it is difficult to model exchange behavior of vermiculitic clay minerals because CEC and Kv vary as a function of monovalent cation load.

Binary exchange isotherms for both clay samples are presented in Fig. 3 . For reference purposes the nonpreference isotherm at 10 mM Cl was included. This nonpreference isotherm was generated by the equation (Sposito, 1981b)

[10]
where ENH4 equals the equivalent fraction of NH+4 on exchanger phase, F = {gamma}2NH4/{gamma}Ca; where {gamma} denotes single-ion activity coefficient, TN represents the total cation normality, and E'NH4 represents equivalent fraction of NH+4 in solution phase. It can be seen from the data in Fig. 3 that vermiculite showed no great preference for either NH+4 or Ca2+ while HIV showed NH+4 preference over Ca2+. It appeared that HAP shielded internal surfaces and their role in cation exchange was limited. Thus, external clay surfaces were most likely responsible for NH+4 preference behavior. It is well known that external clay surface sites exhibit low charge density (Shainberg et al., 1980). This low surface charge density, because of entropy effects, formed relatively stronger surface complexes with monovalent cations than divalent cations (Shainberg et al., 1980).



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 3. Exchange isotherms for NH4-Ca exchange on vermiculite and hydroxy-Al interlayered vermiculite (HIV).

 
The data in Fig. 4 show dependence of Kv on exchangeable loads of monovalent cations. The HIV clay exhibited at least two major classes of exchange sites. At low exchangeable monovalent cation loads the clay system exhibited extremely high affinity sites (high Kv) for the monovalent cation. These sites were most likely edge sites with highly steric wedge influences or retardation of Ca2+ diffusion in the clay interlayer (Kozak and Huang, 1971). At increasing loads of exchangeable monovalent cations, Kv approached 1. On the other hand, vermiculite maintained an approximately constant Kv for NH4-Ca exchange throughout the exchange isotherm. This suggested nearly constant surface complexation uniformity throughout the entire exchange isotherm.



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 4. Vanselow selectivity coefficients of NH4-Ca exchange on vermiculite and hydroxy-Al interlayered vermiculite (HIV).

 
The influence of a third ion (K at 6.0 mM) on NH4-Ca exchange on both clay samples are presented in Fig. 5 and 6 . Ternary Kv values were calculated by including exchangeable K+ in the mole fractions (Sposito and LeVesque, 1985; Sposito et al., 1983). For the HIV mineral, adding K+ dramatically suppressed KV (Fig. 5). In contrast, adding K+ to vermiculite enhanced NH+4 selectivity while shifting the maximum KV to higher ENH4 values (Fig. 6). The mechanism for this apparent reversal in the vermiculite's NH+4 selectivity by the addition of K+ is not clear at this time. However, one can hypothesize that in the presence of K+ the interlayer collapsed (pinching effect) thus preventing Ca2+ from exchanging with NH+4 (Chappell and Evangelou, 2000 a,b). These results imply that this mechanism is only operable in the absence of HAP. Note also that the decrease in {sum}Exi at high NH+4 loading may also suggest that the interlayers of vermiculite are collapsing and thus fixing NH+4 and precluding exchange by Cs+. This implies that the exchange isotherms maybe biased in that highly selective sites for NH+4 are unaccounted for at high mole fractions of NH+4. Accordingly, calculated Kv values will be low at higher ENH4 making it appear that the solid phase has little preference for NH+4 vs. Ca2+.



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 5. Vanselow selectivity coefficients for NH4-Ca exchange in the presence or absence of K+ (6.0 mM KCl in solution) on hydroxy-Al interlayered vermiculite (HIV).

 


View larger version (17K):
[in this window]
[in a new window]
  		Fig. 6. Vanselow selectivity coefficients for NH+4-Ca exchange in the presence or absence of K+ (6.0 mM KCl in solution) on vermiculite.

 
Direct comparison of the HIV-exchange data to that of vermiculite maybe problematic since the two systems differ in pH. In other words, to a degree the pH difference alone between the two clay mineral systems could explain the difference in exchange behavior. High pH is commonly associated with high surface electrical potential and divalent cations are preferred by the clay surface (Pratt et al., 1962). However, a high clay-surface electrical potential can also be attributed to overlapping clay-interlayer double layers independently of pH (Shainberg et al., 1980). On the other hand, low pH is commonly associated with low clay-surface electrical potential and monovalent cations are preferred. However, clay dispersion may also induce a similar behavior independently of pH (McBride, 1980). One could still argue that the cation-exchange data in Fig. 3 are consistent with the expected cation exchange behavior because of influence of pH. However, the sharp increase in Kv at low K+ loads in HIV (Fig. 4 and 5) cannot be attributed to pH. This increase in Kv (high preference for K+) is commonly attributed to an interlayer monovalent cation-sieving effect (Beckett and Nafady, 1967).

In the case of agricultural soils, the most useful portion of a cation-exchange isotherm is that at low monovalent cation loads, e.g., low EK or ENH4. For this reason, an analysis of cation-exchange isotherm behavior at low monovalent cation fractional loads was carried out below. For a binary system, e.g., NH4-Ca exchange, the slope of the isotherm (CRNH4 vs. ENH4) as CRNH4 approaches zero can be used to estimate magnitude of Kv at the low fractional loads. This relationship is given by (Lumbanraja and Evangelou, 1990)

[11]

In the case of a ternary exchange system, e.g., NH4-K-Ca, the solution of dENH4/dCRNH4 at a constant CRK as CRNH4 approaches zero is given by (Lumbanraja and Evangelou, 1990)

[12]
where Kv1 is the Vanselow exchange selectivity coefficient for NH4-Ca exchange with NH+4 as the displacing cation, and Kv2 is the Vanselow exchange selectivity coefficient for K-Ca exchange with K+ as the displacing cation. Note that in Eq. [11] and [12] CRNH4 was employed instead of ARNH4 because for the ionic strength tested CRNH4 and ARNH4 or CRK and ARK are nearly indistinguishable (Evangelou et al., 1986).

The NH4-Ca exchange isotherms for vermiculite and HIV are shown in Fig. 7 and 8 , respectively. The higher slopes of the isotherms representing the ternary exchange modes as CRNH4 approached zero suggested that the NH4-Ca Kv of the ternary system was greater than the NH4-Ca Kv of the binary system. Note that ENH4 values for the ternary systems were calculated using all three exchangeable cations, e.g., ExK, ExNH4, and ExCa1/2. Assuming that the NH4-Ca exchange Kv of the binary and ternary exchange systems were similar, the slope of the ternary plot (as CRNH4 approached zero) in Fig. 7 would be smaller than the slope of the binary plot. This is because (Kv1/2) > (Kv1/[4 + (Kv2CRK)2]1/2). On the other hand, if the slope of the ternary plot was slightly greater than the slope of the binary plot as CRNH4 approached zero, then the ternary Kv of NH+4-Ca exchange would be greater than the binary Kv of the same cation-exchange reaction (see also Fig. 6).



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 7. Relationship between ammonium concentration ratio (CRNH4) and equivalent fraction of ammonium on the exchange phase (ENH4) on the exchange phase of vermiculite in the presence and absence of K (ternary exchange equivalent fractions were calculated utilizing cation-exchange capacity (CEC) as the sum of all three cations, NH+4, K+, and Ca2+, on the exchange phase).

 


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 8. Relationship between ammonium concentration ratio (CRNH4) and ammonium on the exchange phase (ENH4) of hydroxy-Al interlayered vermiculite (HIV) in the presence and absence of K (ternary exchange equivalent fractions were calculated utilizing cation-exchange capacity (CEC) as the sum of all three cations, NH+4, K+, and Ca2+, on the exchange phase).

 
In the case of HIV, the slopes of the binary- and ternary-exchange modes also differ, as CRNH4 approached zero (Fig. 8). The exchange isotherm representing the binary mode reveals a slightly greater slope than the exchange isotherm representing the ternary mode. This implied that (Kv/2) > Kv/[4 + (Kv2CRK)2]1/2 as CRNH4 approached zero. Furthermore, the Kv representing the binary exchange mode may be equal to or greater than the ternary Kv as CRNH4 approached zero. Support for this conclusion comes from Fig. 5 where the presence of K+ significantly decreased the Kv of the NH4-Ca exchange at low ENH4 values.

Theoretically, Kv values of ternary exchange systems calculated on the basis of only two exchangeable cations (see Eq. [4], [5], and [6]) should be smaller than the actual binary Kv values (Evangelou and Phillips, 1987). Figures 9 and 10 were plotted based on this reasoning. Note that in Fig. 9 the slope of the ternary mode, with ExK omitted, is greater than the slope of the actual binary mode. However, in the case of HIV the two Kv values are indistinguishable (Fig. 10). Some justification for treating the HIV ternary system as binary comes from the fact that exchangeable K+ and solution K+ remained nearly constant throughout the entire exchange isotherm (Table 6). This is not so for the vermiculite (Table 4) which appeared to behave as a ternary exchange system. The present study's solution K+ of ~6 mM (Tables 4 and 6) would be much higher than what one would expect to find in actual soils. However, our previous data (Lumbanraja and Evangelou, 1990) showed that at much lower soil solution K+ concentration a similar behavior with respect to magnitude of Kv in NH+4-Ca exchange in a vermiculitic soil was observed.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 9. Relationship between ammonium concentration ratio (CRNH4) and ammonium on the exchange phase (ENH4) of vermiculite in the presence and absence of K (ternary exchange equivalent fractions were calculated utilizing cation-exchange capacity [CEC] as the sum of the two cations [NH+4 and Ca2+] on the exchange phase).

 


View larger version (22K):
[in this window]
[in a new window]
  		Fig. 10. Relationship between ammonium concentration ratio (CRNH4) and ammonium on the exchange phase (ENH4) of hydroxy-Al interlayered vermiculite (HIV) in the presence and absence of K (ternary exchange equivalent fractions were calculated utilizing cation-exchange capacity [CEC] as the sum of the two cations [NH+4 and Ca2+] on the exchange phase).

 

   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Theoretical Considerations
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Formation of HAP in vermiculite increased the mineral's affinity for NH+4 while addition of K+ decreased the vermiculite's affinity for NH+4. On the other hand, in the absence of HAP the presence of K+ induced high affinity for NH+4 at increasing load of NH+4. However, upon continuing to increase NH+4 load, NH+4 affinity decreased. The findings suggest that in the case of vermiculite, binary exchange data alone may not be able to predict ternary exchange data. However, in the case of HIV, binary data maybe able to predict ternary data as long as the third cation's distribution between exchange and solution phases remains constant across the isotherm.

The findings appear to have important implications in the management of fertilizers in soil systems with vermiculite-like clay minerals with and without HAP. It appears that the clay surface exhibits high specificity for K+ at low-K fractional loads which cause the vermiculite interlayer to collapse and thus a large proportion of interlayer NH+4 most likely becomes fixed. The available-exchange sites appear to have relatively lower affinity for NH+4. Therefore, the data indicate that availability of applied ammonium in vermiculitic soils would depend in the presence or absence of applied K. Applied K would have a tendency to increase NH+4 fixation but any remaining exchangeable NH+4 would be weakly held by the clay surface.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Theoretical Considerations
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Contribution from the Department of Agronomy, Iowa State University, Ames, IA 50011-1010. The investigation reported in this paper (No. IOWA03321) is in connection with a project of the Iowa State Agric. Exp. Stn. and is published with the approval of the director.

Received for publication June 6, 2001.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Theoretical Considerations
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
T. M. DeSutter, G. M. Pierzynski, and L. R. Baker
Flow-Through and Batch Methods for Determining Calcium-Magnesium and Magnesium-Calcium Selectivity
Soil Sci. Soc. Am. J., February 27, 2006; 70(2): 550 - 554.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
T. M. DeSutter and G. M. Pierzynski
Evaluation of Soils for Use as Liner Materials: A Soil Chemistry Approach
J. Environ. Qual., April 20, 2005; 34(3): 951 - 962.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
J. J. Wang, D. L. Harrell, and P. F. Bell
Potassium Buffering Characteristics of Three Soils Low in Exchangeable Potassium
Soil Sci. Soc. Am. J., March 1, 2004; 68(2): 654 - 661.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow