Soil Science Society of America Journal 65:1348-1349 (2001)
© 2001 Soil Science Society of America
COMMENTS & LETTERS TO THE EDITOR
A Discussion of the Surface Complexation Modeling in the Paper by Sarkar et al. (1999)
Johannes Lützenkirchen
Forschungszentrum Karlsruhe GmbH Institut für Nukleare Entsorgung Postfach 3640 D-76021 Karlsruhe, Germany
luj{at}colenco.ch
 |
ABSTRACT
|
|---|
The application of the triple layer surface complexation model by Sarkar et al. is discussed, and it is shown that the authors obtain for gibbsite an acid/base model which is not in agreement with typical experimental data for that sorbent. Furthermore, it is concluded that experimental surface charge must be obtained data for the sorbent sample, because there is no unique surface charge for gibbsite based on the overall properties such as sorbent mass or specific surface area.
Abbreviations: SCM, surface complexation model
|
INTRODUCTION
|
|---|
IN THEIR PAPER on mercury adsorption of Hg(II) to quartz and gibbsite, Sarkar et al. (1999) model their experimental data using a surface complexation model (SCM). For such an application, acid/base properties of the sorbents quartz and gibbsite are required. Sarkar et al. (1999) take the corresponding SCM parameters from several literature sources and fit some others to the Hg(II) adsorption data. Some aspects of their procedure can, to say the least, be assessed.
|
Choice of parameters from the literature
|
|---|
The sources of the acid/base SCM parameters to describe the charging behaviors of the two crystalline sorbents are given in Table 1. At least three independent sources seem to be involved for each of the solids, some of them (at least the relevant ones by Meng and Letterman) pertaining to amorphous sorbents. This is a dangerous procedure, since a consistent set of parameters is required to assure that the properties of some solid be actually described. Most striking in Table 1 is the complete neglect of the anion binding constant for gibbsite. Even for quartz (if the formation constant for XOH+2 was not given the wrong sign by accident), anion binding would need to be considered. However, it is possible that there is a typing error with respect to the sign of this constant in Table 3 of Sarkar et al. (1999).
To fit the electrolyte binding constants to Hg(II), sorption data is most probably not a good idea. For gibbsite, it yields a rather large electrolyte binding constant of 101.7. A priori, it is impossible to foresee whether constants fitted in this way will accurately describe the charging behavior of a sorbent, because Hg(II) adsorption data contain no experimental information pertaining directly to the charging.
The above remarks suggest that the acid/base models should be compared to pertaining experimental data. Attempts to compare actual experimental titration data (Riese, 1999) indicated that for quartz, a good description of the data is obtained. Figure 1 shows the results of calculations for gibbsite. Neglect of the anion binding constant causes charging curves which are not in agreement with any previously published experimental data. For variations of ionic strength, shifting points of zero charge result (not shown).
Supposing that the anion binding was simply forgotten in Table 3 of Sarkar et al. (1999), and assuming symmetrical electrolyte binding yields a modified prediction with a point of zero charge of slightly below 7, which is not in agreement with the rather well-established value of around 10 (Hiemstra et al., 1999).
|
Sample inherent variations of surface charge
|
|---|
Different samples of nominally identical sorbents can exhibit different surface charge vs. pH behaviour. This is shown in Figure 1 on the basis of parameters published by Hiemstra et al. (1999), who also summarized older gibbsite surface charge studies. The compilation of the experimental data shows a huge variability. This means that it is necessary to identify the sample inherent charging properties, and apply a consistent model to them.
The conclusion from these two points is that the probability is very low for the gibbsite acid base model to correspond to something even close to reality. For quartz, one might wonder why the parameters have not been simply taken from a recent compilation (Sahai and Sverjensky, 1997). Even if the Hg(II) adsorption data can to some extent be described, the extracted stability constants can not later be combined with different (tabulated or more realistic) surface properties; the internal consistency would be lost, because fitted Hg(II) adsorption parameters depend on the underlying surface properties, as well as on the Hg(II) solution speciation scheme.
|
Further aspects
|
|---|
Also, the XOHg
-2 surface complex might be questioned. The species dominating in solution is, in the reasoning of the authors, the one to be most probably adsorbed. However, Hg(II) complexes involving two ligands (Cl- or OH-) appear to be the most stable ones over wide pH ranges. Furthermore, for SCMs, the dominating species in solution has not necessarily the biggest impact on the surface complex formed (Bargar et al., 2000). Maybe, with a more appropriate acid/base model (ideally based on sample inherent titration data), one might come to a more realistic Hg(II) sorption model.
The sample inherent differences make it very difficult to envision generally valid surface complexation parameters, if their purpose is to predict. So, if one has to distinguish between several samples of, for example, gibbsite, then it is necessary to use a model that takes care of sample variability. This is, to some extent, possible with the model used by Hiemstra et al. (1999). Even then, though, predictions will require the characterization of the sorbent sample.
Received for publication September 27, 2000.
|
REFERENCES
|
|---|
- Bargar, J., R. Reimeyer, and J.A. Davis. 2000. Spectroscopic confirmation of uranium (VI)-carbonato complexes on hematite. Environ. Sci. Technol. 33:2481–2484.
- Hiemstra, T., H. Yong, and W.H. van Riemsdijk. 1999. Interfacial charging phenomena of aluminum (hydr)oxides. Langmuir 15:5942–5955.[ISI]
- Sarkar, D., M.E. Essington, and K.C. Misra. 1999. Adsorption of mercury(II) by variable charge surfaces of quartz and gibbsite. Soil Sci. Soc. Am. J. 63:1626–1636.[Abstract/Free Full Text]
- Sahai, N., and D.A. Sverjensky. 1997. Evaluation of internally consistent parameters for the triple layer model by the systematic analysis of oxide surface titration data. Geochim. Cosmochim. Acta 61: 2801–2826.
TOP
ABSTRACT
INTRODUCTION
