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 (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ge, Y.
Right arrow Articles by Hendershot, W.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Ge, Y.
Right arrow Articles by Hendershot, W.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Ge, Y.
Right arrow Articles by Hendershot, W.
Related Collections
Right arrow Soil Chemistry
Published in Soil Sci. Soc. Am. J. 68:82-88 (2004).
© 2004 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA

DIVISION S-2—SOIL CHEMISTRY

Evaluation of Soil Surface Charge Using the Back-Titration Technique

Ying Ge and William Hendershot*

Dep. of Natural Resource Sciences, Macdonald Campus of McGill Univ., 21,111 Lakeshore Road, Ste-Anne-de-Bellevue, QC, Canada H9X 3V9

* Corresponding author (William.Hendershot{at}mcgill.ca).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Variable surface charge (Qv) is one of the most important soil properties controlling ion adsorption on the soil solid phase. In this study, the back-titration technique was used to determine the Qv of soils with a wide range of properties. The procedure defines the Qv as the OH– consumption by surface reactions corrected for dissolution of the solid phase and other solution reactions (e.g., metal hydrolysis). The Qv, varying from 0 to 80 cmolc kg–1, was dependent on the pH of the soil suspension and the amount of soil organic matter. We used the non-ideal competitive adsorption (NICA)–Donnan model to simulate the surface charge, assuming a bimodal distribution of H+ affinity on the soil solid phase. With the charge data and Microsoft Excel, the NICA-Donnan model parameters were optimized. The model provided an excellent fit to the experimental data. When the pH was below 8, the surface charge was dominantly distributed to the Type 1 sites; the Type 2 sites started to contribute to the total surface charge at pH > 8. Multiple linear regressions showed that the charge maxima (Qmax) of the two sites were related to soil cation-exchange capacity (CEC) and organic C (Org. C); these significant statistical relationships may be used to estimate the surface charge of soils using values of commonly measured soil parameters.

Abbreviations: CEC, cation-exchange capacity • EGME, ethylene glycol monoetheyl ether • HS, humic substances • K, median affinity constant • m, intrinsic site heterogeneity • NICA, Non-Ideal Competitive Adsorption • Org. C, organic carbon • Qmax, maximum charge • Qp, permanent charge • Qv, variable surface charge • SSA, specific surface area


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE SURFACE OF THE SOIL SOLID PHASE consists of a variety of components such as clay minerals, oxides, and organic matter. The reactive surfaces of these materials provide sites with pH-dependent charge (Qv) properties and can exist as colloids or coatings and thus represent a significant fraction of surfaces available for soil–solution interactions (Sposito, 1998). The reactions taking place on the surface often involve proton (H+) association and dissociation, making the surfaces electrically charged (Evans, 1989). The surface charge has a marked influence on partitioning of cations and anions between the soil solid phase and the soil solution. Therefore, it is one of the most important soil properties controlling the solubility and bioavailability of trace metals (Bolan et al., 1999).

Potentiometric titration methods have been widely used to assess the surface charge of pure minerals, sediments, and soils. For example, Duquette and Hendershot (1993a)(b) used a back-titration technique to measure the surface charge of soils from temperate regions. The charge was defined as the OH– consumption by the soil surface. This method consists of sample and reference titrations, and provides a fast and reliable means to evaluate the amount of variable charge and its distribution as a function of soil pH. Duquette and Hendershot (1993b) interpreted the charge data using a four-site model, in which H+ adsorption at each site could be described using a Langmuir equation. The multisite model assumes that the surfaces of soil samples are heterogeneous, that is, there are several classes of sites involved in the H+/OH– adsorption processes. In addition, the charge data may be used to calibrate the surface acid-base parameters such as the affinity constants and site concentrations in surface complexation models (Dzombak and Morel, 1990).

Recently, a model to simulate cation binding, namely the NICA–Donnan model, has been developed and is used frequently in studies of humic acids (e.g., Kinniburgh et al., 1996; Robertson and Leckie, 1999; Milne et al., 2001). It assumes that two classes of sites (carboxylic- and phenolic-type) contribute to the proton complexation on the organic molecule. The specific-binding part of the model, NICA, implicitly assumes a continuous distribution of site affinities. Each site is characterized by three parameters: the median affinity constant (K), the intrinsic heterogeneity of the humic material (m), and the maximum site concentration (Qmax) (Milne et al., 2001). The first two parameters define the strength and variability of binding at each site. Qmax represents the total surface charge of that site as a result of H+ binding. The electrostatic part of the binding reaction is considered by the Donnan model, in which humic material is assumed to possess an electrically neutral gel phase (Benedetti et al., 1996). The concentrations of ions inside the Donnan phase are related to those in the bulk solution by the Boltzmann factor (Kinniburgh et al., 1996). Usually the pH inside the Donnan phase is one to two units lower than in the bulk solution (Milne et al., 2001).

In the boreal temperate regions, organic matter usually accumulates in the surface and subsurface layers of the soil profile as either bulk materials or coatings on inorganic particulate matter (McKeague et al. 1986). It is a major reactive component in such soils providing much of the ability of the soil to retain metal cations (Temminghoff et al., 1997). In the present study, we hypothesize that the materials making up the soil particle surfaces possess proton-binding properties similar to humic substances (HS) and therefore the NICA-Donnan model may be used to predict the variable surface charge generated by H+/OH– adsorption. Although we are assuming that the reactive surfaces of soil particles are dominated by organic matter, we cannot in fact distinguish between organic and inorganic surfaces because the pKa values of many mineral functional groups [i.e., Fe(OH)3, Al(OH)3, Si(OH)4] lie in the same range as those of organic matter.

Soil samples in this study were collected around metal smelters and varied in surface charge and soil characteristics. In these soils, clay minerals, organic matter, and oxides are the major components capable of binding protons and metal ions. Therefore, it may be possible to link the NICA–Donnan model parameters (Qmax, log K, and m) with soil properties using a statistical approach. The objectives of this study are: (i) to assess the surface charge of soils with a wide range of properties; (ii) to evaluate the ability of the NICA–Donnan model to describe the pH-dependent surface charge; and (iii) to establish the relationships between surface charge parameters and soil properties.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Sample Collection
In 1999 and 2000, soil sampling was performed at different locations around the metal smelters in Rouyn-Noranda and Valleyfield, QC and Sudbury, ON. Thirty-four soils were collected from organic (F and H horizons) and mineral layers (Ah, Bh, Bf, and Bm horizons) so that we have a wide range of soil properties.

Characterization of Soil Samples
Soil pH was measured with a soil/water ratio of 1:2 for mineral soils and 1:10 for organic soils (Hendershot et al., 1993a). A 0.1 M BaCl2 extraction was used to analyze CEC and exchangeable cations (Hendershot et al., 1993b). Particle-size analysis was performed with the procedure in Sheldrick and Wang (1993). Organic C was measured by the wet oxidation method (Tiessen and Moir, 1993). Iron and Al oxides were quantified using the acid ammonium oxalate extraction (Ross and Wang, 1993). Specific surface area (SSA) was determined using the ethylene glycol monoetheyl ether (EGME) adsorption technique (Heilman et al., 1965).

Back-Titration
We adapted the back-titration method of Duquette and Hendershot (1993b) to determine the surface charge. The soil (<2-mm fraction) was weighed (0.25 g for organic and 0.5 g for mineral soil) into a 100-mL beaker and 25 mL of 0.01 M Ca(NO3)2 solution was added; the suspension was then stirred until the pH stabilized. We first titrated the suspension with standardized 0.1 M HNO3 until a pH of 3 was reached. After the equilibration of the suspension for 2 min, the back-titration procedure was performed with standardized 0.005 M Ca(OH)2 at a rate of 0.5 mL min–1 to pH 10. The volumes of acid and base and pH were recorded. Based on previous studies of similar soils in eastern Canada, we assumed that the soils in this study have a very low anion exchange capacity and are mostly negatively charged (Duquette and Hendershot, 1993a). Hence the surface reactions involve only neutral ({equiv}SiOH) or negatively charged surface functional groups ({equiv}SiO–). In the sample back-titration, the following reactions are assumed to take place: (1) surface reactions: {equiv}SiOH + OH– = {equiv}SiO– + H2O, (2) metal hydrolysis: x Mz+ + y OH– = Mx(OH)y(xz – y)+, and (3) any unknown reactions in the solution [e.g., precipitation of Al(OH)3]. Since we are mainly interested in Reaction (1), a reference titration is needed to correct for the H+/OH– consumption from Reactions (2) and (3).

The reference back-titration was performed on a supernatant separated from the suspension (pH 3) by centrifugation (2000 x g [3000 rpm] for 20 min) and filtration (0.45-µm Millipore polycarbonate filter (Millipore, Billerica, MA). This procedure accounted for the metal hydrolysis and other unknown reactions in the supernatant. Therefore, the pH-dependent charge (Qv) due to the OH– consumption may be calculated by subtracting the OH– consumption of the reference from the sample back-titration at the same pH (Duquette and Hendershot, 1993a).

[1]
where Qs = OH– consumed by the sample, cmolc kg–1; Qr = OH– consumed by the reference solution, cmolc kg–1; cbase = concentration of the base used to back-titrate, mol L–1; Vs = volume of the base added during the sample titration, mL; Vr = volume of the base added during the reference titration, mL; Vo = volume of the supporting electrolyte and added acid, mL; Vsn = volume of the supernatant, mL; Wsoil = weight of the soil, g; a = unit conversion factor, 100 (1000 g kg–1 x 0.001 L mL–1 x 100 cmolc molc–1).

NICA–Donnan Model of Soil Surface Charge
The details of NICA–Donnan model have been presented elsewhere (Kinniburgh et al., 1996). Briefly, cation complexation is assumed to occur through specific binding between the cations and negatively charged surface functional groups. The distribution of binding affinities is composed of two parts, which are considered to be due to carboxylic- and phenolic-type groups (Kinniburgh et al., 1996). The total amount of bound H+ (QH, in mol kg–1) is given by the following equation (Milne et al., 2001):

[2]
where Qmax, K, and m stand for the maximum concentration, median affinity constant, and intrinsic heterogeneity of the site, respectively. The [HD] is the concentration of H+ close to the surface of the humic particle (i.e., within the Donnan phase). In this paper, the Suffixes 1 and 2 denote Type 1 and Type 2 sites, respectively.

We evaluated the effectiveness of NICA–Donnan model to fit the surface charge data. In the NICA–Donnan model, the reaction taking place on surface functional groups is described as: {equiv}SiO– + H+ = {equiv}SiOH, in which Qmax i = {equiv}SiOH + {equiv}SiO– and QVi = {equiv}SiO–.

Combined with Eq. [2], variable surface charge due to the H+ binding at two sites can be expressed with the following Langmuir-Freundlich adsorption equation:

[3]
where Q0 is the surface charge at the reference pH (pH 3). This parameter can be fitted by the relationship of Qv1 + Qv2 – Q0 = 0 because the consumption of OH– is zero at pH 3.

The charge curves generally demonstrate two inflection points (Fig. 1 and 2) , suggesting that a bimodal affinity distribution may be identified for the soil surfaces. The initial estimates of Qmax and log K for the two sites were taken from the surface charge curve. Furthermore, the model parameters need to be corrected for the electrostatic effects associated with the ionic strength (I) in the solutions. The Donnan volume, VD, was calculated by the empirical relationship: log VD = b x (1 – log I) – 1. We assumed the surface properties of soils are similar to those of humic acid. Thus a value of 0.49 was chosen for b (Milne et al., 2001). The concentration ratio between the Donnan phase and bulk solution is defined by the Boltzmann factor (Kinniburgh et al., 1996), which can be solved using the Eq. [4] through [10] specified in Kinniburgh et al. (1996) and the Goal Seek function in Microsoft Excel (Microsoft Corp., Redmond, WA). We used the coefficient of determination (R2) as the criteria of data fitting. The model parameters (Qmax1, Qmax2, log K1, log K2, m1, and m2) were tested in Excel to find the best fit, where the maximum value of R2 was reached between a measured and a predicted variable surface charge.



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 1. Variable surface charge of three organic soils from Sudbury (SUD 12), Rouyn-Noranda (RN 7), and Valleyfield (VAL 4) ({square}: measured by the back-titration; solid line: predicted by the non-ideal competitive adsorption [NICA]–Donnan model).

 


View larger version (16K):
[in this window]
[in a new window]
  		Fig. 2. Variable Surface charge of three mineral soils from Sudbury (SUD 1), Rouyn-Noranda (RN 1), and Valleyfield (VAL 1) ({square}: measured by the back-titration; solid line: predicted by the non-ideal competitive adsorption [NICA]–Donnan model).

 
The distribution of soil variable surface charge within two types of sites was calculated as a function of pH. To establish relationships between the parameters and soil properties, the values of Qmax, log K, and m were related to CEC, clay, SSA, Org. C, and amorphous Al and Fe oxides (oxalate extraction). Statistical analyses were made using SYSTAT 8.0 (Wilkinson, 1998).


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
The Properties of Soil Samples
The samples were collected from forested, residential, and industrial areas. The soils covered a wide range of properties with pH varying from 3.5 to 7.5, clay from 3.3 to 42%, Org. C from 0.86 to 44%, Al and Fe oxides from 0.15 to 10%, surface area from 8.1 to 299 m2 g–1 (Table 1). High surface area was often observed when a soil contained a high amount of either organic matter or clay minerals. The oxalate-extractable Al and Fe were higher in sandy soils, indicating the migration of Al and Fe in this soil type was more pronounced than in clayey soils.


View this table:
[in this window]
[in a new window]
  		Table 1. Selected chemical properties of the soils.

 
Soil CEC varied from 0.44 to 60.4 cmolc kg–1. Calcium was the most abundant exchangeable cation (up to 88%; Table 2). Most of the soils in this study have high base saturation except for three soils from Sudbury and one from Rouyn-Noranda that had a very high percentage of BaCl2–extractable Al, probably because of their low pH and high content of Al oxide.


View this table:
[in this window]
[in a new window]
  		Table 2. Cation-exchange capacity (CEC) and exchangeable cations of the soils.

 
Surface Charge Data
The back-titration procedure records the OH– consumed through proton dissociation from the soil surface sites. The amount of variable surface charge was highly dependent on the Org. C content of samples, varying from 10 to 60 cmolc kg–1 for mineral soils and from 10 to 80 cmolc kg–1 for organic soils (Fig. 1 and 2). The charge was usually much higher for the organic soils than for the mineral soils (Fig. 1 and 2). Only six soils (two from each site) are presented here because of the similar shape of the surface charge curves for all soils. Our charge data for mineral soils were within the range reported in Duquette and Hendershot (1993b) (10–80 cmolc kg–1). The charge curve for the organic soil (VAL 4) was also similar to that of a humic acid titrated at an ionic strength of 0.01 M by Robertson and Leckie (1999).

The contribution of each site to the Qv was found to be similar for both organic and mineral soils. When the pH of soil suspension was lower than 8, surface charges of all soils were predominantly distributed to the first site, as the Qv1 constituted more than 70% of total charge. More surface charge was distributed to the second site when the pH exceeded 8 as the H+ started to dissociate from the site. The charge distribution is useful to indicate the type of the site contributing to surface charge in any pH range.

Parameter Estimation
Surface charge curves (Fig. 1 and 2) reveal two types of binding sites on the surface. Each site behaved as an acid group, which may be described with a median log K value. The fitted two log K values varied from 2 to 4.2 for the first site and from 7.7 to 9.5 for the second site. The results are within the range of carboxylic (log Ka of 2–4) and phenolic sites (log Ka of 6–10) on HS (Milne et al., 2001), indicating that surface sites behave in a manner similar to soil organic matter. Therefore, the surface materials appear to behave as organic matter diluted by other soil components and have H+ binding properties similar to HS.

The results of maximum H+ binding (Qmax) for both sites varied within a wide range mainly because of the large variation of soil chemical properties. Organic soils possessed more variable surface charge; therefore, they have higher Qmax than mineral soils (Table 3). The binding distribution parameters, m1 and m2, were within the range of 0.6 to 0.99 and 0.45 to 0.99, respectively (Table 3). These results were comparable with what has been reported for HS, 0.38 to 0.89 for m1 and 0.14 to 0.86 for m2 (Milne et al., 2001). The m defines the width of the binding site distribution. A small value of m indicates a highly heterogeneous and hence a wide distribution of site affinities. In this case, the inflection points are not very clear, making for a relatively flat titration curve as is seen in the shape of the titration curve for the Valleyfield organic soil (VAL 4; Fig. 1) with an m2 value of 0.45. The shape of this curve can be compared with those of the Sudbury and Rouyn-Noranda organic soils SUD 12 and RN 7, which have values of m in the range 0.85 to 0.99 (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 3. Optimized parameters and goodness of fit for the nonideal competitive adsorption (NICA)–Donnan model.

 
The NICA–Donnan model provided an excellent fit (e.g., Fig. 1 and 2) to the acid-base titration data, as shown by the highly significant coefficient of determination, R2 (>0.95; Table 3). The agreement between the experimental results and model predictions indicates that the NICA-Donnan model, developed from humic acids, may be extended to soils with different properties, since the pKa values for two types of sites lie in the range for both organic and inorganic functional groups. In addition, the surface proton binding of soil samples may be described by the bimodal distribution with a reasonable number of adjustable model parameters (six in total).

Relationships between Surface Charge and Soil Characteristics
Surface components such as clay, oxides, and organic matter may all contribute to proton binding since they possess either inorganic hydroxyl groups or organic functional groups ({equiv}SO–) (Bolan et al., 1999). It may be possible to estimate the components that control the surface acid-base chemistry by statistical means. For example, Duquette and Hendershot (1993b) considered that the surface adsorption maxima (Qm) were a function of soil properties; Qm results were significantly correlated with SSA, clay, Org. C, and oxalate extractable Al and Fe. In this study, we applied backward stepwise linear regressions to establish relationships between Qmax, log K, and m of each site and soil properties such as CEC, SSA, clay, Org. C, and Al and Fe oxides. Variables were entered into the model if they met a 0.05 significance level. Exploratory data analysis revealed that Qmax1 was significantly correlated with the log of Org. C and CEC (Fig. 3) . The regression equations are

[4]

[5]

[6]



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 3. Type 1 maximum surface charge (Qmax1) as a function of soil (a) organic C (Org. C) and (b) cation-exchange capacity (CEC).

 
Likewise, the log of Org. C strongly affected Qmax2 (Fig. 4a) . Interestingly, although soil CEC was not correlated with the maximum charge for the Type 2 sites (Eq. [8]), its combination with Org. C significantly influenced Qmax2 (Eq. [9]). This may imply that this type of sites is comprised of both organic matter and mineral functional groups. The equations representing these relationships are

[7]

[8]

[9]



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 4. Type 2 maximum surface charge (Qmax2) as a function of soil (a) organic C (Org. C) and (b) cation-exchange capacity (CEC).

 
On the surface of soil particles, minerals such as clay and oxides may form strong bonds with organic matter. Consequently, the reactive surfaces may be a complex mixture of organic and inorganic soil materials and this may explain the fact that Org. C is the common property related to the Qmax of both sites.

However, there were no significant relationships between soil properties and log K and m, the parameters describing the binding affinity and its distribution. The lack of statistical relationships may be attributed to the small range of these parameters. In addition, since log K and m are essentially composite parameters that describe the influence of organic and inorganic functional groups on soil charging behavior, it is not surprising that these parameters were poorly correlated with individual soil properties. More importantly, since the significant equations for Qmax were derived from the soils with a wide range of chemical properties, they may be applicable to other soils. Therefore, we may be able to estimate the variable surface charge characteristics by using our equations to predict site concentrations for other soils without carrying out detailed acid-base titrations for each soil. Further research will determine the applicability of the above equations and the NICA–Donnan model to soil surface charge under diverse environmental conditions.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
The NICA–Donnan model successfully simulated the surface charge data, suggesting that the soil particle surfaces have H+ binding properties similar to HS and may be described by a bimodal binding distribution. The maximum proton adsorption (Qmax) was found to be related to soil properties (CEC and Org. C). The relationships between surface parameters and soil characteristics may provide a useful means to predict the surface charge properties in a wide range of field conditions.


   ACKNOWLEDGMENTS
 
We thank Hélène Lalande for her assistance in the chemical analysis and three anonymous reviewers for their constructive comments. The financial support for this project from the Metals In The Environment Research Network (MITE-RN), Toxic Substances Research Initiative (TSRI) of Health Canada, and a doctoral scholarship from "Fonds québécois de la recherche sur la nature et les technologies" (formerly FCAR) are greatly acknowledged.

Received for publication October 22, 2002.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
Y. Jiao, W. H. Hendershot, and J. K. Whalen
Modeling Phosphate Adsorption by Agricultural and Natural Soils
Soil Sci. Soc. Am. J., June 18, 2008; 72(4): 1078 - 1084.
[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 (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ge, Y.
Right arrow Articles by Hendershot, W.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Ge, Y.
Right arrow Articles by Hendershot, W.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Ge, Y.
Right arrow Articles by Hendershot, W.
Related Collections
Right arrow Soil Chemistry

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome