«THERMODYNAMIC MODELING OF METAL ADSORPTION AND MINERAL SOLUBILITY IN GEOCHEMICAL SYSTEMS A Dissertation Submitted to the Graduate School of the ...»
THERMODYNAMIC MODELING OF METAL ADSORPTION AND MINERAL
SOLUBILITY IN GEOCHEMICAL SYSTEMS
Submitted to the Graduate School
of the University of Notre Dame
in Partial Fulfillment of the Requirements
for the Degree of
Doctor of Philosophy
Daniel Scott Alessi
Graduate Program in Civil Engineering and Geological Sciences
Notre Dame, Indiana
UMI Number: 3364983
INFORMATION TO USERSThe quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
® UMI UMI Microform 3364983 Copyright 2009 by ProQuest LLC All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.
ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 © Copyright 2009 Daniel Scott Alessi
THERMODYNAMIC MODELING OF METAL ADSORPTION AND MINERAL
SOLUBILITY IN GEOCHEMICAL SYSTEMSAbstract by Daniel Scott Alessi Developing geochemical models that can accurately predict the fate and transport of heavy metals and actinides in soils and aquifers requires detailed information about the quantity and reactivity of each component, across a wide range of environmental conditions. The studies I present in this dissertation provide critical insights into the mechanisms of heavy metal and actinide transport and immobilization by answering questions including: (1) To what extent to monovalent cations adsorb to the cell walls of bacteria? (2) How accurate is the chloroform fumigation-extraction method in determining biomass carbon in naturally occurring soils? (3) Can surface complexation models that are developed for the adsorption of metals on to single sorbents be combined ' to predict metal distribution in multi-sorbent systems? (4) Does the incorporation of Np(V) into the structure of the uranyl silicate affect its solubility, and therefore thermodynamic stability?
The results of study (1) demonstrate that monovalent cations adsorb to the cell walls of bacteria, albeit weakly. I model the adsorption data invoking discrete surface functional groups on the bacteria, and show how this approach is a simpler alternative to
adsorbs to clays during fumigation, artificially inflating the organic carbon pool by entering into the subsequent extraction solutions. Because the increase in total organic carbon in the extracts is interpreted as biomass, the method must be corrected for the effects of chloroform contamination to be accurate. Study (3) demonstrates that surface complexation models of metal adsorption to individual soil components can be combined to predict Cd distribution in mixtures of these sorbents in many cases;
however, to describe Cd adsorption behavior in the presence of a dissolved organic ligand, models must invoke ternary complexes between the sorbents, Cd, and the organic ligand. Finally, in study (4), it is revealed that a relatively small substitution of Np for U in the structure of soddyite causes a dramatic decrease in its solubility. This result has important implications for geochemical modeling of repositories because Npincorporated soddyite will release substantially less U into solution than pure soddyite.
To my family.
2.1 (A) Lithium adsorption to B. subtilis as a function of ionic strength and pH.
Initial experimental conditions were 20 g l"1 B. subtilis cells and 2.34 x 10'5 M Li. The pH 5 model curve represents the best-fit model that accounts for Li adsorption onto Site 2 only. Curves for the pH 7 and pH 9 models show the extent of adsorption that would be predicted using the KNC and Ku values determined from modeling the pH 5 data, and assuming no additional adsorption of Li onto Sites 3 or 4. (B) Best-fit model for Li adsorption to Site 2 of B. subtilis at pH 5 (solid curve). Dashed curves are models resulting from a ± 0.2 variation in the best-fitting log stability constant value of Ku 19 2.2 (A) Rubidium adsorption to B. subtilis as a function of ionic strength and pH. Initial experimental conditions were 20 g l"1 B. subtilis cells and 2.34 x 10"5 M Rb. The pH 5 model curve represents the best-fit model that accounts for Rb adsorption onto Site 2 only. Curves for the pH 7 and pH 9 models show the extent of adsorption that would be predicted using the K^a and A^j values determined from modeling the pH 5 data, and assuming no additional adsorption of Rb onto Sites 3 or 4. (B) Best-fit model for Rb adsorption to Site 2 of B.
subtilis at pH 5 (solid curve). Dashed curves are models resulting from a ± 0.2 variation in the best-fitting log stability constant value of KRb 20 2.3 (A) Cd adsorption to B. subtilis cells in Na-, K-, and Li-perchlorate electrolytes.
Initial experimental conditions were 10 g l"1 B. subtilis cells and
8.90 x 10'5 M Cd in a 0.1 M perchlorate solution. Curves indicate the best-fit models for Cd adsorption onto bacterial Site 2 with (solid curve) and without (dashed curve) inclusion of the Na-Site 2 complexation reaction in the model.
The speciation of Site 2, with and without Na-Site 2 complexation, is depicted in (B) and (C), respectively 22
3.4 Results for fumigation experiments with a constant 50 g of silica sand and varying amounts of bacteria, plotted in terms of the concentration of extracted TOC as a function of the initial mass of bacteria present in the sample 46
4.1 Adsorption of (A) 8.9 x 10"5 M Cd(II) and (B) 8.9 x lO-6 M Cd(II) to 1 g T1 HFO ( • ), 1 g l'1 B. subtilis cells ( • ), and 1 g l"1 kaolinite. Curves represent best-fit models to HFO (grey line), B. subtilis (solid line), and kaolinite (dashed line) Cd adsorption data 60
4.2 Adsorption of (A) 8.9 x 10"5 M Cd(II) and (B) 8.9 x 10"6 M Cd(II) to mixtures of HFO and B. subtilis cells. Dashed lines represent best-fit models to 1 g l"1 HFO and B. subtilis end members from Figure 1. Darkened symbols and lines represent adsorption data and predicted adsorption behavior for two-sorbent mixtures, including 0.75 g l"1 HFO + 0.25 g l"1 B. subtilis cells (A, thin line), and
0.25 g l'1 HFO + 0.75 g V B. subtilis cells ( •, thick line) 71
4.3 Adsorption of (A) 8.9 x 10"5 M Cd(II) and (B) 8.9 x 10-6 M Cd(II) to mixtures of kaolinite and B. subtilis cells. Dashed lines represent best-fit models to 1 g l'1 kaolinite and B. subtilis end members from Figure 1. Darkened symbols and lines represent adsorption data and predicted adsorption for two-sorbent mixtures, including 0.75 g l"1 B. subtilis cells + 0.25 g l"1 kaolinite (A, thin line), and 0.25 g l"1 B. subtilis cells + 0.25 g l"1 kaolinite ( •, thick line) 73
4.4 Adsorption of (A) 8.9 x 10"5 M Cd(II) and (B) 8.9 x 10"6 M Cd(II) to mixtures of HFO and kaolinite. Dashed lines represent best-fit models to 1 g l"1 HFO and kaolinite end members from Figure 1. Darkened symbols and lines
4.5 Adsorption of 8.9 x 10"5 M Cd to 1 g l"1 A) HFO, C) B. subtilis, and E) kaolinite, and 8.9 x 10"6 M Cd to B) HFO, D) B. subtilis, and F) kaolinite.
Darkened symbols and lines represent adsorption data and predicted adsorption for mixtures without acetate. Open symbols are adsorption data in the presence of 0.3 M acetate. Dashed line indicates the predicted extent of adsorption in acetate experiments without the adsorption of a Cd-acetate complex, and grey line is the best-fit model including a Cd-acetate ternary surface complex.77
4.6 Adsorption of 8.9 x 10"6 M Cd to three-component mixtures of HFO, kaolinite, and B. subtilis, and in the presence of 0.3 M dissolved acetate.
Darkened symbols and lines represent adsorption data and predicted adsorption for mixtures without acetate. Open symbols and dashed lines are adsorption data and predicted adsorption in the presence of 0.3 M acetate, respectively. Sorbent mixtures include A) 0.8 g l"1 HFO + 0.1 g l"1 B. subtilis + 0.1 g l'1 kaolinite, B)
0.1 g l-1 HFO + 0.8 g l:1 B. subtilis + 0.1 g T' kaolinite, C) 0.1 g l"1 HFO + 0.1 g 1" B. subtilis + 0.8 g l"1 kaolinite, D) 0.33 g l'1 HFO + 0.33 g l"1 B. subtilis + 0.33 g l"1 kaolinite 81
4.1 Proton and Cd reactions at Bacillus subtilis cell walls, kaolinite, and HFO 61 '
4.2 Molal site concentrations of sorbents in two- and three- component mixtures 70
5.2 Equilibrium species concentrations and pH from each solubility experiment used for thermodynamic calculations, reported as log molalities 93
I would like to thank my advisor, Prof. Jeremy Fein, for his mentoring and support during my doctoral studies. His enthusiasm, availability, and patience made all the difference. I would also like to thank the members of my committee, including Profs.
Peter Burns, Patricia Maurice, and Robert Nerenberg, for reading my dissertation and for providing advice and assistance throughout my time at Notre Dame. I thank Jennifer Szymanowski for giving invaluable help in completing many of the experiments included here, and Dennis Birdsell and Jon Loftus for frequent help with analytical instruments.
My stipend and tuition were largely subsidized by a University of Notre Dame Presidential Doctoral Fellowship provided by the Arthur J. Schmitt Foundation. Thanks to the trustees and fellows of the Schmitt Foundation for their generous support. I was partially supported by an Environmental Molecular Science Institute grant from the National Science Foundation to the University of Notre Dame.
1.1 Overview and Research Questions The ability to accurately predict the mobility of metal cations and radionuclides in geologic systems depends on accurate characterization of the reactivity of every component in the system with the elements of interest. Adsorption, dissolution, and precipitation reactions at bacterial, organic, and mineral surfaces control the fate of these aqueous species in geologic systems. My dissertation research addresses several
important questions that relate to metal or radionuclide mobility, including:
(1) Do monovalent metal cations adsorb to bacterial cell walls? If they do adsorb, can we quantify the stability of the important bacterial surface complexes, and by so doing can we account for the effect of ionic strength on the adsorption of higher charged cations onto bacteria? It is often assumed that monovalent cations exhibit negligible site-specific binding to cell wall functional groups, and therefore only influence the adsorption of higher charged metals through non-specific electrostatic surface charging effects. However, this assumption has not been rigorously tested. If site-specific monovalent metal binding occurs, then accounting for site competition between monovalent and higher valence ions may be a more straightforward approach to quantifying the effect of ionic strength on metal adsorption than the current models of surface electric field effects. In Chapter 2,1 report results from experiments that measured the adsorption of Li+ and Rb+ onto the Gram-positive soil bacteria Bacillus subtilis as a function of pH and electrolyte ionic strength to determine stability constants for these metals and Na+. Using the stability constant determined for the surface complexes between these monovalent cations and the bacteria, we correct previously determined apparent stability constants for metal-bacterial surface complexes involving Cd2+.
(2) Is the fumigation-extraction procedure for determining biomass content of geologic samples accurate and/or precise enough to enable calculations of metal speciation in mass transport models? Metal binding to microbial cells can significantly affect the speciation and overall mobility of metals in groundwater and soils, so the biomass content must be precisely and accurately determined in order to model the fate and transport of metals in these systems. The fumigation-extraction procedure is the most commonly-accepted method of determining soil microbial biomass content, but control experiments for the procedure have never been conducted, so its accuracy and precision are not constrained. I performed chloroform fumigation experiments on individual soil components, including silica sand, montmorillonite, kaolinite, a humic acid, and Bacillus subtilis bacterial cells, to determine if the method can accurately determine biomass amounts, and if chloroform adsorption to these surfaces during fumigation affects the results of this procedure. Based on the data from these experiments, I discuss the validity of the fumigation-extraction method for various soil types in Chapter 3.
(3) Can we develop models that accurately account for metal binding and distribution among the various components of soils? Although there are many empirical studies of bulk adsorption of metals onto soils, the ability of surface complexation models (SCM) to model the speciation and distribution of metals in multi-sorbent systems has not been widely tested. In this study (Chapter 4), Cd2+ adsorption to mixtures of soil components is measured at varying pH and ionic strength conditions to determine the capability of a non-electrostatic SCM to independently account for the distribution of the metal to the surfaces.