| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Earth and Physical Sciences Division University of Texas San Antonio, TX 78249-0663
b Department of Plant and Soil Sciences The University of Tennessee Knoxville, TN 37901-1071
Abbreviations: SCM, surface complexation model • TLM, triple-layer model
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION REFERENCES |
|---|
Dr. Lützenkirchen is highly critical of the surface parameters selected and employed in the TLM examination of Hg(II) adsorption. His first concern is associated with the selection of literature values used to describe surface protonation and deprotonation reactions, stating that model predictions should be compared with pertaining experimental data, inherently requiring the acid/base characterization of the adsorbent. We agree with this approach. Ideally, all input parameters to the TLM should be evaluated for the employed adsorbents and experimental conditions. However, obtaining input parameters for SCMs from relevant literature is a very common practice, and alleviates the need to ‘reinvent the wheel.’ Even Lützenkirchen has coauthored a paper that used not only literature-obtained surface parameters, but even literature-obtained experimental Hg(II) adsorption results (Tiffreau et al., 1995). Further, the use of SCMs to predict adsorptive behavior on minerals with more than one type of surface functional group (e.g., kaolinite) or in chemically complex environments (e.g., soils), will require a reliance on functional group-specific data. For quartz, TLM parameters were not taken from more recent compilations (e.g., Sahai and Sverjensky, 1997), but from literature dealing with Min-U-Sil5, the commercial product used in this study. Similarly, TLM parameters for gibbsite were obtained from literature dealing with Superfine-4, the Al(OH)3 employed in Sarkar et al. (1999). Further, the surface ionization constants employed for Superfine-4 are not inconsistent with those reported for Al oxides: log K+ = 5.2 to 7.9 and log K- = -8.1 to -10.0 (Schindler and Stumm, 1987; Goldberg, 1992). Although the site density values obtained from Meng and Letterman (1993) were not specific to Superfine-4, employing these values does not constitute a ‘dangerous procedure.’ Hayes et al. (1991) performed a sensitivity analysis and showed that the TLM was insensitive to changes in the value of site density over the range of values that are reasonable for mineral oxides (2–20 nm-2). Therefore, we considered the selection of a site density value of 8 nm-2 for Superfine-4 to be justified.
One might conclude by the absence of anion binding values in Table 1 of Sarkar et al. (1999) that we neglected nonspecific anion retention in the TLM. Clearly, we were remiss, not in omitting these reactions, but in our discussion. The inclusion of anion retention reactions for NO-3, Cl-, and SO2-4 had no impact on the predicted retention of Hg(II). Thus, in the final analysis, these reactions were excluded. Further, it is not clear to us how Lützenkirchen obtained his charging curves for gibbsite or the electrolyte (Na+) binding constant (log K = 1.7; for 
AlO--Na+ formation) reported in his letter. Our preliminary modeling results indicated that the inclusion of the anion binding reaction: AlOH0 + H+ + NO-3 = AlOH+2-NO-3 
had no impact on the charge characteristics of the gibbsite surface in the 0.1 M NaNO3 system; although the point of zero charge does fall between 7.3 and 8 [which is consistent with the pK+ and pK- values obtained from Johnson (1995) for Superfine-4]. The lack of impact of nonspecifically-bound anions on Hg(II) retention stands to reason, since the predicted mechanisms for Hg(II) retention are specific [XOHg
-2] and nonspecific cation retention (XO--HgOH0).
Lützenkirchen also questions the validity of the XOHg-2 surface complex, via the adsorption of the Hg02 species, arguing that the dominant species in solution is not necessarily the species that has the greatest impact on surface complex formation. We agree; however, Hg02 is the only Hg(II) species in solution above pH 5 in a 0.1 M NaNO3 matrix. Since compound adsorption is at least partially a function of aqueous speciation, it is not unrealistic to entertain the notion that Hg02 might contribute to Hg(II) adsorption. Further, we indeed considered the formation of the XOHgOH0 inner-sphere surface species during our attempts to model the experimental data. We theorized that the adsorption of the Hg02 species would proceed via the displacement of one of the Hg2+-coordinated OH- groups by the surface-bound oxygen, forming XOHgOH0. As is evident in Figs. 3 and 4 of Sarkar et al. (1999), consideration of the XOHgOH0 surface complex alone did not result in an adequate fit to the experimental adsorption data for either quartz or gibbsite. It appears that Lützenkirchen's resistence to the formation of XOHg-2 is based on the results of a modeling exercise he coauthored (Tiffreau et al., 1995). In an attempt to model the experimental Hg(II) retention data (to quartz) that were generated by MacNaughton (1973), Tiffreau et al. (1995) considered the formation of the inner-sphere XOHgOH0 complex alone to be significant. Their chemical model adequately predicted the Hg(II) adsorption edge (pH 3); but underestimated the Hg(II) adsorption maxima (pH 4 to 7) and overestimated Hg(II) adsorption in the pH 8 to 9 region. Our predicted Hg(II) adsorption envelops for both the quartz and gibbsite systems, considering only the XOHgOH0 surface complex, were similar to those produced by Tiffreau et al. (1995) and clearly did not adequately predict Hg(II) adsorption behavior. This observation prompted the consideration and acceptance of a chemical model that included the formation of the outer-sphere XO--HgOH+ complex and the inner-sphere XOHg-2 complex.
Finally, Lützenkirchen questions the reality of our proposed Hg(II) adsorption model (particularly with respect to Hg(II) retention on gibbsite). If one assumes the validity of the surface parameters employed for Min-U-Sil5 and Superfine-4, then reality must be based on goodness-of-fit. In Sarkar et al. (1999), the chemical model that included the formation of the XOHg-2 complex provided a substantially better fit to the experimental data, relative to the model that considered XOHgOH0 alone. Thus, the relevancy of the XOHg-2 complex was inferred. More problematic is the situation where two differing chemical models produce the same predicted behavior. Sarkar et al. (2000) report that the adsorption of Hg(II) by quartz, gibbsite, and kaolinite in a system containing Cl- can be partially explained by the formation of either XOHgOHCl- or XOHgCl0. The question then is which surface species forms and which does not? Which is reality? Clearly, the answer to such a question can not be ascertained from SCMs, as the results of a modeling exercise do not provide the final word on adsorption mechanisms. Relative to the chemical models proposed by Sarkar et al. (1999), Sarkar et al. (2000) evaluated the possibility of predicting Hg(II) adsorption on a multicomponent surface (kaolinite) using adsorption constants generated from the quartz and gibbsite studies of Sarkar et al. (1999). The goodness-of-fit of the modeled Hg(II) adsorption data to the experimental adsorption data on kaolinite in a 0.1 M NaNO3 matrix tends to support the reality of the chemical models developed by Sarkar et al. (1999).
Received for publication December 12, 2000.
|
REFERENCES |
|---|
|
TOPINTRODUCTIONREFERENCES |
|---|
-quartz. J. Colloid Interface Sci. 172:82–93.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
