The Pennsylvania State University
The Graduate School
Department of Geosciences
CONTAMINANT SEQUESTRATION AND PHASE TRANSFORMATION
PROPERTIES OF BIRNESSITE-LIKE PHASES (δ-MnO2)
A Dissertation in
Geosciences
by
Claire R. Fleeger
2012 Claire R. Fleeger
Submitted in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
May 2012
The dissertation of Claire R. Fleeger was reviewed and approved* by the following:
Peter J. Heaney Professor of Mineral Sciences Dissertation Advisor Chair of Committee
James D. Kubicki Professor of Geosciences
Matthew S. Fantle Assistant Professor of Geosciences
William D. Burgos Professor of Environmental Engineering
Chris J. Marone Professor of Geosciences Associate Department Head of Graduate Programs
*Signatures are on file in the Graduate School
iii
ABSTRACT
For the first time, time-resolved X-ray diffraction (TR-XRD) and inductively coupled plasma-mass spectrometry (ICP-MS) have been coupled to monitor contaminant
(e.g., Cs and Mn) cation exchange reactions into two birnessite-like phases, which exhibit triclinic and hexagonal symmetry. Cs is a principal contaminant at several nuclear waste sites, including the DOE Hanford Site, where high-level, high pH nuclear waste is leaking from tens of underground storage tanks. Our results indicate that Cs uptake by hexagonal H-birnessite increases as pH increases from pH 3 to pH 11; the increase in Cs uptake (wt%) correlates to an incremental increase in unit-cell volume as a function of pH. As Cs adsorption increases and pH increases, the rate of Cs cation exchange reaction decreases. On the other hand, triclinic Na-birnessite adsorbs less Cs at a neutral pH than at high or low pH. Again, greater uptake of Cs correlates to a decrease in Cs cation exchange rate as the pH becomes more acidic or basic. The Cs exchange rate equations determined from this study are as follows:
Hexagonal H-birnessite, pH 3 to 10:
d - Triclinic Na-birnessite, pH 4 to 6: x - d
d Triclinic Na-birnessite, pH 6.5 to 11: d
During the course of these variable pH reactions, we observed that a phase transformation occurred between the triclinic and hexagonal phases as a function of pH even in the presence of competing cations (i.e., Cs). At pH 3, triclinic Na-birnessite transformed to hexagonal H- or Cs-birnessite, and at pH 13, hexagonal H-birnessite
iv transformed to triclinic Na- or Cs-birnessite. At both low and high pH, cation exchange with the dissolved cation (H, Na, or Cs) preceded the symmetry-changing transformation.
The interlayer cation remained within the structure during the phase transformations, indicating that these materials act as sequestering agents no matter the pH or crystalline phase.
For the last set of experiments, we exchanged Mn2+ into triclinic Na-birnessite to assess the possibility of a redox-driven phase transformation to hexagonal H-birnessite.
We determined that aqueous Mn2+ catalyzed the phase transformation at pH 3, and the rate of phase transformation increased as [Mn2+] increased. However, at pH 4.5 and 6,
2+ Mn (aq) promoted the breakdown of the manganese oxide octahedral sheets. Initially the
Mn2+ exchanged into the interlayer region, indicated by a decrease in unit-cell volume, followed by an increase in the Mnoct-Ooct bond distances to 2.01 Å, suggesting that the
4+ 3+ octahedral Mn (s) was reduced to Mn (s). We supported this observation by completing the same reaction in a batch reactor, where the final solid-state products were a mixture of triclinic birnessite, hausmannite, and three Mn3+OOH products (manganite, groutite, and feitknechtite).
This investigation reveals the adaptability of birnessite as a function of chemical environment. Mn oxides control heavy metal mobility not only by high sorption and cation exchange capacities but by redox reactivity, and these properties vary with the particular phase of Mn oxide. Thus, each phase needs to be characterized individually to fully understand how mixtures will interact in natural systems.
v
TABLE OF CONTENTS
LIST OF FIGURES ...... viii
LIST OF TABLES ...... xiv
ACKNOWLEDGEMENTS ...... xvi
DEDICATION ...... xviii
Chapter 1 Introduction ...... 1
Importance of manganese oxides ...... 1 The effect of Mn oxides on contaminants: 137Cs ...... 5 Research objectives ...... 8 References ...... 9 Figures ...... 19
Chapter 2 Exchange of Cs into hexagonal H-birnessite as a function of pH ...... 21
Abstract ...... 21 Introduction ...... 21 Experimental Methods ...... 24 Material synthesis ...... 24 Chemical analysis ...... 25 Kinetic analysis ...... 27 Results...... 30 Discussion ...... 32 Controls of Cs uptake ...... 32 Transformation from hexagonal to triclinic symmetry ...... 33 Back reaction ...... 35 Mechanism of exchange ...... 35 Acknowledgements ...... 37 References ...... 37 Figures ...... 41 Tables ...... 55
Chapter 3 The rate and mechanism of Cs cation exchange into triclinic Na- birnessite ...... 60
Abstract ...... 60 Introduction ...... 60 Experimental Methods ...... 63 Material synthesis ...... 63
vi
Chemical analysis ...... 64 Kinetic analysis ...... 66 Results...... 69 Discussion ...... 72 Cs exchange rates and loading ...... 72 Back reaction ...... 75 Mechanism of exchange ...... 76 Conclusion ...... 77 Acknowledgements ...... 77 References ...... 78 Figures ...... 82 Tables ...... 99
Chapter 4 Exploring the phase transformation between triclinic and hexagonal birnessite at pH 3 and 13 ...... 104
Abstract ...... 104 Introduction ...... 105 Experimental section ...... 109 Material synthesis ...... 109 Structural analysis ...... 110 Kinetic analysis ...... 112 Results...... 113 Discussion: Triclinic Na-birnessite reactions at pH 3 ...... 113 Rietveld analyses ...... 113 ICP-MS analyses ...... 114 Phase transformations ...... 115 Kinetics of the triclinic-hexagonal transformations ...... 117 Discussion: Hexagonal H-birnessite reactions at pH 13 ...... 118 Transformation to triclinic Na-birnessite ...... 118 Transformation to triclinic Cs-birnessite ...... 119 Kinetics of the hexagonal-triclinic transformations ...... 121 Acknowledgements ...... 122 References ...... 122 Figures ...... 126 Tables ...... 143
Chapter 5 Reduction of triclinic Na-birnessite by Mn2+ cation exchange ...... 146
Abstract ...... 146 Introduction ...... 147 Experimental section ...... 150 Material synthesis ...... 150 Chemical analysis ...... 150 Cation exchange coefficients ...... 152
vii
Kinetic analysis ...... 153 One-stage cation exchange reactions ...... 153 Two-stage cation exchange reactions ...... 155 Phase transformation reactions ...... 157 Results and Discussion ...... 158 Unit-cell variations in triclinic birnessite ...... 159 Kinetic analyses based on unit-cell variations ...... 159 Kinetic and mechanistic insights from eluate measurements ...... 161 Phase transformation reactions and kinetics ...... 163 Acknowledgements ...... 165 References ...... 166 Figures ...... 169 Tables ...... 191
viii
LIST OF FIGURES
Figure 1.1. The crystal structure of triclinic Na-birnessite. The Mn-O octahedral sheets contain Mn3+/4+ in all octahedral positions with hydrated Na+ as interlayer cations (adapted from Post 2002)...... 19
Figure 1.2. The crystal structure of hexagonal H-birnessite with Mn4+-O + 2+ octahedral sheets and H /H2O and Mn as interlayer cations (adapted from Silvester et al. (1997))...... 20
Figure 2.1. The crystal structure of hexagonal H-birnessite with Mn4+-O + 2+ octahedral sheets and H /H2O and Mn as interlayer cations (adapted from Silvester et al. (1997))...... 41
Figure 2.2. Flow-through TR-XRD experiment setup at beam line 13-BM-C at APS...... 42
Figure 2.3. A schematic of the two stage cation exchange reaction: stage one is the complete exchange of cations while stage two is the reordering of cations within the interlayer...... 43
Figure 2.4. Stacked TR-XRD patterns of hexagonal H-birnessite exchanged with 0.001 M Cs+ at pH , with θ (°) along the x-axis, intensity along the y-axis, and time (min) along the z-axis where each pattern represents 5 min in time. .... 44
Figure 2.5. Unit cell parameters a (Å) are plotted versus time (min) as a function of pH: (A) pH 3, (B) pH 5, (C) pH 6.5, (D) pH 9, and (E) pH 10...... 45
Figure 2.6. Unit cell parameter c (Å) plotted versus time (min) at (A) pH 3, (B) pH 5, (C) pH 6.5, (D) pH 9, and (E) pH 10...... 46
Figure 2.7. Volume, V, (Å3) plotted versus time (min) fit with Vegard’s Law -1 (Eqn. 1.8) with k1 (min ) as the rate constant of reaction stage one and k2 (min-1) as the rate constant of reaction stage two at (A) pH 3, (B) pH 5, (C) pH 6.5, (D) pH 9, and (E) pH 10...... 47
Figure 2.8. The log of the initial rate plotted as a function of pH. The change in initial cation exchange rate is weakly dependent on pH...... 48
Figure 2.9. The change in volume (Å3) versus time (min) verses pH. The unit- cell volume increases as pH increases, indicating more Cs+ enters the interlayer region at a higher pH...... 49
ix
Figure 2.10. Plots of Cs adsorbed (wt%) from the flow-through solution as a function of time. The bold line indicates 0 wt% adsorption, while positive percentages are adsorption and negative percentages are desorption. More Cs is adsorbed into the hexagonal H-birnessite as the pH increases. Each plot represents a cation exchange reaction completed at a different pH: (A) pH 3, (B) pH 5, (C) pH 6.5, (D) pH 9, and (E) pH 10...... 50
Figure 2.11. Plots of the Mn concentration (ppm) in the leachate as a function of time (min). Each plot represents a cation exchange reaction completed at a different pH: (A) pH 3, (B) pH 5, (C) pH 6.5, (D), pH 9, and (E) pH 10...... 51
Figure 2.12. Mn(oct) site occupancy increases to 1.0 as pH increases, indicating the vacancies within the octahedral Mn-O sheet are filled with oxidized Mn2/3+ from the interlayer...... 52
Figure 2.13. Difference electron Fourier (DELF) maps of the interlayer region calculated by GSAS for 0.001 M Cs+ exchange with hexagonal H-birnessite at pH 6.5, for exchange times (A) 0 min, (B) 60 min, (C) 120 min, (D) 180 min, (E) 240 min, and (F) 300 min. The contour interval is 0.1 e-/Å3...... 53
Figure 2.14. ASEM elemental maps showing the distribution of Mn, O, and Cs on partially exchanged hexagonal H-birnessite at pH 9 (40 min)...... 54
Figure 3.1. The crystal structure of triclinic Na-birnessite. The Mn-O octahedral sheets contain Mn3+/4+ in all octahedral positions with hydrated Na+ as interlayer cations (adapted from Post 2002)...... 82
Figure 3.2. Flow-through TR-XRD experimental setup at beamline 13-BM-C at APS...... 83
Figure 3.3. A schematic of the two-stage cation exchange reaction: stage one is the complete exchange of cations while stage two is the reordering of cations within the interlayer...... 84
Figure 3.4. Stacked TR-XRD patterns of triclinic Na-birnessite exchanged with 0.001 M Cs+ at pH 11, with 2θ (°) along the x-axis, intensity along the y-axis, and time (min) on the z-axis where each pattern represents 5 min in time...... 85
Figure 3.5. Unit cell parameter a (Å) plotted versus time (min) for (A) pH 4.7, (B) pH 6.7, (C) pH 8.9, and (D) pH 10.7...... 86
Figure 3.6. Unit cell parameter c (Å) plotted versus time (min) for (A) pH 4.7, (B) pH 6.7, (C) pH 8.9, and (D) pH 10.7...... 87
Figure 3.7. Unit cell parameter β (°) plotted versus time (min) for (A) pH 4.7, (B) pH 6.7, (C) pH 8.9, and (D) pH 10.7...... 88
x
Figure 3.8. Unit cell parameter γ (°) plotted versus time (min) for (A) pH 4.7, (B) pH 6.7, (C) pH 8.9, and (D) pH 10.7...... 89
Figure 3.9. Volume, V, (Å3) plotted as a function of time (min) fit with Eqn. 3.8 for (A) pH 4.7, (B) pH 6.7, (C) pH 8.9, and (D) pH 10.7...... 90
Figure 3.10. The log of the initial rate plotted as a function of pH from three replicate reactions: (■) NSLS, March , (♦) APS, June , and (▲) APS, November Data from Lopano et al ( ) is also plotted (●) ...... 91
Figure 3.11. Plots of Cs adsorbed (wt%) from the flow-through solution as a function of time (min). Each plot represents a cation exchange reaction completed at a different pH: (A) pH 5, (B) pH 6.7, (C) pH 9, and (D) pH 10. .... 92
Figure 3.12. Na (ppm) released during the Cs cation exchange reaction plotted as a function of time (min). Each plot represents a cation exchange reaction completed at a different pH: (A) pH 5, (B) pH 6.7, (C) pH 9, and (D) pH 10. .... 93
Figure 3.13. Mn (ppm) released during the Cs cation exchange reaction plotted as a function of time (min). Each plot represents a cation exchange reaction completed at a different pH: (A) pH 5, (B) pH 6.7, (C) pH 9, and (D) pH 10. .... 94
Figure 3.14. Cs loading (mg) as a function of pH. Note that the amount of triclinic Na-birnessite is assumed to remain constant at 4 mg for each experiment...... 95
Figure 3.15. Cs loading (mg) as a function of time (min) at pH 10. Cs is sorbed during two distinct time frames, which may indicate there is a second sorption site on birnessite at pH 10...... 96
Figure 3.16. Na released (x 10-6 mol) from the triclinic birnessite as Cs (x 10-6 mol) is sequestered from solution. Each data point represents the mass of each element as the reaction proceeds forward in time as a function of pH: (A) pH 5, (B) pH 6.7, (C) pH 9, and (D) pH 10...... 97
Figure 3.17. ASEM elemental maps showing the distribution of Mn, O, Na, and Cs on partially exchanged triclinic Na-birnessite at pH 7 (60 min)...... 98
Figure 4.1. The crystal structure of triclinic Na-birnessite with Mn3+/4+ in all octahedral positions with hydrated Na+ in the interlayer positions, adapted from Post et al. (2002)...... 126
Figure 4.2. The crystal structure of hexagonal H-birnessite with Mn4+ occupying 2+/3+ most of the octahedral positions. Mn and H(H2O)x are within the interlayer. Adapted from Silvester et al. (1997)...... 127
xi
Figure 4.3. Flow-through TR-XRD experimental setup at beam line 13-BM-C at APS...... 128
Figure 4.4. TR-XRD patterns of triclinic Na-birnessite transforming to hexagonal H-birnessite in 0.001 M HCl...... 129
Figure 4.5. Triclinic Na-birnessite transforms to hexagonal Cs-birnessite in 0.001 M CsCl at pH 3...... 130
Figure 4.6. Hexagonal H-birnessite begins to transform to triclinic Na-birnessite in 0.1 M NaOH...... 131
Figure 4.7. Hexagonal H-birnessite transforms to triclinic Cs-birnessite in 0.1 M CsOH at pH 13...... 132
Figure 4.8. Changes in volume during pH 3 reactions in (A) 0.001 M HCl and (B) 0.001 M CsCl and 0.001 M HCl...... 133
Figure 4.9. Changes in Na and Mn concentration over the course of the reaction of triclinic Na-birnessite at pH 3...... 134
Figure 4.10. Changes in Cs, Na, and Mn concentration over the course of the reaction of triclinic Na-birnessite exchanged with 0.001 M CsCl at pH 3...... 135
Figure 4.11. The change in percentage of triclinic Na-birnessite present as a function of time at pH 3: (A) 0.001 M HCl and (B) 0.001 M HCl and 0.001 M CsCl...... 136
Figure 4.12. The rate of triclinic Na-birnessite exchanged with (A) 0.001 M HCl and (B) 0.001 M CsCl and 0.001 M HCl, both buffered to pH 3, follow first- order kinetics...... 137
Figure 4.13. Changes in volume during pH 13 reactions in (A) 0.1 M NaOH and (B) 0.1 M CsOH...... 138
Figure 4.14. Changes in Na and Mn concentration over the course of the reaction of hexagonal H-birnessite exchange with 0.1 M NaOH at pH 13...... 139
Figure 4.15. Percentage of Cs adsorbed to hexagonal H-birnessite exchanged with 0.1 M CsOH (pH 13). Na and Mn concentrations were below detection limits...... 140
Figure 4.16. The change in percentage of hexagonal H-birnessite present as a function of time at pH 13: (A) 0.1 M NaOH and (B) 0.1 M CsOH...... 141
xii
Figure 4.17. The rate of hexagonal H-birnessite exchanged with (A) 0.1 M NaOH and (B) 0.1 M CsOH at pH 13 follow first-order kinetics...... 142
Figure 5.1. The crystal structure of triclinic Na-birnessite with Mn3+/4+ in all octahedral positions with hydrated Na+ in the interlayer positions, adapted from Post et al. (2002)...... 169
Figure 5.2. The crystal structure of hexagonal H-birnessite with Mn4+ and vacancies in the octahedral positions with proton/water complexes in the interlayer and Mn2+ on either side of the vacancies...... 170
Figure 5.3. Flow-through TR-XRD experimental setup at beam line 13-BM-C at APS...... 171
Figure 5.4. Stacked TR-XRD patterns of triclinic Na-birnessite exchanged with 0.001 M MnCl2 at pH , with θ (°) along the x-axis, intensity along the y- axis, and time (min) on the z-axis where each pattern represents 2 min in time...... 172
Figure 5.5. Stacked TR-XRD patterns of triclinic Na-birnessite exchanged with 0.001 M MnCl2 at pH , with θ (°) along the x-axis, intensity along the y- axis, and time (min) on the z-axis where each pattern represents 5 min in time...... 173
Figure 5.6. Stacked TR-XRD patterns of triclinic Na-birnessite exchanged with 0.001 M MnCl2 at pH , with θ (°) along the x-axis, intensity along the y- axis, and time (min) on the z-axis where each pattern represents 5 min in time...... 174
Figure 5.7. Unit-cell volume changes of triclinic Na-birnessite exchanged at pH 3 with (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2...... 175
Figure 5.8. Unit-cell volume changes of triclinic Na-birnessite exchanged at pH 4.5 with (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2...... 176
Figure 5.9. Unit-cell volume changes of triclinic Na-birnessite exchanged at pH 6 with (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2...... 177
Figure 5.10. The log of the initial cation exchange rate (dV/dt) plotted as a function of (A) pH and (B) -log([Mn2+])...... 178
Figure 5.11. Plots of Mn adsorbed (wt%) from the flow-through solution as a function of time (min) at pH 3. Each plot represents a cation exchange reaction completed at (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2...... 179
xiii
Figure 5.12. Na (ppm) released during the Mn cation exchange reaction at pH 3 plotted as a function of time (min): (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2...... 180
Figure 5.13. Plots of Mn adsorbed (wt%) from the flow-through solution as a function of time (min) at pH 4.5. Each plot represents a cation exchange reaction completed at (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2...... 181
Figure 5.14. Na (ppm) released during the Mn cation exchange reaction at pH 4.5 plotted as a function of time (min): (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2...... 182
Figure 5.15. Plots of Mn adsorbed (wt%) from the flow-through solution as a function of time (min) at pH 6. Each plot represents a cation exchange reaction completed at (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2...... 183
Figure 5.16. Na (ppm) released during the Mn cation exchange reaction at pH 6 plotted as a function of time (min): (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2...... 184
Figure 5.17. Mn (x 10-6 mol) adsorbed onto triclinic Na-birnessite as Na (x 10-6 mol) is released from the birnessite interlayer during Mn cation exchange in 0.001 M MnCl2 at (A) pH 3.0, (B) pH 4.5, and (C) pH 6.0. Each data point represents the summation of Mn loaded and Na released from the birnessite as time progresses; therefore, the slope corresponds to the overall Kex value...... 185
Figure 5.18. Kex plotted against time (min) at (A) pH 3.0, (B) pH 4.5, and (C) pH 6.0. Each data point represents the Kex value associated with each individual fractionated solution...... 186
Figure 5.19. The change in percentage of triclinic Na-birnessite as a function of time at pH 3: (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2. .... 187
Figure 5.20. A single XRD pattern of triclinic Na-birnessite reacted with 0.1 M MnCl2 at pH for hr as a batch reaction (λ = 0.7093 Å)...... 188
Figure 5.21. The rate of phase transformation of triclinic birnessite to hexagonal birnessite at pH 3. First-order kinetics are followed at (A) 0.001 M MnCl2 and (B) 0.01 M MnCl2. Zero-order kinetics are followed at (C) 0.1 M MnCl2. .. 189
Figure 5.22. (A) The change in percentage of triclinic Na-birnessite as a function of time at 0.01 M MnCl2, pH 4.5. (B) The rate of the phase transformation of triclinic birnessite to hexagonal birnessite at 0.01 M MnCl2, pH 4.5...... 190
xiv
LIST OF TABLES
Table 2.1. Final Rietveld refinement parameter for end-member hexagonal H- birnessite exchanged with 0.001 M Cs+ at varying pHs...... 55
Table 2.2. Atomic coordinates and isotropic displacement factors for Cs- exchanged hexagonal H-birnessite across a wide pH range...... 56
Table 2.3. Selected bond distance values for Cs-exchanged hexagonal H- birnessite as a function of pH...... 57
+ Table 2.4. The first stage rate constant, k1, and initial rate, dV/dt, of Cs exchange into hexagonal H-birnessite as a function of pH...... 58
Table 2.5. [Cs] in effluent as a function of pH and time...... 59
Table 3.1. Final Rietveld refinement parameters for end-member triclinic Na- birnessite exchange with 0.001 M Cs+ at varying pHs...... 99
Table 3.2. Atomic coordinates and isotropic displacement factors for Cs- exchanged triclinic Na-birnessite across a wide pH range...... 100
Table 3.3. Bond distance values for Cs-exchanged triclinic Na-birnessite as a function of pH...... 101
+ Table 3.4. The first stage rate constant, k1, and initial rate, dV/dt, of Cs exchange into triclinic Na-birnessite as a function of pH for each replicate reaction...... 102
Table 3.5. [Cs] in effluent as a function of pH and time...... 103
Table 4.1. Rietveld refinement parameters for the final transformation structures. ... 143
Table 4.2. Atomic coordinates and isotropic displacement factors for transformed birnessite in the absence and presence of Cs at pH 3 and 13...... 144
Table 4.3. Bond distance values for transformed birnessite in the absence and presence of Cs at pH 3 and 13...... 145
Table 5.1. Final Rietveld refinement parameters for end-member triclinic Na- birnessite exchanged with Mn2+ at varying [Mn] and pH...... 191
Table 5.2. Atomic coordinates and isotropic displacement factors for Mn- exchanged triclinic Na-birnessite as a function of [Mn] and pH...... 192
Table 5.3. Bond distance values for Mn-exchanged triclinic Na-birnessite as a function of [Mn] and pH...... 193
xv
Table 5.4. The first-stage rate constant, k1, and initial rate, dV/dt, of the Mn- exchange reaction with triclinic Na-birnessite as a function of pH and [Mn]...... 194
Table 5.5. Zero-order and first-order rate constants, k, with respect to [triclinic birnessite] of the triclinic Na-birnessite phase transformation to hexagonal birnessite under varying experimental conditions...... 195
xvi
ACKNOWLEDGEMENTS
This research was made possible due to several funding sources: NSF-DOE
Center for Environmental Kinetics Analysis (CEKA) Grant No. CHE-0431328 and NSF
Grant No. EAR07-45374. Use of the Advanced Photon Source was supported by the U.
S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under
Contract No. DE-AC02-06CH11357. The staff at APS beam line 13-BM-C were fundamental to my success: Peter Eng, Nancy Lazarz, and Joanne Stubbs. Use of the
National Synchrotron Light Source, Brookhaven National Laboratory, was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02-98CH10886. Thanks to Jonathan Hanson at beam line
X7B. ASEM was completed at the Mineral Sciences Analytical Laboratory at the
Smithsonian Institution with Timothy Rose and Jeffrey Post. Mark Angelone, John
Cantolina, Waleska Castro, and Melanie Saffer assisted with XRD, SEM, and ICP-MS at the Materials Characterization Laboratory at the Pennsylvania State University.
My research group members were invaluable resources, as teachers, co-workers, and friends during this process. They include Christina Lopano, Daniel Hummer,
Timothy Fischer, Andrew Wall, and Kristina Peterson. Christina and Tim, thank you for answering my emails and phone calls even after you left birnessite behind and moved on in your careers. Kris, thank you for being such an awesome travelling companion and office mate; you made my graduate career much more bearable and humorous.
xvii
My committee has also offered invaluable advice and critiques: James Kubicki,
William Burgos, and Matthew Fantle. I especially thank my advisor, Peter Heaney, for accepting me as a chemist and for giving me the freedom to pursue my own interests.
Thank you for your guidance and support.
xviii
DEDICATION
This dissertation is dedicated to my family: Mom, Matt, and Kimberly. Your love, support, and prayers are the only reasons I felt adequate enough to conceive and attempt such a large dream.
1 Chapter 1 Introduction
Importance of manganese oxides
Manganese oxides (MnOx) compose up to one weight percent of soils and possess sorption and redox properties that control the mobility of heavy metals. There are more than 30 manganese oxide minerals and each one has distinctive properties (Post 1999).
Phyllomanganates, layered Mn oxides, are the most commonly occurring Mn oxides and include birnessite, ranciéite, and lithiophorite (Taylor et al. 1964; Lind and Hem 1993;
Bilinski et al. 2002; Vaniman et al. 2002). These are found in caves as coatings, dendrites, or nanocrystalline deposits (White et al. 2009; Xu et al. 2010; Frierdich et al.
2011). They are also found on the ocean floor, in island arcs, and in hydrothermal vents as ferromanganese nodules (Rona et al. 1976; Chukhrov et al. 1985; Post 1999; Dubinin et al. 2008; Reolid et al. 2011). In soils, manganese oxides are most likely a product of biogenic oxidation of Mn2+ and occur as coatings, nodules, and fine-grained aggregates
(Post 1999; Tebo et al. 2004; Saratovsky et al. 2006).
Birnessite is the most commonly occurring phyllomanganate in nature and is characterized by a 7 Å interlayer spacing (Taylor et al. 1964; McKenzie 1977). It was first found by Jones and Milne (1956) near Birness, Scotland in a fluvio-glacial deposit of gravel and occurs with triclinic or hexagonal symmetry. An early crystallographic study by Glover (1977) reported a hexagonal structure of natural birnessite taken from a core from the Gulf of Mexico. Other birnessites found in the Pacific Ocean and Pinal Creek,
AZ also indicate a hexagonal or pseudo-hexagonal crystal structure (Chukhrov et al.
2 1985; Bargar et al. 2009), and Tan et al. (2010) collected both hexagonal and triclinic birnessite at two sites in western PA. Frierdich et al. (2011) found that phyllomanganates deposited in caves were mixtures of hexagonal birnessite and vernadite (δ-MnO2).
4+ Triclinic birnessite consists of octahedral MnO6 sheets composed of Mn and
3+ 4+ 3+ Mn with the chemical formula Na0.56(Mn 1.44Mn 0.56)O4• H2O (Fig. 1.1). The presence of Mn3+ in place of Mn4+ results in negatively charged sheets, which are charge balanced by hydrated cations that reside between each of the sheets. Na/H2O complexes act as interlayer cations in synthetic triclinic Na-birnessite and reside halfway between two octahedral MnO6 sheets (Post et al. 2002). In hexagonal birnessite, the octahedral
4+ 3+ MnO6 sheets are composed of Mn , Mn , and Mn vacancies (□) with the chemical
3+ 2+ 4+ 3+ formula H0.33Mn 0.111Mn 0.055(Mn 0.722Mn 0.111□0.167)O2 (Fig. 1.2). The presence of
Mn3+ and vacancies in place of Mn4+ also results in negatively charged sheets, which are charge balanced by interlayer Mn2+/3+ and protons. These interlayer cations reside on opposite sides of an octahedral vacancy (Silvester et al. 1997).
The interlayer cations within triclinic and hexagonal birnessite are exchangeable with alkali and alkaline elements, trace metals, and heavy metals. Cationic Li, K, Mg,
Ca, and Ba all replace Na interlayer cations when the birnessite interacts with the alkali or alkaline enriched solutions (Golden et al. 1986; Paterson et al. 1986; Wang et al. 1996;
Lopano et al. 2007). Antimony and Bi were found to coprecipitate with Mn2O4 in a low pH solution (Burke 1970) while U sorbs onto birnessite at a high pH and undergoes a cation exchange reaction with birnessite at a low pH (Al-Attar and Dyer 2002; Brennecka et al. 2011). Pu will preferentially sorb onto natural MnOx rather than iron oxides or smectites in zeolitic tuff (Duff et al. 1999). Other cation exchange and sorption reactions
3 involve Pb2+ (O'Reilly and Hochella 2003; Zhao et al. 2009; Kwon et al. 2010; Zhao et al.
2010; Lee et al. 2011; Zhao et al. 2011), Cu2+ (Fu et al. 1991; Sherman and Peacock
2010; Wang et al. 2010), iron complexes (Duckworth et al. 2008; Kang et al. 2010), Zn2+
(Stahl and James 1991; Toner et al. 2006; Boonfueng et al. 2009), Co, Ni, Fe, Mn, and
Cd (Taylor and McKenzie 1966; Jenne 1968; Burns 1976; McKenzie 1980; Drits et al.
2002; Lanson et al. 2002; Villalobos et al. 2005; Meng et al. 2009; Novikov et al. 2010;
Pena et al. 2010; Roque-Rosell et al. 2010; Zhu et al. 2010b; Cho et al. 2011). Birnessite also undergoes redox reactions with cations such as Cr(III) where the crystal structure transforms from triclinic to hexagonal (Negra et al. 2005; Feng et al. 2006; Dai et al.
2009; Fischer 2011).
Apart from controlling trace and heavy metal mobility in soils, manganese oxides can also be used for energy storage options such as batteries and supercapacitors
(Armstrong and Bruce 1996; Liu et al. 2000; Donne et al. 2010; Coleman et al. 2011).
They can exhibit large magnetoresistance (Mathews et al. 1997; Tan et al. 2003) and catalytic properties (Brock et al. 1998). MnOx exhibits strong oxidative properties and therefore acts as an oxidizing agent that converts toxic compounds into less toxic forms in the environment. Birnessite is known to oxidize benzylic and allylic alcohols into their corresponding aldehydes (Kamimura et al. 2011), iodide to form iodoalkanes (Allard et al. 2010), and atrazine, a common carcinogenic herbicide (Shin et al. 2000). It also activates persulfate into a sulfate radical, resulting in a stronger oxidant (Ahmad et al.
2010).
Natural birnessites are typically products of bacterial or fungal precipitation and there has been some debate over pathways and mechanisms of formation (Tebo et al.
4 2004; Petkov et al. 2009; Spiro et al. 2010). The crystallinity of the birnessite is dependent on the formation reaction pathway and the environmental conditions under which it is precipitated. Some studies (Saratovsky et al. 2009; Grangeon et al. 2010) conclude that fungi produce either vernadite or todorokite, but Petkov et al. (2009) argue that fungal MnOx is todorokite whereas bacterial MnOx is vernadite. Bacterial MnOx is produced by Mn(II) oxidizing bacteria such as Bacillus sp. strain SG-1 (Mandernack et al. 1995; Bargar et al. 2005; Webb et al. 2005b; Webb et al. 2005a), Pseudomonas putida
(Villalobos et al. 2003; Toner et al. 2006; Villalobos et al. 2006; Feng et al. 2010),
Leptothrix discophora SP6 (Jurgensen et al. 2004; Saratovsky et al. 2006), and
Brachybacterium sp. strain Mn32 (Wang et al. 2009). The majority of these laboratory- based biologically produced MnOx phases have hexagonal symmetry, but natural nanoparticles of MnOx are generally poorly crystalline and have variable layer structures, making them difficult to analyze.
The hexagonal and triclinic birnessite structures can also transform from one phase to the other depending on the environmental conditions of the soil. The transformation can be initiated by particular cations in solution during the biogenic precipitation of birnessite (Webb et al. 2005b; Zhu et al. 2010a) or by bacterial reduction
(Fischer et al. 2008). Abiotic transformations tend to be controlled by redox reactions: low pH induces a disproportionation reaction of Mn(III) (Drits et al. 1997; Lanson et al.
2000) and cations such as Co2+ act as reducing agents of octahedral Mn3+/4+ (Manceau et al. 1997).
5 The effect of Mn oxides on contaminants: 137Cs
The DOE Hanford Site in Richland, Washington was a plutonium production site from 1943 to 1991. Approximately 60% of the total amount of weapons-grade plutonium produced by the United States was produced at this site. The high-level nuclear waste is the result of three processes used to extract Pu, resulting in large quantities of aqueous nitric acid, uranium, and other fission products. The solutions were neutralized with
NaOH and stored in 177 underground single- and double-shell steel tanks (Edgemon et al. 2009). Direct measurements of tank inventories are not possible due to high radiation but ionic concentrations are modeled by Lichtner and Felmy (2003). Waste types now include 707,921 m3 solid waste, 1.1 billion m3 contaminated soil and groundwater, 216 million L of high-level nuclear waste, and 708 m3 of nuclear materials. The single-shell tanks began leaking as early as 1956 and the most significant leak of 435,275 L occurred in 1973 (Edgemon et al. 2009). About 3,785,000 L of waste have leaked from the single- shell tanks in part due to residual stress from welding, nitrate ions, high temperatures, and corrosion from the high pH (Gee et al. 2007; Edgemon et al. 2009). Twenty eight double-shell tanks were constructed between 1968 and 1986; they contain both solid and liquid waste and no leaks have occurred to date (Edgemon et al. 2009).
The Hanford and Ringold formations underlie the tank farm at the Hanford Site.
The Hanford formation consists of unconsolidated sediments with gravel, sand, and silt grain sizes of granitic and basaltic provenance. The Ringold formation is predominantly sand and cobble-sized gravel with significant amounts of clay, silt, and sand (Lindsey and
Gaylord 1990; Christensen et al. 2007; Zachara et al. 2007). These sediments are coated
6 in discrete iron and manganese oxide layers, and the MnOx makes up 0.4-0.93 wt% of the soil (Barnett et al. 2002; Fredrickson et al. 2004). The sediments are glacial-fluvial in origin where large amounts of poorly consolidated sands and gravels were deposited.
Clastic dikes tend to act as cutoff walls so contaminant plumes move primarily in the vertical direction (Gee et al. 2007). The contaminant plume is initially pH 14 but the pH decreases with increasing distance from the nuclear waste storage tank (Wan et al. 2004).
The plume travels through the Hanford formation which ends 20 m beneath the surface; the Ringold formation spans the next 40 m and crosses the water table. Once the plume reaches the water table, it has the potential to contaminate the Colombia River and other major water sources (Wan et al. 2004; Zachara et al. 2007).
Cs-137 is a byproduct of the Pu production process and is responsible for the majority of the radioactivity of the high-level nuclear waste at Hanford (McKinley et al.
2001; Gee et al. 2007). It was also released from nuclear power plant failures in
Chernobyl (Ukraine) and Fukushima (Japan). Its half life is ~30 years before decaying into metastable Ba-137m, which emits a gamma particle to achieve its stable state. If ingested, Cs-137 acts like K and is absorbed into the bloodstream, increasing the risks of cancer. In the first 11 days of the Chernobyl core meltdown in April 1986, 9 x 1016 Bq of
137Cs were released into the atmosphere, and these initially hot particles were too chemically inert to be soluble or exchangeable. One percent of the released 137Cs became exchangeable by 1991, and this fraction is more easily adsorbed onto the fine sediments in slow moving waters found in the Pripyat and Dnieper Rivers (Onishi et al. 2007). The most recent nuclear power plant partial meltdown occurred at the Fukushima Daiichi
Nuclear Power Plant in March 2011. Radionuclides such as 131I and 137Cs were released
7 but their concentrations were extremely low (Takemura et al. 2011). The largest amount of 137Cs measured in the air was 1200 Bq/m2, measured in Yagamata, Japan on March 28,
2011. This plume traveled to Thessaloniki, Greece where 137Cs was measured at 145
Bq/m3 on April 4, 2011, which is a negligible concentration (Manolopoulou et al. 2011).
The degree of contamination produced by the Fukushima disaster, however, is still only poorly known.
Cs can undergo several different chemical reactions as it travels through the soil profile underneath the Hanford tank farm. At the high temperature and pH underneath the tanks, Cs is incorporated into an albite-analcime assemblage through dissolution and recrystallization (Redkin and Hemley 2000), but a high temperature decreases the preference for Cs during cation exchange reactions (Liu et al. 2003). Cs also exchanges into zeolites such as cancrinite, sodalite, allophone, and zeolite A, replacing Na and K from the zeolite cages (Mon et al. 2005; El-Kamash 2008; Borai et al. 2009). Cs mobility is primarily controlled by cation exchange reactions into materials like illite and smectite and is retarded more than other alkali metals in the Hanford sediments (Comans et al.
1991; Jeong et al. 1996; Lichtner et al. 2004; Thompson et al. 2010). A high ionic strength decreases the ability of Hanford sediments to sequester Cs, indicating Cs becomes more mobile as the ionic strength increases (Flury et al. 2004).
Cs interaction with MnOx has also been studied in terms of nuclear waste remediation. Singh and Tandon (1977) concluded that a hydrated manganese oxide has a lower adsorption capacity for Cs than cations such as Pb, Cd, and Zn, and the adsorption capacity decreases with increasing alkali concentration. Mikhail and Misak (1988) completed Cs and Co adsorption studies on a structure similar to nsutite (γ-Mn2O3),
8 where Cs adsorption followed a Freundlich isotherm and was greatest at a neutral pH. Cs also had the largest distribution coefficient over other radionuclides such as Sr, Y, Ce, and Eu. However, Dyer et al. (2000) completed Cs, Sr, and Co exchange on birnessite in the presence of competing alkali and alkaline cations and competing complexing agents such as EDTA, tetraborate, and citrate. They found that the Cs distribution coefficient was the smallest in the presence of the competing species, which is the opposite conclusion of Mikhail and Misak (1988), but may be attributed to differences in the crystal structures of nstutite and birnessite. Al-Attar et al. (2003) completed cation exchange studies on synthetic Na- and H-birnessite (crystallography unspecified) with radioactive waste from the Ginna and Diablo Canyon Nuclear Power Plants. H-birnessite uptook more 60Co, 125Sb, and 110Ag than 134/137Cs due to the low pH of the equilibrium solution. Na-birnessite did not significantly reduce Cs concentration in the solution. A hydrous manganese oxide was found to uptake more Cs as Cs concentration, temperature, and pH increased (Mishra and Vijaya 2007). Liu et al. (2007) synthesized a Cs-type layered manganese oxide with hexagonal symmetry where the Cs was exchanged with H+ in 1 M HNO3; other alkali metals can then be inserted into the H-layered MnOx interlayer. This material also exhibits ion-sieve properties for the adsorption of cations.
Research objectives
This dissertation addresses the following questions:
1. How does a variation in pH, such as occurs in the Hanford Site contamination
plume, affect Cs+ exchange into triclinic Na- and hexagonal H-birnessite?
9 2. What are the rates of Cs cation exchange into triclinic Na- and hexagonal H-
birnessite? Is one phase of birnessite more efficient than the other?
3. What controls the polymorph of natural birnessite, and does phase cycling exist in
soils?
2+ 4. What effect does Mn (aq) have on the crystal structure of birnessite?
In this study, the structural changes of triclinic Na- and hexagonal H-birnessite have been monitored in situ with time-resolved X-ray diffraction (TR-XRD) as a function of pH and cation concentration. Cs cation exchange reactions are completed from pH 3 to 13 to model the man-made environmental conditions found at the Hanford Site and to represent how a pH gradient affects the cation exchange properties and phase of natural birnessite.
These reactions produced a third study that supports the idea of phase cycling between triclinic and hexagonal birnessite which is dependent on solution pH, or redox conditions.
Lastly, hexagonal-H birnessite contains interlayer Mn2+, which may serve as a second control of the birnessite symmetry. The final study consists of Mn2+ cation exchange reactions with triclinic Na-birnessite to determine whether interlayer Mn can reduce octahedral Mn and create octahedral vacancies that may induce a phase transformation to hexagonal birnessite.
References
Ahmad, M., Teel, A.L., and Watts, R.J. (2010) Persulfate activation by subsurface minerals. Journal of Contaminant Hydrology, 115(1-4), 34-45. Al-Attar, L., and Dyer, A. (2002) Sorption behaviour of uranium on birnessite, a layered manganese oxide. Journal of Materials Chemistry, 12(5), 1381-1386.
10 Al-Attar, L., Dyer, A., Paajanen, A., and Harjula, R. (2003) Purification of nuclear wastes by novel inorganic ion exchangers. Journal of Materials Chemistry, 13(12), 2969- 2974. Allard, S., Gallard, H., Fontaine, C., and Croue, J.P. (2010) Formation of methyl iodide on a natural manganese oxide. Water Research, 44(15), 4623-4629. Armstrong, A.R., and Bruce, P.G. (1996) Synthesis of layered LiMnO2 as an electrode for rechargeable lithium batteries. Nature, 381(6582), 499-500. Bargar, J.R., Fuller, C.C., Marcus, M.A., Brearley, A.J., and De la Rosa, M.P. (2009) Structural characterization of terrestrial microbial Mn oxides from Pinal Creek, AZ. Geochimica Et Cosmochimica Acta, 73, 889-910. Bargar, J.R., Tebo, B.M., Bergmann, U., Webb, S.M., Glatzel, P., Chiu, V.Q., and Villalobos, M. (2005) Biotic and abiotic products of Mn(II) oxidation by spores of the marine Bacillus sp. strain SG-1. American Mineralogist, 90(1), 143-154. Barnett, M.O., Jardine, P.M., and Brooks, S.C. (2002) U(VI) adsorption to heterogeneous subsurface media: Application of a surface complexation model. Environmental Science & Technology, 36(5), 937-942. Bilinski, H., Giovanoli, R., Usui, A., and Hanzel, D. (2002) Characterization of Mn oxides in cemented streambed crusts from Pinal Creek, Arizona, U.S.A., and in hot-spring deposits from Yuno-Taki Falls, Hokkaido, Japan. American Mineralogist, 87, 580-591. Boonfueng, T., Axe, L., Yee, N., Hahn, D., and Ndiba, P.K. (2009) Zn sorption mechanisms onto sheathed Leptothrix discophora and the impact of the nanoparticulate biogenic Mn oxide coating. Journal of Colloid and Interface Science, 333(2), 439-447. Borai, E.H., Harjula, R., Malinen, L., and Paajanen, A. (2009) Efficient removal of cesium from low-level radioactive liquid waste using natural and impregnated zeolite minerals. Journal of Hazardous Materials, 172(1), 416-422. Brennecka, G.A., Wasylenki, L.E., Bargar, J.R., Weyer, S., and Anbar, A.D. (2011) Uranium Isotope Fractionation during Adsorption to Mn-Oxyhydroxides. Environmental Science & Technology, 45(4), 1370-1375. Brock, S.L., Duan, N.G., Tian, Z.R., Giraldo, O., Zhou, H., and Suib, S.L. (1998) A review of porous manganese oxide materials. Chemistry of Materials, 10(10), 2619-2628. Burke, K.E. (1970) Study of Scavenger Properties of Manganese(IV) Oxide with Atomic Absorption Spectrometry - Determination of Microgram Quantities of Antimony, Bismuth, Lead, and Tin in Nickel. Analytical Chemistry, 42(13), 1536-1540. Burns, R.G. (1976) Uptake of Cobalt into Ferromanganese Nodules, Soils, and Synthetic Manganese(Iv) Oxides. Geochimica Et Cosmochimica Acta, 40(1), 95-102. Cho, Y., Jang, S., Kim, Y., Komarneni, S., and Kim, S. (2011) Uptake of cadmium, copper, and lead by microporous synthetic Na-birnessite. Journal of Porous Materials, 18(1), 125-131. Christensen, J.N., Conrad, M.E., DePaolo, D.J., and Dresel, P.E. (2007) Isotopic studies of contaminant transport at the Hanford Site, Washington. Vadose Zone Journal, 6(4), 1018-1030.
11 Chukhrov, F.V., Sakharov, B.A., Gorshkov, A.I., Drits, V.A., and Dikov, Y.P. (1985) Crystal structure of birnessite from the Pacific Ocean. International Geology Review, 27, 1082-1088. Coleman, J.N., Lotya, M., O'Neill, A., Bergin, S.D., King, P.J., Khan, U., Young, K., Gaucher, A., De, S., Smith, R.J., Shvets, I.V., Arora, S.K., Stanton, G., Kim, H.Y., Lee, K., Kim, G.T., Duesberg, G.S., Hallam, T., Boland, J.J., Wang, J.J., Donegan, J.F., Grunlan, J.C., Moriarty, G., Shmeliov, A., Nicholls, R.J., Perkins, J.M., Grieveson, E.M., Theuwissen, K., McComb, D.W., Nellist, P.D., and Nicolosi, V. (2011) Two-Dimensional Nanosheets Produced by Liquid Exfoliation of Layered Materials. Science, 331(6017), 568-571. Comans, R.N.J., Haller, M., and Depreter, P. (1991) Sorption of Cesium on Illite - Nonequilibrium Behavior and Reversibility. Geochimica Et Cosmochimica Acta, 55(2), 433-440. Dai, R., Liu, J., Yu, C., Sun, R., Lan, Y., and Mao, J.D. (2009) A comparative study of oxidation of Cr(III) in aqueous ions, complex ions and insoluble compounds by manganese-bearing mineral (birnessite). Chemosphere, 76, 536-541. Donne, S.W., Hollenkamp, A.F., and Jones, B.C. (2010) Structure, morphology and electrochemical behaviour of manganese oxides prepared by controlled decomposition of permanganate. Journal of Power Sources, 195(1), 367-373. Drits, V.A., Lanson, B., Bougerol-Chaillout, C., Gorshkov, A.I., and Manceau, A. (2002) Structure of heavy-metal sorbed birnessite: Part 2. Results from electron diffraction. American Mineralogist, 87(11-12), 1646-1661. Drits, V.A., Silvester, E., Gorshkov, A.I., and Manceau, A. (1997) Structure of synthetic monoclinic Na-rich birnessite and hexagonal birnessite.1. Results from X-ray diffraction and selected-area electron diffraction. American Mineralogist, 82(9- 10), 946-961. Dubinin, A., Uspenskaya, T., Gavrilenko, G., and Rashidov, V. (2008) Geochemistry and genesis of Fe-Mn mineralization in island arcs in the west Pacific Ocean. Geochemistry International, 46(12), 1206-1227. Duckworth, O.W., Bargar, J.R., and Sposito, G. (2008) Sorption of ferric iron from ferrioxamine B to synthetic and biogenic layer type manganese oxides. Geochimica Et Cosmochimica Acta, 72(14), 3371-3380. Duff, M.C., Hunter, D.B., Triay, I.R., Bertsch, P.M., Reed, D.T., Sutton, S.R., Shea- McCarthy, G., Kitten, J., Eng, P., Chipera, S.J., and Vaniman, D.T. (1999) Mineral associations and average oxidation states of sorbed Pu on tuff. Environmental Science & Technology, 33, 2163-2169. Dyer, A., Pillinger, M., Harjula, R., and Amin, S. (2000) Sorption characteristics of radionuclides on synthetic birnessite-type layered manganese oxides. Journal of Materials Chemistry, 10(8), 1867-1874. Edgemon, G.L., Anda, V.S., Barman, H.S., Johnson, M.E., and Boomer, K.D. (2009) History and Operation of the Hanford High-Level Waste Storage Tanks (Reprinted from Proceedings of the CORROSION/2008 Research Topical Symposium). Corrosion, 65(3), 163-174.
12 El-Kamash, A.M. (2008) Evaluation of zeolite A for the sorptive removal of Cs+ and Sr2(+) ions from aqueous solutions using batch and fixed bed column operations. Journal of Hazardous Materials, 151(2-3), 432-445. Feng, X.H., Zhai, L.M., Tan, W.F., Zhao, W., Liu, F., and He, J.Z. (2006) The controlling effect of pH on oxidation of Cr(III) by manganese oxide minerals. Journal of Colloid and Interface Science, 298(1), 258-266. Feng, X.H., Zhu, M.Q., Ginder-Vogel, M., Ni, C.Y., Parikh, S.J., and Sparks, D.L. (2010) Formation of nano-crystalline todorokite from biogenic Mn oxides. Geochimica Et Cosmochimica Acta, 74(11), 3232-3245. Fischer, T.B. (2011) Structural transformations of birnessite (delta-MnO2) during biological and abiological reduction. Geosciences, Doctorate of Philosophy. The Pennsylvania State University, University Park. Fischer, T.B., Heaney, P.J., Jang, J.H., Ross, D.E., Brantley, S.L., Post, J.E., and Tien, M. (2008) Continuous time-resolved X-ray diffraction of the biocatalyzed reduction of Mn oxide. American Mineralogist, 93(11-12), 1929-1932. Flury, M., Czigany, S., Chen, G., and Harsh, J.B. (2004) Cesium migration in saturated silica sand and Hanford sediments as impacted by ionic strength. Journal of Contaminant Hydrology, 71, 111-126. Fredrickson, J.K., Zachara, J.M., Kennedy, D.W., Kukkadapu, R.K., McKinley, J.P., Heald, S.M., Liu, C., and Plymale, A.E. (2004) Reduction of TcO4- by sediment- associated biogenic Fe(II). Geochimica et Cosmochimica Acta, 68(15), 3171- 3187. Frierdich, A.J., Hasenmueller, E.A., and Catalano, J.G. (2011) Composition and structure of nanocrystalline Fe and Mn oxide cave deposits: Implications for trace element mobility in karst systems. Chemical Geology, 284(1-2), 82-96. Fu, G., Allen, H.E., and Cowan, C.E. (1991) Adsorption of cadmium and copper by manganese oxide. Soil Science, 152(2), 72-81. Gee, G.W., Oostrom, M., Freshley, M.D., Rockhold, M.L., and Zachara, J.M. (2007) Hanford site vadose zone studies: An overview. Vadose Zone Journal, 6(4), 899- 905. Glover, E.D. (1977) Characterization of a Marine Birnessite. American Mineralogist, 62(3-4), 278-285. Golden, D.C., Dixon, J.B., and Chen, C.C. (1986) Ion exchange, thermal transformations, and oxidizing properties of birnessite. Clays and Clay Minerals, 34(5), 511-520. Grangeon, S., Lanson, B., Miyata, N., Tani, Y., and Manceau, A. (2010) Structure of nanocrystalline phyllomanganates produced by freshwater fungi. American Mineralogist, 95(11-12), 1608-1616. Jenne, E.A. (1968) Controls on Mn, Fe, Co, Ni, Cu, and Zn Concentrations in Soils and Water - Significant Role of Hydrous Mn and Fe Oxides. Advances in Chemistry Series(73), 337-387. Jeong, C.H., Kim, C.S., Kim, S.J., and Park, S.W. (1996) Affinity of radioactive cesium and strontium for illite and smectite clay in the presence of groundwater ions. Journal of Environmental Science and Health Part a-Environmental Science and Engineering & Toxic and Hazardous Substance Control, 31(9), 2173-2192.
13 Jones, L.H.P., and Milne, A.A. (1956) Birnessite, a new manganese oxide mineral from Aberdeenshire, Scotland. The Mineralogical Magazine, 31, 283-288. Jurgensen, A., Widmeyer, J.R., Gordon, R.A., Bendell-Young, L.I., and Moore, M.M. (2004) The structure of the manganese oxide on the sheath of the bacterium Leptothrix discophora: An XAFS study. American Mineralogist, 89(7), 1110- 1118. Kamimura, A., Nozaki, Y., Ishikawa, S., Inoue, R., and Nakayama, M. (2011) K- birnessite MnO2: a new selective oxidant for benzylic and allylic alcohols. Tetrahedron Letters, 52(4), 538-540. Kang, D.H., Schwab, A.P., Johnston, C.T., and Banks, M.K. (2010) Adsorption of iron cyanide complexes onto clay minerals, manganese oxide, and soil. Journal of Environmental Science and Health Part A-Toxic/Hazardous Substances & Environmental Engineering, 45(11), 1391-1396. Kwon, K.D., Refson, K., and Sposito, G. (2010) Surface complexation of Pb(II) by hexagonal birnessite nanoparticles. Geochimica Et Cosmochimica Acta, 74(23), 6731-6740. Lanson, B., Drits, V.A., Gaillot, A.C., Silvester, E., Plancon, A., and Manceau, A. (2002) Structure of heavy-metal sorbed birnessite: Part 1. Results from X-ray diffraction. American Mineralogist, 87(11-12), 1631-1645. Lanson, B., Drits, V.A., Silvester, E., and Manceau, A. (2000) Structure of H-exchanged hexagonal birnessite and its mechanism of formation from Na-rich monoclinic buserite at low pH. American Mineralogist, 85(5-6), 826-838. Lee, S., An, J., Kim, Y.J., and Nam, K. (2011) Binding strength-associated toxicity reduction by birnessite and hydroxyapatite in Pb and Cd contaminated sediments. Journal of Hazardous Materials, 186(2-3), 2117-2122. Lichtner, P.C., and Felmy, A.R. (2003) Estimation of Hanford SX tank waste compositions from historically derived inventories. Computers & geosciences, 29(3), 371-383. Lichtner, P.C., Yabusaki, S., Pruess, K., and Steefel, C.I. (2004) Role of competitive cation exchange on chromatographic displacement of cesium in the vadose zone beneath the Hanford S/SX tank farm. Vadose Zone Journal, 3(1), 203-219. Lind, C.J., and Hem, J.D. (1993) Manganese minerals and associated fine particulates in the streambed of Pinal Creek, Arizona, U.S.A.: a mining-related acid drainage problem. Applied Geochemistry, 8, 67-80. Lindsey, K.A., and Gaylord, D.R. (1990) Lithofacies and sedimentology of the Miocene- Pliocene Ringold Formation, Hanford Site, South-Central Washington. Northwest Science, 64(3), 165-180. Liu, C., Zachara, J.M., Qafoku, O., and Smith, S.C. (2003) Effect of temperature on Cs+ soprtion and desorption in subsurface sediments at the Hanford Site, U.S.A. Environmental Science & Technology, 37, 2640-2645. Liu, Z., Ooi, K., Kanoh, H., Tang, W., and Tomida, T. (2000) Swelling and delamination behaviors of birnessite-type manganese oxide by intercalation of tetraalkylammonium ions. Langmuir, 16(9), 4154-4164.
14 Liu, Z.H., Kang, L.P., Zhao, M.Z., and Ooi, K. (2007) Preparation, ion-exchange, and electrochemical behavior of Cs-type manganese oxides with a novel hexagonal- like morphology. Journal of Materials Research, 22(9), 2437-2447. Lopano, C.L., Heaney, P.J., Post, J.E., Hanson, J., and Komarneni, S. (2007) Time- resolved structural analysis of K- and Ba-exchange reactions with synthetic Na- birnessite using synchrotron X-ray diffraction. American Mineralogist, 92(2-3), 380-387. Manceau, A., Drits, V.A., Silvester, E., Bartoli, C., and Lanson, B. (1997) Stuctural mechanism of Co2+ oxidation by the phyllomanganate buserite. Mandernack, K.W., Post, J., and Tebo, B.M. (1995) Manganese Mineral Formation by Bacterial-Spores of the Marine Bacillus, Strain Sg-1 - Evidence for the Direct Oxidation of Mn(Ii) to Mn(Iv). Geochimica Et Cosmochimica Acta, 59(21), 4393-4408. Manolopoulou, M., Vagena, E., Stoulos, S., Ioannidou, A., and Papastefanou, C. (2011) Radioiodine and radiocesium in Thessaloniki, Northern Greece due to the Fukushima nuclear accident. Journal of Environmental Radioactivity, 102(8), 796-797. Mathews, S., Ramesh, R., Venkatesan, T., and Benedetto, J. (1997) Ferroelectric field effect transistor based on epitaxial perovskite heterostructures. Science, 276(5310), 238-240. McKenzie, R.M. (1977) Manganese Oxides and Hydroxides. In J.B. Dixon, and S.B. Weed, Eds. Minerals in Soil Environments, p. 181-193. Soil Science Society of America, Madison, WI. -. (1980) Adsorption of Lead and Other Heavy-Metals on Oxides of Manganese and Iron. Australian Journal of Soil Research, 18(1), 61-73. McKinley, J.P., Zeissler, C.J., Zachara, J.M., Serne, R.J., Lindstrom, R.M., Schaef, H.T., and Orr, R.D. (2001) Distribution and retention of Cs-137 in sediments at the Hanford Site, Washington. Environmental Science & Technology, 35(17), 3433- 3441. Meng, Y.T., Zheng, Y.M., Zhang, L.M., and He, J.Z. (2009) Biogenic Mn oxides for effective adsorption of Cd from aquatic environment. Environmental Pollution, 157(8-9), 2577-2583. Mikhail, E.M., and Misak, N.Z. (1988) Sorption of Cesium and Cobalt Radionuclides on a New Manganese Oxide. Applied Radiation and Isotopes, 39(11), 1121-1124. Mishra, S.P., and Vijaya. (2007) Removal behavior of hydrous manganese oxide and hydrous stannic oxide for Cs(I) ions from aqueous solutions. Separation and Purification Technology, 54(1), 10-17. Mon, J., Deng, Y., Flury, M., and Harsh, J.B. (2005) Cesium incorporation and diffusion in cancrinite, sodalite, zeolite, and allophane. Microporous and Mesoporous Materials, 86, 277-286. Negra, C., Ross, D.S., and Lanzirotti, A. (2005) Soil manganese oxides and trace metals: Competitive sorption and microfocused synchrotron X-ray fluorescence mapping. Soil Science Society of America Journal, 69(2), 353-361.
15 Novikov, G.V., Sval'nov, V.N., Bogdanova, O.Y., and Sivtsov, A.V. (2010) Ion exchange properties of manganese and iron minerals in oceanic micronodules. Lithology and Mineral Resources, 45(5), 410-424. O'Reilly, S.E., and Hochella, M.F. (2003) Lead sorption efficiencies of natural and synthetic Mn and Fe-oxides. Geochimica Et Cosmochimica Acta, 67(23), 4471- 4487. Onishi, Y., Kivva, S.L., Zheleznyak, M.J., and Voitsekhovich, O.V. (2007) Aquatic assessment of the chernobyl nuclear accident and its remediation. Journal of Environmental Engineering-Asce, 133(11), 1015-1023. Paterson, E., Clark, D.R., Russell, J.D., and Swaffield, R. (1986) Cation-Exchange in Synthetic Manganates .2. the Structure of an Alkylammonium-Saturated Phyllomanganate. Clay Minerals, 21(5), 957-964. Pena, J., Kwon, K.D., Refson, K., Bargar, J.R., and Sposito, G. (2010) Mechanisms of nickel sorption by a bacteriogenic birnessite. Geochimica et Cosmochimica Acta, 74, 3076-3089. Petkov, V., Ren, Y., Saratovsky, I., Pasten, P., Gurr, S.J., Hayward, M.A., Poeppelmeier, K.R., and Gaillard, J.F. (2009) Atomic-Scale Structure of Biogenic Materials by Total X-ray Diffraction: A Study of Bacterial and Fungal MnOx. Acs Nano, 3(2), 441-445. Post, J.E. (1999) Manganese oxide minerals: Crystal structures and economic and environmental significance. Proceedings of the National Academy of Sciences of the United States of America, 96(7), 3447-3454. Post, J.E., Heaney, P.J., and Hanson, J. (2002) Rietveld refinement of a triclinic structure for synthetic Na-birnessite using synchrotron powder diffraction data. Powder Diffraction, 17(3), 218-221. Redkin, A.F., and Hemley, J.J. (2000) Experimental Cs and Sr sorption on analcime in rock-buffered systems at 250-300 degrees C and P-sat and the thermodynamic evaluation of mineral solubilities and phase relations. European Journal of Mineralogy, 12(5), 999-1014. Reolid, M., El Kadiri, K., Abad, I., Oloriz, F., and Jimenez-Millan, J. (2011) Jurassic microbial communities in hydrothermal manganese crust of the Rifian Calcareous Chain, Northern Morocco. Sedimentary Geology, 233(1-4), 159-172. Rona, P.A., Harbison, R.N., Bassinger, B.G., Scott, R.B., and Nalwalk, A.J. (1976) Tectonic Fabric and Hydrothermal Activity of Mid-Atlantic Ridge Crest (Lat 26 Degrees). Geological Society of America Bulletin, 87(5), 661-674. Roque-Rosell, J., Mosselmans, J.F.W., Proenza, J.A., Labrador, M., Gali, S., Atkinson, K.D., and Quinn, P.D. (2010) Sorption of Ni by "lithiophorite-asbolane" intermediates in Moa Bay lateritic deposits, eastern Cuba. Chemical Geology, 275(1-2), 9-18. Saratovsky, I., Gurr, S.J., and Hayward, M.A. (2009) The Structure of manganese oxide formed by the fungus Acremonium sp strain KR21-2. Geochimica Et Cosmochimica Acta, 73(11), 3291-3300. Saratovsky, I., Wightman, P.G., Pasten, P.A., Gaillard, J.F., and Poeppelmeier, K.R. (2006) Manganese oxides: Parallels between abiotic and biotic structures. Journal of the American Chemical Society, 128(34), 11188-11198.
16 Sherman, D.M., and Peacock, C.L. (2010) Surface complexation of Cu on birnessite (delta-MnO2): Controls on Cu in the deep ocean. Geochimica Et Cosmochimica Acta, 74(23), 6721-6730. Shin, J.Y., Buzgo, C.M., and Cheney, M.A. (2000) Mechanochemical degradation of atrazine adsorbed on four synthetic manganese oxides. Colloids and Surfaces a- Physicochemical and Engineering Aspects, 172(1-3), 113-123. Silvester, E., Manceau, A., and Drits, V.A. (1997) Structure of synthetic monoclinic Na- rich birnessite and hexagonal birnessite. 2. Results from chemical studies and EXAFS spectroscopy. American Mineralogist, 82(9-10), 962-978. Singh, O.V., and Tandon, S.N. (1977) Studies on Adsorption of Cesium and Strontium Radionuclides on Hydrated Manganese Oxide. International Journal of Applied Radiation and Isotopes, 28(8), 701-704. Spiro, T.G., Bargar, J.R., Sposito, G., and Tebo, B.M. (2010) Bacteriogenic Manganese Oxides. Accounts of Chemical Research, 43(1), 2-9. Stahl, R.S., and James, B.R. (1991) Zinc Sorption by Manganese-Oxide-Coated Sand as a Function of Ph. Soil Science Society of America Journal, 55(5), 1291-1294. Takemura, T., Nakamura, H., Takigawa, M., Kondo, H., Satomura, T., Miyasaka, T., and Nakajima, T. (2011) A Numerical Simulation of Global Transport of Atmospheric Particles Emitted from the Fukushima Daiichi Nuclear Power Plant. Sola, 7, 101- 104. Tan, G.T., Dai, S.Y., Duan, P., Zhou, Y.L., Lu, H.B., and Chen, Z.H. (2003) Structural, electric and magnetic properties of the electron-doped manganese oxide: La1- xTexMnO3 (x=0.1, 0.15). Journal of Applied Physics, 93(9), 5480-5483. Tan, H., Zhang, G.X., Heaney, P.J., Webb, S.M., and Burgos, W.D. (2010) Characterization of manganese oxide precipitates from Appalachian coal mine drainage treatment systems. Applied Geochemistry, 25(3), 389-399. Taylor, R.M., and McKenzie, R.M. (1966) The association of trace elements with manganese minerals in Australian soils. Australian Journal of Soil Research, 4, 29-39. Taylor, R.M., McKenzie, R.M., and Norrish, K. (1964) The mineralogy and chemistry of manganese in some Australian soils. Australian Journal of Soil Research, 2, 235- 248. Tebo, B.M., Bargar, J.R., Clement, B.G., Dick, G.J., Murray, K.J., Parker, D., Verity, R., and Webb, S.M. (2004) Biogenic manganese oxides: Properties and mechanisms of formation. Annual Review of Earth and Planetary Sciences, 32, 287-328. Thompson, A., Steefel, C.I., Perdrial, N., and Chorover, J. (2010) Contaminant Desorption during Long-Term Leaching of Hydroxide-Weathered Hanford Sediments. Environmental Science & Technology, 44(6), 1992-1997. Toner, B., Manceau, A., Webb, S.M., and Sposito, G. (2006) Zinc sorption to biogenic hexagonal-birnessite particles within a hydrated bacterial biofilm. Geochimica Et Cosmochimica Acta, 70(1), 27-43. Vaniman, D.T., Chipera, S.J., Bish, D.L., Duff, M.C., and Hunter, D.B. (2002) Crystal chemistry of clay-Mn oxide associations in soils, fractures, and matrix of the Bandelier Tuff, Pajarito Mesa, New Mexico. Geochimica Et Cosmochimica Acta, 66(8), 1349-1374.
17 Villalobos, M., Bargar, J., and Sposito, G. (2005) Trace metal retention on biogenic manganese oxide nanoparticles. Elements, 1(4), 223-226. Villalobos, M., Lanson, B., Manceau, A., Toner, B., and Sposito, G. (2006) Structural model for the biogenic Mn oxide produced by Pseudomonas putida. American Mineralogist, 91(4), 489-502. Villalobos, M., Toner, B., Bargar, J., and Sposito, G. (2003) Characterization of the manganese oxide produced by Pseudomonas putida strain MnB1. Geochimica Et Cosmochimica Acta, 67(14), 2649-2662. Wan, J.M., Larsen, J.T., Tokunaga, T.K., and Zheng, Z.P. (2004) pH neutralization and zonation in alkaline-saline tank waste plumes. Environmental Science & Technology, 38(5), 1321-1329. Wang, F.L., Yu, T.R., Huang, P.M., Krishnamurti, G.S.R., Li, S.X., and Fairhurst, D. (1996) Competitive adsorption of potassium and calcium on manganese oxide as studied with two ion-selective electrodes. Zeitschrift Fur Pflanzenernahrung Und Bodenkunde, 159(1), 93-99. Wang, M.X., Wang, Y.P., Tan, W.F., Liu, F., Feng, X.H., and Koopal, L.K. (2010) Effect of 1-1 electrolyte concentration on the adsorption/desorption of copper ion on synthetic birnessite. Journal of Soils and Sediments, 10(5), 879-885. Wang, W.M., Shao, Z.Z., Liu, Y.J., and Wang, G.J. (2009) Removal of multi-heavy metals using biogenic manganese oxides generated by a deep-sea sedimentary bacterium - Brachybacterium sp strain Mn32. Microbiology-Sgm, 155, 1989- 1996. Webb, S.M., Tebo, B.M., and Bargar, J.R. (2005a) Structural characterization of biogenic Mn oxides produced in seawater by the marine bacillus sp strain SG-1. American Mineralogist, 90(8-9), 1342-1357. -. (2005b) Structural influences of sodium and calcium ions on the biogenic manganese oxides produced by the marine Bacillus sp., strain SG-1. Geomicrobiology Journal, 22(3-4), 181-193. White, W.B., Vito, C., and Scheetz, B.E. (2009) The Mineralogy and Trace Element Chemistry of Black Manganese Oxide Deposits from Caves. Journal of Cave and Karst Studies, 71(2), 136-143. Xu, H.F., Chen, T.H., and Konishi, H. (2010) HRTEM investigation of trilling todorokite and nano-phase Mn-oxides in manganese dendrites. American Mineralogist, 95(4), 556-562. Zachara, J.M., Serne, J., Freshley, M., Mann, F., Anderson, F., Wood, M., Jones, T., and Myers, D. (2007) Geochemical processes controlling migration of tank wastes in Hanford's vadose zone. Vadose Zone Journal, 6(4), 985-1003. Zhao, W., Feng, X., Tan, W., Liu, F., and Ding, S. (2009) Relation of lead adsorption on birnessites with different average oxidation states of manganese and release of Mn2+/H+/K+. Journal of Environmental Sciences, 21, 520-526. Zhao, W., Tan, W., Feng, X., Liu, F., Xie, Y., and Xie, Z. (2011) XAFS studies on surface coordination of Pb2+ on birnessites with different average oxidation states. Colloids and surfaces. A, Physicochemical and engineering aspects, 379(1- 3), 86-92.
18 Zhao, W., Wang, Q.Q., Liu, F., Qiu, G.H., Tan, W.F., and Feng, X.H. (2010) Pb2+ adsorption on birnessite affected by Zn2+ and Mn2+ pretreatments. Journal of Soils and Sediments, 10(5), 870-878. Zhu, M.Q., Ginder-Vogel, M., Parikh, S.J., Feng, X.H., and Sparks, D.L. (2010a) Cation Effects on the Layer Structure of Biogenic Mn-Oxides. Environmental Science & Technology, 44(12), 4465-4471. Zhu, M.Q., Ginder-Vogel, M., and Sparks, D.L. (2010b) Ni(II) Sorption on Biogenic Mn- Oxides with Varying Mn Octahedral Layer Structure. Environmental Science & Technology, 44(12), 4472-4478.
19 Figures
Figure 1.1. The crystal structure of triclinic Na-birnessite. The Mn-O octahedral sheets contain Mn3+/4+ in all octahedral positions with hydrated Na+ as interlayer cations (adapted from Post 2002).
20
Figure 1.2. The crystal structure of hexagonal H-birnessite with Mn4+-O octahedral + 2+ sheets and H /H2O and Mn as interlayer cations (adapted from Silvester et al. (1997)).
21 Chapter 2 Exchange of Cs into hexagonal H-birnessite as a function of pH
Abstract
The leakage of solutions containing high concentrations of radioactive Cs-137 into the subsurface threatens groundwater integrity at several nuclear waste and failed nuclear reactor sites (e.g., Hanford, Chernobyl, and Fukushima). In order to provide a quantitative measure of the uptake of radioactive Cs in soils that are rich in Mn oxides, we applied time-resolved synchrotron X-ray diffraction (TR-XRD), inductively coupled plasma-mass spectrometry (ICP-MS), and analytical scanning electron microscopy
(ASEM) to determine the kinetics and mechanisms of aqueous Cs substitution in the phyllomanganate hexagonal H-birnessite at pH values ranging from 3 to 10. These data demonstrate that the Cs+ exchange rate is weakly dependent on pH, but the amount of Cs+ that will exchange into the interlayer increases dramatically as pH increases. Our work supports delamination-reassembly as a mechanism of cation exchange.
Introduction
Cesium-137 is a byproduct of the plutonium production process, and it is a significant contaminant during nuclear fallout, nuclear plant meltdown, and nuclear waste storage. Following the nuclear plant disasters at Chernobyl (Soviet Union), and most recently, Fukushima (Japan), 137Cs was distributed widely through radioactive plumes that settled into soils and surface waters. Moreover, 137Cs is a major component of the
22 nuclear wastes monitored by the Department of Energy (DOE) at such sites as Hanford,
WA, Savannah River, SC, and Oak Ridge, TN.
At the Hanford site, 137Cs is responsible for roughly 40% of the radioactivity of the nuclear waste (McKinley et al. 2001; Gee et al. 2007). Hanford was a plutonium production factory from 1943 to 1991, and its high-level nuclear waste is stored in 177 underground single- and double-shell steel tanks. Approximately 3.8 million L of waste have leaked from the single-shell tanks, partly because the high alkalinity of the waste
(pH 13-14) corroded the steel. The leaking solutions are concentrated in radioactive
137Cs (2 x 1010 Bq/L, equivalent to 0.04 mmol/L), and they have contaminated approximately 28,300 m3 of soil (Gee et al. 2007). Although the contaminant solutions leaking from the tanks initially are at pH 14, the alkalinity is neutralized with increasing distance of transport as the solutions are buffered by soil minerals (Wan et al. 2004).
Consequently, in order to model the migration of Cs-rich fluids, it is necessary to study
Cs uptake by soil minerals over a range of pH.
Underlying the tank farm at Hanford is the Ringold Formation, which contains predominantly sand- and cobble-sized gravel with significant amounts of clay, silt, and sand (Lindsey and Gaylord 1990). These sediments are glacial-fluvial in origin and are coated in distinct iron and manganese oxides layers (Barnett et al. 2002; Fredrickson et al. 2004). These manganese oxides make up 0.4-0.93 wt% of the Ringold Formation
(Barnett et al. 2002; Fredrickson et al. 2004). Despite their minor abundances, these phases play a major role in controlling heavy metal mobility. As coatings on glacial clasts at Hanford, they occur at the interface between pore solutions and the primary soil minerals. Moreover, the high cation exchange, redox, and adsorption capacities of Mn
23 oxides are well documented (Fu et al. 1991; Violante and Pigna 2002; O'Reilly and
Hochella 2003; Weaver and Hochella 2003; Negra et al. 2005; Zhao et al. 2009).
Hydrous manganese oxides with open structures are most frequently observed in these soils, such as the phyllomanganates birnessite and ranciéite, and the tunneled manganate todorokite (Taylor et al. 1964; Vaniman et al. 2002; Fredrickson et al. 2004). All of these phases occur as submicron grains with large surface areas, but birnessite is typically the most reactive.
In this study, we have investigated the interactions of aqueous Cs with a synthetic hexagonal H-birnessite. Birnessite structures have been shown to exhibit either triclinic
(Lanson et al. 2002; Post et al. 2002) or hexagonal (Silvester et al. 1997) symmetry. The structure of hexagonal birnessite differs from that of triclinic birnessite in the near- absence of Mn3+ and the presence of Mn vacancies (□) in the octahedral sheet, resulting
3+ 2+ 4+ 3+ in the following chemical formula: H0.33Mn 0.111Mn 0.055(Mn 0.722Mn 0.111□0.167)O2
(Fig. 2.1) (Silvester et al. 1997). The charge deficits incurred by the vacancies and Mn3+ in place of octahedral Mn4+ result in negatively charged sheets, which are electrostatically balanced by Mn2+ and hydrated cations between the sheets.
To the best of our knowledge, most prior studies of cation exchange in the phyllomanganate system have focused on triclinic birnessite (e.g., Mikhail and Misak
1988; Dyer et al. 2000; Al-Attar et al. 2003; Liu et al. 2007; Lopano et al. 2007; Lopano et al. 2009). However, many scientists have demonstrated that natural soils contain both triclinic and hexagonal birnessite (Webb et al. 2005; Post et al. 2008; Tan et al. 2010;
Santelli et al. 2011). In light of the importance of the hexagonal varieties of birnessite in
24 soil systems, we report the effects of pH on Cs cation exchange into hexagonal H- birnessite.
Experimental Methods
Material synthesis
Triclinic Na-birnessite was initially synthesized according to the procedure described in Golden et al. (1986). At room temperature, oxygen was bubbled (4.2 L/min) through a mixture of chilled 250 mL 5.5 M NaOH and 200 mL 0.5 M MnCl2 for five hours. The precipitate was divided evenly among 12 centrifuge tubes and centrifuged.
The solution was poured off and replaced with fresh DI water; after five replicate rinsing cycles, the birnessite was stored in fresh DI water until needed for experiments. For experiments, aliquots of the stored triclinic Na-birnessite were filtered through a μm hydrophilic polypropylene membrane filter (GH Polypro, Pall) and allowed to air dry.
Dry triclinic Na-birnessite (250 mg) was ground under acetone in an agate mortar to disaggregate clumps. After the acetone had fully evaporated, the powder was reacted with 250 mL 0.001 M HCl for three hours. The resultant synthetic hexagonal H- birnessite was filtered and then rinsed with 300 mL DI water, filtered again through a 0.2
μm filter, and allowed to air dry The material was initially characterized with a conventional sealed Mo tube source on a Rigaku II D/MAX-RAPID microdiffractometer
(Materials Characterization Laboratory, Pennsylvania State University).
25 Chemical analysis
Time-resolved X-ray diffraction (TR-XRD) experiments were completed at beam line 13-BM-C at the Advanced Photon Source (APS) at Argonne National Laboratory
(ANL). Approximately 4 mg of hexagonal H-birnessite were packed between two cotton plugs in a 0.7 mm quartz capillary and attached to a flow-through apparatus based on the design of Wall et al. (2011) (Fig. 2.2). The five Cs-rich cation exchange solutions had the same initial Cs concentration of 0.001 M, and pH ranged from 3 to 10 (pH 3, 5, 6.5, 9,
10). The pH was adjusted with CsOH, and the cesium concentration was controlled with
CsCl. Experiments were not completed at pH 13 because increasing the alkalinity using
CsOH led to a Cs concentration of 0.1 M. The solution flow rate was 1 drop/min (~ 0.05 mL/min). TR-XRD patterns were collected every 60 seconds for 5 to 9 hours using a
MAR165 CCD detector. The sample to collector distance was 97.6 mm, allowing us to collect over a θ range of -50°. Rietveld refinements (Rietveld 1969) were completed for each TR-XRD pattern using the EXPGUI interface (Toby 2001) of the general structure analysis system (GSAS) (Larson and Von Dreele 1994). Starting structures were drawn from the model structure of hexagonal H-birnessite proposed by Lanson et al.
(2000) with atom positions determined by Heaney et al. (2003). The background intensities for the TR-XRD patterns were fit using up to 15 terms of a linear interpolation function. The peak profiles were modeled by a pseudo-Voigt profile function as parameterized by Thompson et al. (1987), with asymmetry correction by Finger et al.
(1994), and microstrain anisotropic broadening terms of Stephens (1999).
26 During initial cycles of refinement only the background, scale, peak profile, and unit-cell parameters were allowed to vary. The position of the O atom in the Mn-O sheet was then refined. Following refinements of the interlayer atom positions, occupancy factors and isotropic atomic displacement factors for the interlayer sites were allowed to vary. Once the interlayer atoms were settled, the isotropic thermal parameters for the Mn and O in the octahedral sheets, followed alternately with those for the interlayer atoms, were refined The final χ2 values for end-member hexagonal Cs-birnessite ranged from
2.55 to 3.83 (Table 2.1). Difference-electron Fourier (DELF) maps were calculated to determine whether the combined O (water) and Cs interlayer atom positions changed over the course of the reaction and whether the atom positions were dependent upon the solution pH.
The leachate was collected as a function of time: a stepping motor controlled a
UR-150 Newport Rotary Stage with an Al tray capable of supporting 15 wide-mouth 15 mL polypropylene vials. The stepping motor was programmed to rotate to the next vial every 20 min. The leachate was analyzed with an X-Series 2 Thermo Scientific quadrupole inductively coupled plasma-mass spectrometer (ICP-MS) in conjunction with
Thermo Scientific PlasmaLab software (Materials Characterization Laboratory,
Pennsylvania State University) to determine the change in Cs and Mn elemental concentrations.
An FEI Nova NanoSEM 600 analytical scanning electron microscope (ASEM) equipped with Thermo Fisher Scientific NSS 2.3.89 software (Mineral Sciences
Analytical Laboratories, Smithsonian Institution) was used to determine the distribution of Cs across partially exchanged hexagonal H-birnessite grains. These partially
27 exchanged products were obtained by flowing 0.001 M CsCl solutions through hexagonal
H-birnessite powders in capillaries for 40 min at pH 9. The birnessite was unpacked from the glass capillary, rinsed with 300 mL DI water, and allowed to air dry at room temperature. Dry samples were mounted on SEM stubs with carbon tape for elemental mapping analysis.
Kinetic analysis
We applied the approach of Lopano et al. (2011) to determine the kinetics of cation substitution into the hexagonal H-birnessite structure. This method assumes that cation exchange in this system obeys Vegard’s Law, which posits that the volume of the unit cell varies linearly with the extent of cation exchange (Denton and Ashcroft 1991).
As suggested in Lopano et al. (2009), this presumption is reasonable, since the large size of Cs+ will expand the interlayer of birnessite-like phases. However, the application of
Vegard’s Law to birnessite is complicated by the appearance of an intermediate product, in which Cs has fully exchanged for the interlayer cation but is positionally disordered within the interlayer. The authors treated this issue by including the disordered state as a distinct phase that contributed to the overall unit-cell volume. Following their lead, we propose that the measured volume of the unit cell (Vtot) at any point during the exchange of Cs into hexagonal H-birnessite can be represented by a summation of the three individual phase fractions (Xx), each multiplied by their respective unit-cell volumes, q, r, and s:
(2.1)
28 where XA is the phase fraction of initial hexagonal H-birnessite, XB is the phase fraction of disordered hexagonal Cs-birnessite, and XC is the phase fraction of ordered hexagonal
Cs-birnessite.
As shown in Lopano et al. (2011), the kinetics of Cs+ exchange into triclinic Na- birnessite by extension can be modeled as a two-stage linear reaction. The first stage involves the complete exchange of cations, and in the second stage cations reorder within the interlayer to more thermodynamically stable positions (Fig. 2.3). The rate constant of
-1 each reaction stage (k1 and k2, respectively, with units min ) was determined according to standard kinetic equations for a first-order compound reaction (Fromherz 1964):
A (2.2)
- e (2.3)
- - (e -e ) (2.4) -
- - e e - - . (2.5) -
where t is time and is the initial phase fraction of starting material, A. According to mass balance:
(2.6) or
. (2.7)
Eqns. 2.3- were substituted into Vegard’s Law (Eqn ):
e- e- - - - e e -e - - . (2.8) - -
29 The dependence of volume versus time for each pH was fit with Eqn. 2.8 using Origin
6.1, solving for q, r, s, k1, and k2. These five variables were allowed to iterate simultaneously using the Levenberg-Marquardt algorithm until the reduced χ2 was minimized. For this study, we were most interested in the initial exchange of cations when XA = 1 and t = 0, and not the reordering of cations; therefore, the initial rate of the cation exchange was determined using the first-stage rate constant (k1) and phase fractions XA and XB. The initial rate in terms of XA, XB, and k1 can be defined with standard first-order reversible kinetics as (Lasaga 1981)
- (2.9)
(2.10)
where n is the reaction order Taking the derivative of Vegard’s Law (Eqn ) with respect to time,
, (2.11)
and substituting Eqn. 2.9 and 2.10 into Eqn. 2.11, the initial rate in is min
lim - (2.12) , or
- . (2.13)
When we plotted the log of the initial rate as a function of pH, we could calculate a linear regression for which the negative slope was the reaction order, n, and the y-intercept was
the log of the initial rate constant, w, with units M min :
30
log - log , (2.14)
which simplifies to
. (2.15)
Thus, Eqn. 2.15 allows us to determine the effect of solution pH on the rate of Cs cation exchange as measured from volume changes of the hexagonal birnessite unit cell.
Results
The substitution of Cs+ into hexagonal H-birnessite is evident from X-ray diffraction patterns through a marked decrease in the ratio of the 001 to 002 diffraction peaks (Fig. 2.4). Cs has a larger scattering factor than the exchangeable proton/water complex in the interlayer of hexagonal H-birnessite, causing a reduction in the peak intensity ratio between the (001) and (002) peaks from roughly 4:1 to 2:1 (Lopano et al.
2009). Rietveld refinement results for the exchange of Cs into hexagonal H-birnessite at pH 3, 5, 6.5, 9, and 10 are compiled in Table 2.1. Atom positions are listed in Table 2.2, and selected bond distances are listed in Table 2.3. The changes in unit-cell parameters a and c as a function of time and pH are recorded in Figures 2.5 and 2.6. All unit-cell parameters increase with extent of reaction, resulting in an increase of unit-cell volume with time. In order to determine the rate constant associated with the first stage of exchange (k1), we fit the unit-cell volume as a function of time using Eqn. 2.8 (Fig. 2.7).
To determine the initial rate of cation exchange for a given pH, we applied Eqn. 2.13, and the results of this analysis are presented in Table 2.4.
31 The dependence of the initial reaction rate on pH (Eqn. 2.14, Fig. 2.8) refined as
log - – R2 = 0.894, (2.16)
which can be simplified to an exponential form where the initial rate constant, w, is the coefficient and the reaction order, n, is the exponent:
. (2.17)
As can be seen in Figure 2.8, the exchange rate of Cs+ into the interlayer of hexagonal H- birnessite decreased only slightly as the pH increased. By extension, the reaction order of the cation exchange was also extremely small (0.079), indicating a weak dependence of exchange rate on pH.
The change in final unit-cell volume increased as pH increased (Fig. 2.9), particularly as pH exceeded 6. To the extent that unit-cell volume is a proxy for Cs+ exchange, we can infer that more Cs+ is exchanged into the interlayer of hexagonal H- birnessite with increasing alkalinity. We tested this conclusion by ICP-MS analysis of Cs and Mn in the leachate, which was collected as a function of time after it interacted with the hexagonal H-birnessite. With the initial concentration of Cs in the exchange solutions equal to 0.001 M Cs (or 133 ppm), it can be seen from Table 2.5 that all of the Cs was sequestered by the birnessite powder in the first reaction stages when pH was high. We reconfigured the results in Table 2.5 to represent the percentage of Cs removed from or released to the flow-through solution as a function of time, and these analyses are included in Fig. 2.10. Positive data points indicate sequestration by the solid, whereas negative data points represent release of Cs by the solid. As can be seen in this figure, only 25 wt% of the Cs in the influent was adsorbed within the first 20 minutes at pH 3,
32 whereas 100 wt% of the Cs was sequestered by the birnessite over the first 100 min at pH
10. For all pH values, the sequestration of Cs eventually was followed by the release of
Cs to the solution.
Discussion
Controls of Cs uptake
We postulate that the rate and amount of Cs+ uptake as a function of pH are controlled by two factors: 1) the charge on the octahedral sheets; and 2) the presence of competing cations (including H+) in solution. The point-of-zero charge (PZC) of birnessite is approximately 3.0 (Tan et al. 2008), and the octahedral MnO6 sheets thus become more negatively charged as pH increases beyond 3 (Murray 1974; McKenzie
1981). In order to maintain neutrality of the surface charge at high pH, more Cs must adsorb to surface sites. Thus, the change in surface charge as a function of pH may explain the increased uptake of Cs in more alkaline solutions.
Secondly, Cs exchange at low pH is not chemically favorable because, according to Le hâtelier’s principle, the excess number of protons in the starting solution drives the reaction in the reverse direction and prevents Cs cation exchange:
+ + H-birnessite(s) + Cs (aq) s -birnessite(s) + H (aq). (2.18)
Finally, acidic dissolution occurs at a pH as low as 3 (Murray 1974; Eary and Ral 1987),
2+ and that may inhibit Cs from entering the interlayer due to Mn (aq) production, where
+ 2+ MnO2(s) + 2H (aq) Mn (aq) + H2O(l) + ½O2(g). (2.19)
33 Measurements of the dissolved Mn in the leachate solutions from our experiments suggest that some dissolution did occur in the earliest stages of reaction (Fig. 2.11). At each pH, a small amount of Mn desorbed or dissolved from the hexagonal H-birnessite, with the greatest measured value (5 ppm after 20 min) occurring at pH 6.5. At pH 3 (Fig.
2.11A), the initial release of Mn reached only 1 ppm, but unlike reactions at higher pH, the concentration of Mn in the leachate achieved a non-zero steady state concentration of approximately 200 ppb for the remainder of the reaction. This observation supports our inference that reductive dissolution of hexagonal H-birnessite occurs at low pH as Cs is exchanged into the interlayer.
Transformation from hexagonal to triclinic symmetry
Although we successfully refined all of the exchanged Cs-birnessite structures within the hexagonal P -3 space group, we nevertheless speculate that the higher loading of Cs within birnessite at higher pH was promoted by a transformation towards triclinic symmetry. At pH 9 and 10, the occupancy of octahedral Mn increased from approximately 0.7 to 1.0 (Fig. 2.12), indicating that the concentration of octahedral vacancies dropped to zero with higher pH. As described in Silvester et al. (1997) and
Post et al. (2008), vacancies in the octahedral sheet are a defining feature of hexagonal birnessite, whereas the absence of these vacancies is more representative of triclinic birnessite (Post et al. 2002).
Another indication that a symmetry transition was initiated at higher pH can be seen in the sharp increase in the change of unit-cell volume above pH 6 (Fig. 2.9). The
34 unit-cell volume of triclinic M-birnessite is greater than that of hexagonal M-birnessite, and thus a hexagonal-triclinic phase transition would be accompanied by a step increase in unit-cell volume. Indeed, when the unit-cell volume behaviors are considered for each pH (Fig. 2.7), the low pH reactions exhibited a different character than those at high pH.
For pH 6.5 and lower, one can clearly detect the two stages of reaction as modeled by
Lopano et al. (2009) as two distinct regions with different slopes for dV/dt. Like Lopano et al. (2009), we interpret the abrupt increase in unit-cell volume in the first stage as representing the full exchange of Cs+ for H+; the comparatively lengthy second stage of minor volume increase reflects a positional ordering of Cs+ within the interlayer. At pH 9 and 10, however, the dependence of the volume change is less easily demarcated into two regimes, perhaps because the exchange of Cs+ continued long after the time that exchange had ceased in the experiments at low pH. Our analyses of the eluate strongly support this conclusion (Fig. 2.10). Consequently, the two-stages observed at low pH were blurred at high pH by the expansion of the interlayer and increased Cs+ uptake.
Although the structures at high pH may have been transforming towards triclinic birnessite, over the time-scale of these experiments (~ 400 min), the transition did not achieve completion, even at pH 10. Rietveld analysis of the diffraction patterns did not justify a lower symmetry for the higher pH products, and difference Fourier syntheses of the interlayers showed no evidence that the positions of the interlayer cations changed as the reaction proceeded (Fig. 2.13).
35 Back reaction
The release of Cs following its uptake was observed at each pH, and that phenomenon requires some explanation. If Cs leached out of the interlayer, we would expect to see the unit-cell volume decrease towards the initial volume of hexagonal H- birnessite. However, in all instances the unit-cell volume remained constant once cation exchange was complete, indicating that Cs remained in the interlayer. Rather, we speculate that Cs was released from edge or frayed sites on the birnessite platelet surfaces, or from the reordering of interlayer cations during stage two of the reaction.
Mechanism of exchange
Cation exchange in layered structures conventionally has been explained through simple diffusion of solvated ions into the solid accompanied by counter-diffusion of cations from the solid to the solution (Helfferich 1962). If this mechanism is operative when Cs+ exchanges into hexagonal H-birnessite, then a Cs-rich reaction rim would form around the edges of each partially-exchanged birnessite grain as the Cs migrated from the edges of the grain towards the core until exchange was complete. Alternatively, Putnis
(2002) has demonstrated that cation exchange commonly involves dissolution of the starting structure followed by reprecipitation of the final structure. This model also may create reaction rims, if the reprecipitation is pseudomorphic (Labotka et al. 2004), or it may be marked by the transient co-existence of endmember phases. In the present experiment, this mechanism would generate the growth of hexagonal Cs-birnessite particles at the expense of hexagonal H-birnessite.
36 The octahedral sheets in birnessite have the capacity to separate into monolayers when intercalated with organic ions and washed, and the monolayers then reconstruct as stacked crystals upon drying (Liu et al. 2000; Yang et al. 2004). By extension, Lopano et al ( ) proposed a “delamination-reassembly” mechanism for the substitution of s into triclinic Na-birnessite, and this exchange model represents a hybrid between simple diffusion and dissolution-reprecipitation. If this model is applied to the present system, adjacent octahedral sheets delaminated, allowing Cs cations to exchange with all of the protons within a single interlayer. Cation exchange then proceeded through a layer-by- layer substitution until all protons swapped out for Cs.
If this mechanism occurred during our experiments, then elemental maps of individual platelets of partially exchanged H-birnessite crystals would not exhibit a reaction rim. Instead, they would display a homogeneous distribution of Cs across a given grain. As can be seen in our ASEM elemental maps of partially exchanged hexagonal H-birnessite at pH 9 (Fig. 2.14), no Cs-rich reaction rims enveloped individual grains of birnessite. Consequently, we rule out diffusion as the exchange mechanism.
Likewise, we do not believe that dissolution-reprecipitation was the exchange mechanism because there were no Cs-rich grains interspersed with Cs-depleted grains. Rather, the homogeneous distribution of Cs supports delamination-reassembly as the driving mechanism for exchange.
In summary, we observed that the rate of Cs cation exchange into hexagonal H- birnessite exhibited only a weak dependence on pH, but the total amount of Cs loading in the interlayer region increased dramatically above pH 6.5. The increase in Cs content at a higher pH may be attributed to the increasingly negative charge on the Mn-O octahedral
37 sheets and perhaps to a structural change towards triclinic symmetry with high pH, while
Mn acidic dissolution at low pH may have inhibited Cs sequestration. The cation exchange reaction occurred by a delamination-reassembly mechanism, shown by a uniform distribution of Cs across partially exchanged grains. Thus, time-resolved XRD studies of Cs cation exchange has provided both kinetic and mechanistic constraints on the uptake of Cs by natural birnessite in soil systems such as those beneath the Hanford site high-level nuclear waste storage tanks.
Acknowledgements
Special thanks to Peter Eng and Nancy Lazarz at APS. Funding for this research was provided by the NSF-DOE Center for Environmental Kinetics Analysis (CEKA) Grant
No. CHE-0431328 and NSF Grant No. EAR07-45374. Use of the Advanced Photon
Source was supported by the U. S. Department of Energy, Office of Science, Office of
Basic Energy Sciences, under Contract No. DE-AC02-06CH11357.
References
Al-Attar, L., Dyer, A., Paajanen, A., and Harjula, R. (2003) Purification of nuclear wastes by novel inorganic ion exchangers. Journal of Materials Chemistry, 13(12), 2969- 2974. Barnett, M.O., Jardine, P.M., and Brooks, S.C. (2002) U(VI) adsorption to heterogeneous subsurface media: Application of a surface complexation model. Environmental Science & Technology, 36(5), 937-942. Denton, A.R., and Ashcroft, N.W. (1991) Vegard's law. Physical Review A, 43(6), 3161. Dyer, A., Pillinger, M., Harjula, R., and Amin, S. (2000) Sorption characteristics of radionuclides on synthetic birnessite-type layered manganese oxides. Journal of Materials Chemistry, 10(8), 1867-1874.
38 Eary, L.E., and Ral, D. (1987) Kinetics of chromium (III) oxidation to chromium (VI) by reaction with manganese dioxide. Environmental Science & Technology, 21, 1187-1193. Finger, L.W., Cox, D.E., and Jephcoat, A.P. (1994) A Correction for Powder Diffraction Peak Asymmetry Due to Axial Divergence. Journal of Applied Crystallography, 27, 892-900. Fredrickson, J.K., Zachara, J.M., Kennedy, D.W., Kukkadapu, R.K., McKinley, J.P., Heald, S.M., Liu, C., and Plymale, A.E. (2004) Reduction of TcO4- by sediment- associated biogenic Fe(II). Geochimica et Cosmochimica Acta, 68(15), 3171- 3187. Fromherz, H. (1964) Physico-chemical calculations in science and industry. Butterworth & Co., London. Fu, G., Allen, H.E., and Cowan, C.E. (1991) Adsorption of cadmium and copper by manganese oxide. Soil Science, 152(2), 72-81. Gee, G.W., Oostrom, M., Freshley, M.D., Rockhold, M.L., and Zachara, J.M. (2007) Hanford site vadose zone studies: An overview. Vadose Zone Journal, 6(4), 899- 905. Golden, D.C., Chen, C.C., and Dixon, J.B. (1986) Synthesis of Todorokite. Science, 231(4739), 717-719. Heaney, P.J., Post, J.E., Lopano, C.L., and Hanson, J.C. (2003) Hydrogen exchange in Na-birnessite: A time-resolved synchrotron X-ray diffraction analysis. Geological Society of America Abstracts with Programs, 35, 620. Helfferich, F. (1962) Ion Exchange. 624 p. McGraw-Hill, New York. Labotka, T.C., Cole, D.R., Fayek, M., Riciputi, L.R., and Stadermann, F.J. (2004) Coupled cation and oxygen-isotope exchange between alkali feldspar and aqueous chloride solution. American Mineralogist, 89(11-12), 1822-1825. Lanson, B., Drits, V.A., Feng, Q., and Manceau, A. (2002) Structure of synthetic Na- birnessite: Evidence for a triclinic one-layer unit cell. American Mineralogist, 87(11-12), 1662-1671. Lanson, B., Drits, V.A., Silvester, E., and Manceau, A. (2000) Structure of H-exchanged hexagonal birnessite and its mechanism of formation from Na-rich monoclinic buserite at low pH. American Mineralogist, 85(5-6), 826-838. Larson, A.C., and Von Dreele, R.B., 1994. GSAS-General Structure Analysis System. Los Alamos National Laboratory, Lasaga, A.C. (1981) Rate Laws of Chemical Reactions. In A.C. Lasaga, and R.J. Kirkpatrick, Eds. Kinetics of Geochemical Processes, 8, p. 1-68. Mineralogical Society of America, Chelsea, MI. Lindsey, K.A., and Gaylord, D.R. (1990) Lithofacies and sedimentology of the Miocene- Pliocene Ringold Formation, Hanford Site, South-Central Washington. Northwest Science, 64(3), 165-180. Liu, Z., Ooi, K., Kanoh, H., Tang, W., and Tomida, T. (2000) Swelling and delamination behaviors of birnessite-type manganese oxide by intercalation of tetraalkylammonium ions. Langmuir, 16(9), 4154-4164.
39 Liu, Z.H., Kang, L.P., Zhao, M.Z., and Ooi, K. (2007) Preparation, ion-exchange, and electrochemical behavior of Cs-type manganese oxides with a novel hexagonal- like morphology. Journal of Materials Research, 22(9), 2437-2447. Lopano, C.L., Heaney, P.J., Bandstra, J.Z., Post, J.E., and Brantley, S.L. (2011) Kinetic analysis of cation exchange in birnessite using time-resolved synchrotron X-ray diffraction. Geochimica Et Cosmochimica Acta, 75(14), 3973-3981. Lopano, C.L., Heaney, P.J., and Post, J.E. (2009) Cs-exchange in birnessite: Reaction mechanisms inferred from time-resolved X-ray diffraction and transmission electron microscopy. American Mineralogist, 94(5-6), 816-826. Lopano, C.L., Heaney, P.J., Post, J.E., Hanson, J., and Komarneni, S. (2007) Time- resolved structural analysis of K- and Ba-exchange reactions with synthetic Na- birnessite using synchrotron X-ray diffraction. American Mineralogist, 92(2-3), 380-387. McKenzie, R.M. (1981) The surface charge on manganese dioxides. Australian Journal of Soil Research, 19, 41-50. McKinley, J.P., Zeissler, C.J., Zachara, J.M., Serne, R.J., Lindstrom, R.M., Schaef, H.T., and Orr, R.D. (2001) Distribution and retention of Cs-137 in sediments at the Hanford Site, Washington. Environmental Science & Technology, 35(17), 3433- 3441. Mikhail, E.M., and Misak, N.Z. (1988) Sorption of Cesium and Cobalt Radionuclides on a New Manganese Oxide. Applied Radiation and Isotopes, 39(11), 1121-1124. Murray, J.W. (1974) The surface chemistry of hydrous manganese dioxide. Journal of Colloid and Interface Science, 46(3), 357-371. Negra, C., Ross, D.S., and Lanzirotti, A. (2005) Soil manganese oxides and trace metals: Competitive sorption and microfocused synchrotron X-ray fluorescence mapping. Soil Science Society of America Journal, 69(2), 353-361. O'Reilly, S.E., and Hochella, M.F. (2003) Lead sorption efficiencies of natural and synthetic Mn and Fe-oxides. Geochimica Et Cosmochimica Acta, 67(23), 4471- 4487. Post, J.E., Heaney, P.J., and Ertl, A. (2008) Rietveld refinement of the rancieite structure using synchrotron powder diffraction data. Powder Diffraction, 23(1), 10-14. Post, J.E., Heaney, P.J., and Hanson, J. (2002) Rietveld refinement of a triclinic structure for synthetic Na-birnessite using synchrotron powder diffraction data. Powder Diffraction, 17(3), 218-221. Putnis, A. (2002) Mineral replacement reactions: from macroscopic observations to microscopic mechanisms. Mineralogical Magazine, 66(5), 689-708. Rietveld, H.M. (1969) A Profile Refinement Method for Nuclear and Magnetic Structures. Journal of Applied Crystallography, 2, 65-71. Santelli, C.M., Webb, S.M., Dohnalkova, A.C., and Hansel, C.M. (2011) Diversity of Mn oxides produced by Mn(II)-oxidizing fungi. Geochimica Et Cosmochimica Acta, 75(10), 2762-2776. Silvester, E., Manceau, A., and Drits, V.A. (1997) Structure of synthetic monoclinic Na- rich birnessite and hexagonal birnessite. 2. Results from chemical studies and EXAFS spectroscopy. American Mineralogist, 82(9-10), 962-978.
40 Stephens, P.W. (1999) Phenomenological model of anisotropic peak broadening in powder diffraction. Journal of Applied Crystallography, 32, 281-289. Tan, H., Zhang, G.X., Heaney, P.J., Webb, S.M., and Burgos, W.D. (2010) Characterization of manganese oxide precipitates from Appalachian coal mine drainage treatment systems. Applied Geochemistry, 25(3), 389-399. Tan, W.F., Lu, S.J., Liu, F., Feng, X.H., He, J.Z., and Koopall, L.K. (2008) Determination of the point-of-zero, charge of manganese oxides with different methods including an improved salt titration method. Soil Science, 173(4), 277- 286. Taylor, R.M., McKenzie, R.M., and Norrish, K. (1964) The mineralogy and chemistry of manganese in some Australian soils. Australian Journal of Soil Research, 2, 235- 248. Thompson, P., Cox, D.E., and Hastings, J.B. (1987) Rietveld Refinement of Debye- Scherrer Synchrotron X-Ray Data from Al2O3. Journal of Applied Crystallography, 20, 79-83. Toby, B.H. (2001) EXPGUI, a graphical user interface for GSAS. Journal of Applied Crystallography, 34, 210-213. Vaniman, D.T., Chipera, S.J., Bish, D.L., Duff, M.C., and Hunter, D.B. (2002) Crystal chemistry of clay-Mn oxide associations in soils, fractures, and matrix of the Bandelier Tuff, Pajarito Mesa, New Mexico. Geochimica Et Cosmochimica Acta, 66(8), 1349-1374. Violante, A., and Pigna, M. (2002) Competitive sorption of arsenate and phosphate on different clay minerals and soils. Soil Science Society of America Journal, 66(6), 1788-1796. Wall, A.J., Heaney, P.J., Mathur, R., Post, J.E., Hanson, J.C., and Eng, P.J. (2011) A flow-through reaction cell that couples time-resolved X-ray diffraction with stable isotope analysis. Journal of Applied Crystallography, 44, 429-432. Wan, J.M., Larsen, J.T., Tokunaga, T.K., and Zheng, Z.P. (2004) pH neutralization and zonation in alkaline-saline tank waste plumes. Environmental Science & Technology, 38(5), 1321-1329. Weaver, R.M., and Hochella, M.F. (2003) The reactivity of seven Mn-oxides with Cr- aq(3+): A comparative analysis of a complex, aq environmentally important redox reaction. American Mineralogist, 88(11-12), 2016-2027. Webb, S.M., Tebo, B.M., and Bargar, J.R. (2005) Structural characterization of biogenic Mn oxides produced in seawater by the marine bacillus sp strain SG-1. American Mineralogist, 90(8-9), 1342-1357. Yang, X.J., Makita, Y., Liu, Z.H., Sakane, K., and Ooi, K. (2004) Structural characterization of self-assembled MnO2 nanosheets from birnessite manganese oxide single crystals. Chemistry of Materials, 16(26), 5581-5588. Zhao, W., Feng, X., Tan, W., Liu, F., and Ding, S. (2009) Relation of lead adsorption on birnessites with different average oxidation states of manganese and release of Mn2+/H+/K+. Journal of Environmental Sciences, 21, 520-526.
41 Figures
Figure 2.1. The crystal structure of hexagonal H-birnessite with Mn4+-O octahedral + 2+ sheets and H /H2O and Mn as interlayer cations (adapted from Silvester et al. (1997)).
42
Figure 2.2. Flow-through TR-XRD experiment setup at beam line 13-BM-C at APS.
43
Figure 2.3. A schematic of the two stage cation exchange reaction: stage one is the complete exchange of cations while stage two is the reordering of cations within the interlayer.
44
Figure 2.4. Stacked TR-XRD patterns of hexagonal H-birnessite exchanged with 0.001 M Cs+ at pH , with θ (°) along the x-axis, intensity along the y-axis, and time (min) along the z-axis where each pattern represents 5 min in time.
45
Figure 2.5. Unit cell parameters a (Å) are plotted versus time (min) as a function of pH: (A) pH 3, (B) pH 5, (C) pH 6.5, (D) pH 9, and (E) pH 10.
46
Figure 2.6. Unit cell parameter c (Å) plotted versus time (min) at (A) pH 3, (B) pH 5, (C) pH 6.5, (D) pH 9, and (E) pH 10.
47
Figure 2.7. Volume, V, (Å3) plotted versus time (min) fit with Vegard’s Law (Eqn ) -1 -1 with k1 (min ) as the rate constant of reaction stage one and k2 (min ) as the rate constant of reaction stage two at (A) pH 3, (B) pH 5, (C) pH 6.5, (D) pH 9, and (E) pH 10.
48
0
-0.5
-1
-1.5 -2 -2.5 log(dV/dt) y = -0.0791x - 0.9885 -3 R² = 0.8943 -3.5 -4 2 4 6 8 10 12 pH
Figure 2.8. The log of the initial rate plotted as a function of pH. The change in initial cation exchange rate is weakly dependent on pH.
49
Figure 2.9. The change in volume (Å3) versus time (min) verses pH. The unit-cell volume increases as pH increases, indicating more Cs+ enters the interlayer region at a higher pH.
50
Figure 2.10. Plots of Cs adsorbed (wt%) from the flow-through solution as a function of time. The bold line indicates 0 wt% adsorption, while positive percentages are adsorption and negative percentages are desorption. More Cs is adsorbed into the hexagonal H-birnessite as the pH increases. Each plot represents a cation exchange reaction completed at a different pH: (A) pH 3, (B) pH 5, (C) pH 6.5, (D) pH 9, and (E) pH 10.
51
Figure 2.11. Plots of the Mn concentration (ppm) in the leachate as a function of time (min). Each plot represents a cation exchange reaction completed at a different pH: (A) pH 3, (B) pH 5, (C) pH 6.5, (D), pH 9, and (E) pH 10.
52
1.2
1.1
1
0.9
site occupancysite 0.8 (oct) (oct)
0.7 Mn
0.6 2 4 6 8 10 pH
Figure 2.12. Mn(oct) site occupancy increases to 1.0 as pH increases, indicating the vacancies within the octahedral Mn-O sheet are filled with oxidized Mn2/3+ from the interlayer.
53
Figure 2.13. Difference electron Fourier (DELF) maps of the interlayer region calculated by GSAS for 0.001 M Cs+ exchange with hexagonal H-birnessite at pH 6.5, for exchange times (A) 0 min, (B) 60 min, (C) 120 min, (D) 180 min, (E) 240 min, and (F) 300 min. The contour interval is 0.1 e-/Å3.
54
Figure 2.14. ASEM elemental maps showing the distribution of Mn, O, and Cs on partially exchanged hexagonal H-birnessite at pH 9 (40 min).
55 Tables
Table 2.1. Final Rietveld refinement parameter for end-member hexagonal H-birnessite exchanged with 0.001 M Cs+ at varying pHs.
pH 3 pH 5 pH 6.5 pH 9 pH 10 Space group P -3 P -3 P -3 P -3 P -3
Unit cell a = b (Å) 2.85090(4) 2.85054(3) 2.85844(3 ) 2.86977(1) 2.87725(3) c (Å) 7.37897(9) 7.36808(2) 7.35841(9) 7.37082(7) 7.37601(6) V (Å3) 51.938(8) 51.848(9) 52.068(4) 52.570(4) 52.882(1)
Refinement No. of diffraction 1676 1676 1676 1676 1676 points No. of reflections 27 29 29 27 21
Diffraction range (° θ) 3.9-37.303 3.9-37.303 3.9-37.303 3.9-37.303 3.9-37.303 No. of variables 32 31 32 31 30 R(F2) 0.0193 0.017 0.0175 0.017 0.018
Rwp 0.0147 0.0105 0.0144 0.0195 0.0195 χ2 4.823 3.788 3.941 3.473 2.756
56 Table 2.2. Atomic coordinates and isotropic displacement factors for Cs-exchanged hexagonal H-birnessite across a wide pH range.
Site U x 102 pH Atom x y z occupancy iso (Å2) factor
3 Mnoct 0 0 0 0.738(4) 1.50(0) O 0.3333 0.6667 0.12145(1) 1.000(0) 2.47(7) oct O 0.6667 0.3333 0.46085(9) 0.667(2) 8.00(0) int Mn 0 0 0.72346(9) 0.167(7) 1.50(0) int
5 Mnoct 0 0 0 0.709(2) 1.50(0) O 0.3333 0.6667 0.11788(9) 1.000(0) 2.47(7) oct O 0.6667 0.3333 0.46895(2) 0.682(0) 8.00(0) int Mn 0 0 0.72635(5) 0.179(0) 1.50(0) int
6.5 Mnoct 0 0 0 0.816(5) 1.50(0) O 0.3333 0.6667 0.12060(1) 1.000(0) 2.47(7) oct O 0.6667 0.3333 0.46532(5) 0.727(7) 8.00(0) int Mn 0 0 0.71783(1) 0.160(7) 1.50(0) int
9 Mnoct 0 0 0 1.000(0) 1.50(0) O 0.3333 0.6667 0.10760(3) 1.000(0) 2.47(7) oct O 0.6667 0.3333 0.45128(2) 0.879(1) 8.00(0) int Mn 0 0 0.69510(1) 0.139(0) 1.50(0) int
10 Mnoct 0 0 0 1.000(0) 1.50(0) O 0.3333 0.6667 0.10019(2) 1.000(0) 2.47(7) oct O 0.6667 0.3333 0.44941(9) 0.892(8) 8.00(0) int Mn 0 0 0.68417(5) 0.126(9) 1.50(0) int
57 Table 2.3. Selected bond distance values for Cs-exchanged hexagonal H-birnessite as a function of pH.
58 + Table 2.4. The first stage rate constant, k1, and initial rate, dV/dt, of Cs exchange into hexagonal H-birnessite as a function of pH.
pH k (min-1) (Å3/min) 1 3 0.195 0.064 5 0.101 0.037 6.5 0.122 0.033 9 0.122 0.016 10 0.066 0.021
59 Table 2.5. [Cs] in effluent as a function of pH and time.
60 Chapter 3 The rate and mechanism of Cs cation exchange into triclinic Na-birnessite
Abstract
Radioactive 137Cs is found in surface soils due to atmospheric nuclear weapons testing and nuclear waste contamination. The migration of 137Cs into groundwaters can be inhibited by sequestration within clay minerals, and recent studies have shown that phyllomanganates are capable of the rapid uptake of Cs. We applied time-resolved synchrotron X-ray diffraction (TR-XRD), inductively coupled plasma-mass spectrometry
(ICP-MS), and analytical scanning electron microscopy (ASEM) to determine the kinetics and mechanisms of aqueous Cs substitution in the phyllomanganate triclinic Na- birnessite at pH values ranging from 5 to 10. These data demonstrate that the Cs+ exchange rate is relatively insensitive to pH; however, lower concentrations of Cs+ are sequestered at circumneutral pH.
Introduction
Cesium-133 is the naturally occurring stable isotope of cesium. It has a concentration of ppm in the Earth’s crust and ppb in seawater Radioactive s-
137 is a component of high- and low-level nuclear waste produced during nuclear fission and is found in surface soils throughout the world due to atmospheric nuclear weapons testing (Cesium Human Health Fact Sheet 2005). In the United States, the DOE Hanford
Site in Richland, WA has experienced serious Cs-137 contamination problems, because
61 high-level nuclear waste storage tanks have been leaking radioactive solutions since
1961. Cs sorption onto Hanford sediments has been studied as a function of pH, competing cations, temperature, and ionic strength (Liu et al. 2003; Lichtner et al. 2004;
Liu et al. 2004; Ainsworth et al. 2005; Giannakopoulou et al. 2007). The sorption of Cs onto soil mineral surfaces and the cation exchange of Cs into soil mineral structures represent important mechanisms for the inhibition of Cs migration in the subsurface.
Thus, modeling the transport of Cs requires the incorporation of cation exchange data to allow scientists to predict the dispersal of Cs on a field-scale (Steefel et al. 2003).
Several minerals have been shown to sequester Cs with high efficiency, such as synthetic and natural zeolites which trap Cs in their cages and channels. El-Kamash
(2008) demonstrated that the capacity of Cs sorption onto zeolite A increases with the solution Cs concentration and follows a pseudo-second order sorption mechanism, while other authors have demonstrated that Cs leachability decreases above pH 8 and Cs+ sorption is controlled by the size of the zeolite cage (Redkin and Hemley 2000; Mon et al. 2005). Cs also can be sequestered within clays such as illite, smectite, and montmorillonite, and studies have revealed two sorption sites that display partial irreversibility (Comans et al. 1991). Cs sorption onto bentonite clays increases as pH increases and ionic strength (controlled by NaClO4) decreases (Sabodina et al. 2006).
The temperature dependence of Cs sorption is controlled by the type of clay: montomorillonite exhibits a higher Cs distribution coefficient as temperature increases while lateritic soil (quartz, kaolinite, and goethite) adsorbs more Cs at lower temperatures
(Tertre et al. 2005; Wang et al. 2008).
62 Compared with aluminosilicate clays, manganese oxides occur in small abundances in most pedalfers (~1 wt%). Nevertheless, they can play a major role in controlling heavy metal mobility through soils. They possess high cation exchange, redox, and adsorption capacities which counteracts their small concentration in soils (Fu et al. 1991; Barnett et al. 2002; Violante and Pigna 2002; O'Reilly and Hochella 2003;
Weaver and Hochella 2003; Fredrickson et al. 2004; Negra et al. 2005; Zhao et al. 2009).
Manganese oxide reactivity is controlled by the Mn oxidation state (II, III, or IV), the crystal structure (layer or tunnel), and its fine grain size, which increases the sorption capabilities. Of the more than 30 known Mn oxide minerals that have been described, todorokite and birnessite-like phases appear to be particularly common in soils (Taylor et al. 1964; Vaniman et al. 2002; Fredrickson et al. 2004).
Birnessite is the most commonly occurring layered manganese oxide and occurs as submicron grains with large surface areas. Synthetic triclinic birnessite has the
4+ 3+ 3+/4+ chemical formula Na0.56(Mn 1.44Mn 0.56)O4• H2O, and it is composed of Mn O6 octahedral sheets that are negatively charged due to the replacement of Mn4+ with Mn3+.
The octahedral sheets are charge balanced by hydrated Na cations that are positioned halfway between the sheets, and these interlayer cations are highly exchangeable (Fig.
3.1) (Post et al. 2002). X-ray absorption spectroscopy studies of Mn oxides in soils have argued that natural birnessite occurs not only with triclinic but also hexagonal symmetry
(Webb et al. 2005; Tan et al. 2010; Santelli et al. 2011). The primary difference in these varieties of birnessite is attributed to the valence of Mn. In hexagonal birnessite, the octahedral sheet contains only Mn4+ cations and vacancies, and Mn2+ resides within the interlayer (Silvester et al. 1997; Lanson et al. 2002; Post et al. 2008).
63 We expect that the ability of birnessite-like phases to take up contaminant cations will be influenced by Mn valence state and, therefore, structural symmetry. Several researchers have determined Cs distribution coefficients for a variety of phyllomanganates as a function of competing cations, other than H+ (Mikhail and Misak
1988; Dyer et al. 2000; Al-Attar et al. 2003; Liu et al. 2007). In a separate study (Chapter
2), we correlated the rate and magnitude of Cs sequestration by hexagonal H-birnessite with solution pH. Here, we correlate the effects of solution pH with Cs adsorption behavior onto triclinic Na-birnessite. In light of the coexistence of triclinic and hexagonal birnessite varieties in natural soils, the data obtained in the present study will, in conjunction with Chapter 2, allow us to model the interaction of 137Cs contaminants with Mn oxides in soils from nuclear reactor sites, such as Hanford (USA), Chernobyl
(Ukraine), and Fukushima (Japan).
Experimental Methods
Material synthesis
Triclinic Na-birnessite was synthesized according to the procedure described in
Golden et al. (1986). At room temperature, oxygen was bubbled (4.2 L/min) through a mixture of chilled 250 mL 5.5 M NaOH and 200 mL 0.5 M MnCl2 for five hours. The precipitate was divided evenly between 12 centrifuge tubes and centrifuged. The solution was poured off and replaced with fresh DI water; after five replicate rinsing cycles, the birnessite was stored in fresh DI water until needed for experiments. Aliquots of the
64 stored triclinic Na-birnessite were filtered through a μm hydrophilic polypropylene membrane filter (GH Polypro, Pall) and allowed to air dry. The dry triclinic Na- birnessite (250 mg) was ground under acetone in an agate mortar to disaggregate clumps.
After the acetone had fully evaporated, the material was characterized with a conventional sealed Mo tube source on a Rigaku II D/MAX-RAPID microdiffractometer
(Materials Characterization Laboratory, Pennsylvania State University).
Chemical analysis
Time-resolved X-ray diffraction (TR-XRD) experiments were completed at beam line 13-BM-C at the Advanced Photon Source (APS) at Argonne National Laboratory
(ANL) and beam line X7B at the National Synchrotron Light Source (NSLS) at
Brookhaven National Laboratory (BNL). Approximately 4 mg of triclinic Na-birnessite were packed between two cotton plugs in a 0.7 mm quartz capillary and attached to a flow-through apparatus (Fig. 3.2) (Wall et al. 2011). The four Cs-rich cation exchange solutions had the same initial Cs concentration of 0.001 M and were adjusted to pH 5.0,
6.7, 9.0, and 11.0. The pH was adjusted with CsOH, and the cesium concentration was controlled with CsCl. Experiments were not completed at pH 3 because a phase transformation occurred, as will be discussed in the following chapter. Experiments were also not completed at pH 13 because increasing the alkalinity using CsOH led to a Cs concentration of 0.1 M.
TR-XRD patterns at APS were collected every 60 seconds for 5 hours using a
MAR165 CCD detector. The solution flow rate was 1 drop/46 sec (~0.07 mL/min). The
65 sample-to-collector distance was 97.6 mm, allowing us to collect over a 2θ range of 4-
50°. The wavelength was 0.8335 Å. TR-XRD patterns at NSLS were collected every 3 minutes for 3 hours using a MAR345 full imaging plate detector. The sample-to-collector distance was 378.6 mm, allowing collection over a 2θ range of 4-30°. The wavelength was 0.3184 Å. Rietveld refinements (Rietveld 1969) were completed for each TR-XRD pattern using the EXPGUI interface (Toby 2001) of the general structure analysis system
(GSAS) (Larson and Von Dreele 1994). The background intensities for the TR-XRD patterns were fit using up to 16 terms of a shifted Chebyschev or linear interpolation function. The peak profiles were modeled by a pseudo-Voigt profile function as parameterized by Thompson et al. (1987), with asymmetry correction by Finger et al.
(1994), and microstrain anisotropic broadening terms of Stephens (1999).
During initial cycles of refinement only the background, scale, peak profile, and unit-cell parameters were allowed to vary. The position of the O atom in the Mn-O sheet was then refined. Following refinements of the interlayer atom positions, occupancy factors and isotropic atomic displacement factors for the interlayer sites were allowed to vary. Once the interlayer atoms were settled, the isotropic thermal parameters for the Mn and O in the octahedral sheets, followed alternately with those for the interlayer atoms, were refined The final χ2 values for end-member triclinic Cs-birnessite ranged from
0.4408-1.351 (Table 3.1).
The leachate was collected as a function of time for experiments completed at
APS: a stepping motor controlled a UR-150 Newport Rotary Stage with an Al tray capable of supporting 15 polypropylene vials. The stepping motor was programmed to rotate to the next vial every 20 min. The leachate was analyzed with an X-Series 2
66 Thermo Scientific quadrupole inductively coupled plasma-mass spectrometer (ICP-MS) in conjunction with Thermo Scientific PlasmaLab software (Materials Characterization
Laboratory, Pennsylvania State University) to determine the change in Cs, Na, and Mn elemental concentrations.
An FEI Nova NanoSEM 600 analytical scanning electron microscope (ASEM) equipped with Thermo Fisher Scientific NSS 2.3.89 software (Mineral Sciences
Analytical Laboratories, Smithsonian Institution) was used to determine the distribution of Cs and Na across partially exchanged triclinic Na-birnessite grains. These partially exchanged products were obtained by flowing 0.001 M CsCl solutions through triclinic
Na-birnessite powders in a 0.7 mm capillary for 60 min at pH 7. The birnessite was unpacked from the glass capillary, rinsed with 300 mL DI water, and allowed to air dry at room temperature. Dry samples were mounted on SEM stubs with carbon tape for elemental mapping analysis.
Kinetic analysis
We applied the approach of Lopano et al. (2011) to determine the kinetics of cation substitution into the triclinic birnessite structure. This method assumes that cation exchange in this system obeys Vegard’s Law, which posits that the volume of the unit cell varies linearly with the extent of cation exchange (Denton and Ashcroft 1991). As suggested in Lopano et al. (2009), this presumption is reasonable, since the large size of
Cs+ will expand the interlayer of birnessite-like phases. However, the application of
Vegard’s Law to birnessite is complicated by the appearance of an intermediate product,
67 in which Cs has fully exchanged for the interlayer cation but is positionally disordered within the interlayer. The authors treated this issue by including the disordered state as a distinct phase that contributed to the overall unit-cell volume. Following their lead, we propose that the measured volume of the unit cell (Vtot) at any point during the exchange of Cs into triclinic Na-birnessite can be represented by a summation of the three individual phase fractions (Xx), each multiplied by their respective unit-cell volumes, q, r, and s:
(3.1) where XA is the phase fraction of initial triclinic Na-birnessite, XB is the phase fraction of disordered triclinic Cs-birnessite, and XC is the phase fraction of ordered triclinic Cs- birnessite.
As shown in Lopano et al. (2011), the kinetics of Cs+ exchange in triclinic Na- birnessite can be modeled as a two-stage linear reaction. The first stage involves the complete exchange of cations, and in the second stage cations reorder within the interlayer to more thermodynamically stable positions (Fig. 3.3). The rate constant of
-1 each reaction stage (k1 and k2, respectively, with units min ) is determined according to standard kinetic equations for a first order compound reaction (Fromherz 1964):
A (3.2)
- e (3.3)
- - (e -e ) (3.4) -
- - e e - - . (3.5) -
68
where t is time and is the initial phase fraction of starting material, A. According to mass balance:
(3.6) or
. (3.7)
Eqns. 3.3-3. are substituted into Vegard’s Law (Eqn ):
e- e- - - - e e -e - - . (3.8) - -
The dependence of volume versus time for each pH is fit with Eqn. 3.8 using Origin 6.1, solving for q, r, s, k1, and k2. These five variables were allowed to iterate simultaneously using the Levenberg-Marquardt algorithm until the reduced χ2 was minimized. At this time, we are most interested in the initial exchange of cations when XA = 1 and t = 0, and not the reordering of cations; therefore, the initial rate of the cation exchange is determined using the first-stage rate constant (k1) and phase fractions XA and XB. The initial rate in terms of XA, XB, and k1 can be defined with standard first-order reversible kinetics as (Lasaga 1981)
- (3.9)
(3.10)
where n is the reaction order Taking the derivative of Vegard’s Law (Eqn ) with respect to time,
, (3.11)
and substituting Eqn. 3.9 and 3.10 into Eqn. 3.11, the initial rate in is min
69
lim - (3.12) , or
- . (3.13)
The log of the initial rate is plotted versus pH, resulting in a linear regression where the negative slope is the reaction order, n, and the y-intercept is the log of the initial rate
constant, w, with units M min :
log - log , (3.14)
which simplifies to
. (3.15)
Eqn. 3.15 allows us to determine how the solution pH affects the rate of Cs cation exchange measured from volume changes of the birnessite unit cell.
Results
The substitution of Cs+ into triclinic Na-birnessite is evident from X-ray diffraction patterns through a marked decrease in the ratio of the (001) to (002) diffraction peaks (Fig. 3.4). Cs has a larger scattering factor than the exchangeable
Na+/water complex in the interlayer of triclinic Na-birnessite, causing a reduction in the peak intensity ratio between the (001) and (002) peaks from roughly 4:1 to 2:1 (Lopano et al. 2009). Refinement results for the exchange of Cs into triclinic Na-birnessite at pH
5, 6.7, 9, and 11 are compiled in Table 3.1. Atom positions are listed in Table 3.2, and selected bond distances are shown in Table 3.3. The changes in selected unit-cell
70 parameters as a function of time and pH are shown in Figures 3.5-3.8; the parameters not shown are excluded because they did not change within error over the course of the reaction. As was expected for increased Cs uptake, the unit-cell volume increased with time for each experiment. The volumes were fit with Eqn. 3.8 for each cation exchange reaction (Fig. 3.9) to determine the rate constant of stage one, k1, and the initial rate of
cation exchange, , according to Eqn. 3.13 (Table 3.4).
As was the case with hexagonal birnessite, the initial reaction rates were weakly dependent on pH, with less than an order of magnitude difference in dV/dt over the pH range investigated, and for both birnessite variants the values for log(dV/dt) were bracketed between -1 and -2 (Fig. 3.10). Unlike hexagonal birnessite, which exhibited a small decrease in exchange kinetics with higher pH, the rate of exchange in triclinic birnessite was maximized at circumneutral pH, within the error of our measurements.
Thus, we calculated separate rate equations for pH 4.0 to 6.0 and from pH 6.5 to 11.0.
The cation exchange rate that we calculated for an acidic environment was
d log - R2 = 0.9779 (3.16) d
which can be simplified to an exponential form where the initial rate constant, w, in M min is the coefficient and the reaction order, n, is the exponent (Eqn. 3.15):
d - x - . (3.17) d
Conversely, the cation exchange rate in a basic environment was
d log - - R2 = 0.9485 (3.18) d which simplifies to (Eqn. 3.15):
71
d . (3.19) d
The initial cation exchange rate is slightly more dependent on pH when the environment is acidic, indicated by the larger reaction order (0.445). In a basic environment, the reaction order is smaller (0.086), indicating a weaker dependence of cation exchange rate on pH.
In a marked departure from the pattern of Cs+ exchange into hexagonal H- birnessite, the final refined unit-cell volumes and the change in unit-cell volume were the same within error for all of pH values tested (Fig. 3.9). As was demonstrated in Chapter
2, the final unit-cell volumes for Cs-exchanged hexagonal H-birnessite exhibited a step- increase at pH 7. Therefore, to the extent that unit-cell volume is a proxy for Cs+ exchange, we can infer that the amount of Cs+ exchanged into the interlayer of triclinic
Na-birnessite was independent of alkalinity. We tested this inference by measuring the concentrations of Cs, Na, and Mn using ICP-MS analysis of the eluate, which was collected as a function of time after it interacted with the triclinic birnessite. With the initial concentration of Cs in the exchange solutions equal to 0.001 M Cs (or 133 ppm), it can be seen from Table 3.5 that 100 wt% of the Cs was sequestered by the birnessite powder in the first reaction stages for all pH values. We reconfigured the results in Table
3.5 to represent the wt percentage of Cs removed from or released to the flow-through solution as a function of time, and the results of these analyses are shown in Figure 3.11.
Positive data points indicate sequestration by the solid, whereas negative data points represent release of Cs by the solid. At pH 6.7 and 9, 100 wt% of the Cs in the influent was adsorbed during the first 40 min, and 100 wt% of the Cs was adsorbed during the
72 first 100 min at pH 5 and 10. The sequestration of Cs was followed by the release of Cs to the solution only at pH 9. As pH increased, more Na was released into solution (Fig.
3.12) while there was only a minor initial release of Mn at the start of each reaction (Fig.
3.13).
Discussion
Cs exchange rates and loading
The Cs+ exchange rate was largest at circumneutral pH, and our measured value agrees fairly well with that determined by Lopano et al. (2011) for pH ~7 (shown in Fig.
3.10). The small disparity in the results may be attributed to minor variations in substrate packing, particle size, and solution flow rate. These variables are extremely difficult to maintain in strict uniformity, even by the same experimentalist. The rates that we measured are also consistent among experiments completed with different synchtroton X- ray sources (i.e., NSLS and APS). Thus, our methodology generated rates that were reproducible within error.
As noted above, the magnitude of the Cs exchange rate as measured by dV/dt was generally between 10-1 and 10-2 Å3/min, which is the same range determined for hexagonal H-birnessite (as presented in Chapter 2). In contrast, K+ exchange into triclinic birnessite in various low pH and 0.01 M KCl solutions resulted in rates with magnitudes between 10-0.4 and 10-0.9 Å3/min, and the rate decreased with increasing pH from pH 4 to 7 (Lopano 2007). However, in a separate study by Lopano et al. (2011), a
73 single experiment of 0.001 M CsCl at pH 7 resulted in a cation exchange rate 10-1.7
Å3/min (Fig. 3.10). It seems likely that the rates of K+ exchange are faster than those reported here for Cs exchange because the K concentration in those experiments was one order of magnitude larger than the Cs concentration used in this study. Therefore, the rate differences between the K+ and Cs+ cation exchange reactions in birnessite is a result of differences in alkali metal concentrations.
While the initial cation exchange rate of Cs+ into hexagonal H-birnessite decreases as pH increases across a wide pH range, the initial Cs+ exchange rate into triclinic Na-birnessite exhibited distinct acidic and basic trends. In a basic solution, the initial cation exchange rate constants, w, and the reaction orders, n, are analogous between the two substrates:
w = 0.176 Å3·M-1·min-1, n = 0.086, R2 = 0.9485 (triclinic Na-birnessite) (3.20) w = 0.103 Å3·M-1·min-1, n = 0.079, R2 = 0.8943 (hexagonal H-birnessite). (3.21)
We can thus conclude that when pH is greater than 7, the initial cation exchange rate of
Cs+ into triclinic Na-birnessite is the same within error as the rate of Cs cation exchange into hexagonal H-birnessite.
In an acidic solution, however, the rate of Cs cation exchange into triclinic Na- birnessite increased as pH increased (Fig. 3.10). This trend is opposite that of the cation exchange rates for Cs+ into hexagonal H-birnessite and K+ into triclinic Na-birnessite
(Lopano 2007). We can explain the kinetic behavior for hexagonal H-birnessite as follows: As described in Chapter 2, hexagonal H-birnessite sequesters considerably less
74 Cs at low pH than at high pH. Consequently, the final hexagonal Cs-birnessite structure at low pH is much more similar to the starting hexagonal H-birnessite structure than is the final hexagonal Cs-birnessite structure at high pH. It follows that the reaction rate for exchange at low pH is faster than at high pH because less structural change is required at low pH. In contrast to the behavior of hexagonal H-birnessite, triclinic Na-birnessite sequestered more total Cs at the lowest and highest pH than at neutral pH (Fig. 3.11).
Thus, greater structural change occurred at low pH, and, by extension, the reaction rate was slightly slower at pH ~3 than at circumneutral pH. Likewise, the reaction rates at pH
10 were slower than those at pH 7 and 9, and our eluate analyses revealed that more Cs was sequestered by triclinic Na-birnessite at pH 10 than at pH 6.7 or 9 (Fig. 3.14). These observations again suggest that the rates mirror the length of the reaction pathway – that is, the degree of structural change involved in the exchange reaction.
From our ICP-MS analyses of the eluate, we can calculate the absolute amount of
Cs sequestered or released as a function of time, and these results show that the loading of Cs by birnessite occurred over two distinct stages at pH 10 (Fig. 3.15). After an initial episode of sequestration tailed off to near zero at 200 minutes, a second episode of Cs uptake occurred. We speculate that this two-stage loading may indicate that more than one Cs sorption site is active at high pH. One site may have a higher preference for Cs than the second, so that the second site only begins to sorb Cs when the first site is fully occupied. Based on our structural refinements, we believe that Cs has the strongest preference for the interlayer region. If the interlayer sites are completely occupied and the octahedral sheets are still negatively charged, as might be the case at the highest pH,
Cs may sorb onto surface or edge sites. Sorption onto external surface sites would not
75 cause an increase in unit-cell volume, explaining why we do not see a change in dV/dt that corresponds to this second Cs uptake episode.
The greatest release of Na also occurs at pH 10, indicating that more Cs must be exchanged into the interlayer region to maintain charge neutrality of the birnessite (Fig.
3.12). The point-of-zero charge (PZC) of birnessite is approximately 3.0 (Tan et al.
2008), and the octahedral MnO6 sheets become more negatively charged as pH increases beyond 3 (Murray 1974; McKenzie 1981). Thus, in order to maintain neutrality of the surface charge at high pH, more Cs must adsorb to surface sites. When the amount of
Na released is plotted against Cs adsorbed at each pH, it is apparent that the amount of Cs adsorbed to the birnessite increases as Na is released into solution, except at pH 9 when
Cs is leached from the substrate (Fig. 3.16). The slope of these plots indicate the ratio of
Na released to Cs exchanged. Due to charge balance, the slope should equal 1 because one mole of Na should be exchanged with one mole of Cs. This is true at pH 5 and 6.7; however, at high pH, the slope increases to 2.5 at pH 9 and 2 at pH 10. This indicates that twice as much Na is released than Cs is adsorbed. Mn is minimally released from the triclinic Na-birnessite (Fig. 3.13) at all pH values, indicating there is no dissolution of the triclinic birnessite and, as expected, there is no exchangeable Mn in the interlayer.
Back reaction
The release of Cs following its uptake was observed only at pH 9, unlike hexagonal H-birnessite, which released Cs at the end of each reaction. The released Cs was most likely desorbed from frayed or edge sites on the birnessite platelet surfaces or
76 from the reordering of interlayer cations during the second stage of the reaction because the unit-cell volume remained constant once cation exchange was complete. If Cs leached out of the interlayer, we would expect to see the unit-cell volume decrease towards the initial volume of triclinic Na-birnessite.
Mechanism of exchange
Three mechanisms can potentially describe the cation exchange of birnessite: diffusion, dissolution-reprecipitation, or delamination-reassembly. Exchange by diffusion (Helfferich 1962) would result in partially exchanged birnessite grains with Cs- rich reaction rims and Na-rich centers. Dissolution-reprecipitation mechanisms result in intermediate products with reaction rims if the reprecipitation is pseudomorphic or the co-existence of endmember phases (i.e., separate Cs-rich particles and Na-rich particles)
(Putnis 2002; Labotka et al. 2004). In the third mechanism, as described by Lopano et al.
(2009), the birnessite sheets delaminate, allowing the cations in that interlayer region to be exchanged simultaneously, followed by the reassembly of those two sheets. The exchange proceeds through a layer-by-layer substitution until all the Na cations are replaced by Cs cations. More thorough explanations of these three exchange mechanisms can be found in the previous chapter.
ASEM elemental maps of partially exchanged triclinic Na-birnessite at pH 7 showed no Cs-rich reaction rims on the perimeter of individual birnessite grains (Fig.
3.17). Consequently, we rule out diffusion as the exchange mechanism. Likewise, we do not believe that dissolution-reprecipitation was the exchange mechanism because Cs-rich
77 grains did not co-exist with Na-rich grains. Rather, the homogeneous distribution of Cs and Na supports delamination-reassembly as the driving mechanism for exchange.
Conclusion
In summary, our results indicate that the rate of Cs cation exchange into triclinic
Na-birnessite is maximized at neutral pH. The initial cation exchange rates are reproducible and independent of X-ray source, despite variations in substrate packing, particle size, and solution flow rate. The total Cs loading into the birnessite interlayer follows the opposite trend as the rate, with more Cs+ sorption at the most acidic or basic pH. The cation exchange reaction occurs by a delamination-reassembly mechanism, shown by a uniform distribution of Cs and Na across partially exchanged grains. Thus,
TR-XRD studies of Cs cation exchange has provided both kinetic and mechanistic constraints on the uptake of Cs by natural birnessite in soil systems.
Acknowledgements
The author thanks Peter Eng and Nancy Lazarz at APS, Jonathan Hanson at NSLS, and
Li Li of the Pennsylvania State University. Funding for this research was provided by the
NSF-DOE Center for Environmental Kinetics Analysis (CEKA) Grant No. CHE-
0431328 and NSF Grant No. EAR07-45374. Use of the Advanced Photon Source was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy
Sciences, under Contract No. DE-AC02-06CH11357. Use of the National Synchrotron
78 Light Source, Brookhaven National Laboratory, was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-
AC02-98CH10886.
References
Ainsworth, C.C., Zachara, J.M., Wagnon, K., McKinley, S., Liu, C., Smith, S.C., Schaef, H.T., and Gassman, P.L. (2005) Impact of highly basic solutions on sorption of Cs+ to subsurface sediments from the Hanford site, USA. Geochimica Et Cosmochimica Acta, 69(20), 4787-4800. Al-Attar, L., Dyer, A., Paajanen, A., and Harjula, R. (2003) Purification of nuclear wastes by novel inorganic ion exchangers. Journal of Materials Chemistry, 13(12), 2969- 2974. Barnett, M.O., Jardine, P.M., and Brooks, S.C. (2002) U(VI) adsorption to heterogeneous subsurface media: Application of a surface complexation model. Environmental Science & Technology, 36(5), 937-942. Cesium Human Health Fact Sheet. 2005. Argonne National Laboratory, Comans, R.N.J., Haller, M., and Depreter, P. (1991) Sorption of Cesium on Illite - Nonequilibrium Behavior and Reversibility. Geochimica Et Cosmochimica Acta, 55(2), 433-440. Denton, A.R., and Ashcroft, N.W. (1991) Vegard's law. Physical Review A, 43(6), 3161. Dyer, A., Pillinger, M., Harjula, R., and Amin, S. (2000) Sorption characteristics of radionuclides on synthetic birnessite-type layered manganese oxides. Journal of Materials Chemistry, 10(8), 1867-1874. Finger, L.W., Cox, D.E., and Jephcoat, A.P. (1994) A Correction for Powder Diffraction Peak Asymmetry Due to Axial Divergence. Journal of Applied Crystallography, 27, 892-900. Fredrickson, J.K., Zachara, J.M., Kennedy, D.W., Kukkadapu, R.K., McKinley, J.P., Heald, S.M., Liu, C., and Plymale, A.E. (2004) Reduction of TcO4- by sediment- associated biogenic Fe(II). Geochimica et Cosmochimica Acta, 68(15), 3171- 3187. Fromherz, H. (1964) Physico-chemical calculations in science and industry. Butterworth & Co., London. Fu, G., Allen, H.E., and Cowan, C.E. (1991) Adsorption of cadmium and copper by manganese oxide. Soil Science, 152(2), 72-81. Giannakopoulou, F., Haidouti, C., Chronopoulou, A., and Gasparatos, D. (2007) Sorption behavior of cesium on various soils under different pH levels. Journal of Hazardous Materials, 149, 553-556. Golden, D.C., Chen, C.C., and Dixon, J.B. (1986) Synthesis of Todorokite. Science, 231(4739), 717-719.
79 Helfferich, F. (1962) Ion Exchange. 624 p. McGraw-Hill, New York. Labotka, T.C., Cole, D.R., Fayek, M., Riciputi, L.R., and Stadermann, F.J. (2004) Coupled cation and oxygen-isotope exchange between alkali feldspar and aqueous chloride solution. American Mineralogist, 89(11-12), 1822-1825. Lanson, B., Drits, V.A., Feng, Q., and Manceau, A. (2002) Structure of synthetic Na- birnessite: Evidence for a triclinic one-layer unit cell. American Mineralogist, 87(11-12), 1662-1671. Larson, A.C., and Von Dreele, R.B., 1994. GSAS-General Structure Analysis System. Los Alamos National Laboratory, Lasaga, A.C. (1981) Rate Laws of Chemical Reactions. In A.C. Lasaga, and R.J. Kirkpatrick, Eds. Kinetics of Geochemical Processes, 8, p. 1-68. Mineralogical Society of America, Chelsea, MI. Lichtner, P.C., Yabusaki, S., Pruess, K., and Steefel, C.I. (2004) Role of competitive cation exchange on chromatographic displacement of cesium in the vadose zone beneath the Hanford S/SX tank farm. Vadose Zone Journal, 3(1), 203-219. Liu, C., Zachara, J.M., Qafoku, O., and Smith, S.C. (2003) Effect of temperature on Cs+ soprtion and desorption in subsurface sediments at the Hanford Site, U.S.A. Environmental Science & Technology, 37, 2640-2645. Liu, C.X., Zachara, J.M., and Smith, S.C. (2004) A cation exchange model to describe Cs+ sorption at high ionic strength in subsurface sediments at Hanford site, USA. Journal of Contaminant Hydrology, 68(3-4), 217-238. Liu, Z.H., Kang, L.P., Zhao, M.Z., and Ooi, K. (2007) Preparation, ion-exchange, and electrochemical behavior of Cs-type manganese oxides with a novel hexagonal- like morphology. Journal of Materials Research, 22(9), 2437-2447. Lopano, C.L. (2007) Time-resolved structural analysis of cation exchange reactions in synthetic birnessite. Department of Geosciences, Doctorate of Philosphy. The Pennsylvania State University, University Park. Lopano, C.L., Heaney, P.J., Bandstra, J.Z., Post, J.E., and Brantley, S.L. (2011) Kinetic analysis of cation exchange in birnessite using time-resolved synchrotron X-ray diffraction. Geochimica Et Cosmochimica Acta, 75(14), 3973-3981. Lopano, C.L., Heaney, P.J., and Post, J.E. (2009) Cs-exchange in birnessite: Reaction mechanisms inferred from time-resolved X-ray diffraction and transmission electron microscopy. American Mineralogist, 94(5-6), 816-826. McKenzie, R.M. (1981) The surface charge on manganese dioxides. Australian Journal of Soil Research, 19, 41-50. Mikhail, E.M., and Misak, N.Z. (1988) Sorption of Cesium and Cobalt Radionuclides on a New Manganese Oxide. Applied Radiation and Isotopes, 39(11), 1121-1124. Mon, J., Deng, Y., Flury, M., and Harsh, J.B. (2005) Cesium incorporation and diffusion in cancrinite, sodalite, zeolite, and allophane. Microporous and Mesoporous Materials, 86, 277-286. Murray, J.W. (1974) The surface chemistry of hydrous manganese dioxide. Journal of Colloid and Interface Science, 46(3), 357-371. Negra, C., Ross, D.S., and Lanzirotti, A. (2005) Soil manganese oxides and trace metals: Competitive sorption and microfocused synchrotron X-ray fluorescence mapping. Soil Science Society of America Journal, 69(2), 353-361.
80 O'Reilly, S.E., and Hochella, M.F. (2003) Lead sorption efficiencies of natural and synthetic Mn and Fe-oxides. Geochimica Et Cosmochimica Acta, 67(23), 4471- 4487. Post, J.E., Heaney, P.J., and Ertl, A. (2008) Rietveld refinement of the rancieite structure using synchrotron powder diffraction data. Powder Diffraction, 23(1), 10-14. Post, J.E., Heaney, P.J., and Hanson, J. (2002) Rietveld refinement of a triclinic structure for synthetic Na-birnessite using synchrotron powder diffraction data. Powder Diffraction, 17(3), 218-221. Putnis, A. (2002) Mineral replacement reactions: from macroscopic observations to microscopic mechanisms. Mineralogical Magazine, 66(5), 689-708. Redkin, A.F., and Hemley, J.J. (2000) Experimental Cs and Sr sorption on analcime in rock-buffered systems at 250-300 degrees C and P-sat and the thermodynamic evaluation of mineral solubilities and phase relations. European Journal of Mineralogy, 12(5), 999-1014. Rietveld, H.M. (1969) A Profile Refinement Method for Nuclear and Magnetic Structures. Journal of Applied Crystallography, 2, 65-71. Sabodina, M.N., Kalmykov, S.N., Sapozhnikov, Y.A., and Zakharova, E.V. (2006) Neptunium, plutonium and Cs-137 sorption by bentonite clays and their speciation in pore waters. Journal of Radioanalytical and Nuclear Chemistry, 270(2), 349-355. Santelli, C.M., Webb, S.M., Dohnalkova, A.C., and Hansel, C.M. (2011) Diversity of Mn oxides produced by Mn(II)-oxidizing fungi. Geochimica Et Cosmochimica Acta, 75(10), 2762-2776. Silvester, E., Manceau, A., and Drits, V.A. (1997) Structure of synthetic monoclinic Na- rich birnessite and hexagonal birnessite. 2. Results from chemical studies and EXAFS spectroscopy. American Mineralogist, 82(9-10), 962-978. Steefel, C.I., Carroll, S., Zhao, P.H., and Roberts, S. (2003) Cesium migration in Hanford sediment: a multisite cation exchange model based on laboratory transport experiments. Journal of Contaminant Hydrology, 67(1-4), 219-246. Stephens, P.W. (1999) Phenomenological model of anisotropic peak broadening in powder diffraction. Journal of Applied Crystallography, 32, 281-289. Tan, H., Zhang, G.X., Heaney, P.J., Webb, S.M., and Burgos, W.D. (2010) Characterization of manganese oxide precipitates from Appalachian coal mine drainage treatment systems. Applied Geochemistry, 25(3), 389-399. Tan, W.F., Lu, S.J., Liu, F., Feng, X.H., He, J.Z., and Koopall, L.K. (2008) Determination of the point-of-zero, charge of manganese oxides with different methods including an improved salt titration method. Soil Science, 173(4), 277- 286. Taylor, R.M., McKenzie, R.M., and Norrish, K. (1964) The mineralogy and chemistry of manganese in some Australian soils. Australian Journal of Soil Research, 2, 235- 248. Tertre, E., Berger, G., Castet, S., Loubet, M., and Giffaut, E. (2005) Experimental sorption of Ni2+, Cs+ and Ln(3+) onto a montmorillonite up to 150 degrees C. Geochimica Et Cosmochimica Acta, 69(21), 4937-4948.
81 Thompson, P., Cox, D.E., and Hastings, J.B. (1987) Rietveld Refinement of Debye- Scherrer Synchrotron X-Ray Data from Al2O3. Journal of Applied Crystallography, 20, 79-83. Toby, B.H. (2001) EXPGUI, a graphical user interface for GSAS. Journal of Applied Crystallography, 34, 210-213. Vaniman, D.T., Chipera, S.J., Bish, D.L., Duff, M.C., and Hunter, D.B. (2002) Crystal chemistry of clay-Mn oxide associations in soils, fractures, and matrix of the Bandelier Tuff, Pajarito Mesa, New Mexico. Geochimica Et Cosmochimica Acta, 66(8), 1349-1374. Violante, A., and Pigna, M. (2002) Competitive sorption of arsenate and phosphate on different clay minerals and soils. Soil Science Society of America Journal, 66(6), 1788-1796. Wall, A.J., Heaney, P.J., Mathur, R., Post, J.E., Hanson, J.C., and Eng, P.J. (2011) A flow-through reaction cell that couples time-resolved X-ray diffraction with stable isotope analysis. Journal of Applied Crystallography, 44, 429-432. Wang, T.H., Li, M.H., Yeh, W.C., Wei, Y.Y., and Teng, S.P. (2008) Removal of cesium ions from aqueous solution by adsorption onto local Taiwan laterite. Journal of Hazardous Materials, 160(2-3), 638-642. Weaver, R.M., and Hochella, M.F. (2003) The reactivity of seven Mn-oxides with Cr- aq(3+): A comparative analysis of a complex, aq environmentally important redox reaction. American Mineralogist, 88(11-12), 2016-2027. Webb, S.M., Tebo, B.M., and Bargar, J.R. (2005) Structural characterization of biogenic Mn oxides produced in seawater by the marine bacillus sp strain SG-1. American Mineralogist, 90(8-9), 1342-1357. Zhao, W., Feng, X., Tan, W., Liu, F., and Ding, S. (2009) Relation of lead adsorption on birnessites with different average oxidation states of manganese and release of Mn2+/H+/K+. Journal of Environmental Sciences, 21, 520-526.
82 Figures
Figure 3.1. The crystal structure of triclinic Na-birnessite. The Mn-O octahedral sheets contain Mn3+/4+ in all octahedral positions with hydrated Na+ as interlayer cations (adapted from Post 2002).
83
Figure 3.2. Flow-through TR-XRD experimental setup at beamline 13-BM-C at APS.
84
Figure 3.3. A schematic of the two-stage cation exchange reaction: stage one is the complete exchange of cations while stage two is the reordering of cations within the interlayer.
85
Figure 3.4. Stacked TR-XRD patterns of triclinic Na-birnessite exchanged with 0.001 M Cs+ at pH 11, with 2θ (°) along the x-axis, intensity along the y-axis, and time (min) on the z-axis where each pattern represents 5 min in time.
86
Figure 3.5. Unit cell parameter a (Å) plotted versus time (min) for (A) pH 4.7, (B) pH 6.7, (C) pH 8.9, and (D) pH 10.7.
87
Figure 3.6. Unit cell parameter c (Å) plotted versus time (min) for (A) pH 4.7, (B) pH 6.7, (C) pH 8.9, and (D) pH 10.7.
88
Figure 3.7. Unit cell parameter β (°) plotted versus time (min) for (A) pH 4.7, (B) pH 6.7, (C) pH 8.9, and (D) pH 10.7.
89
Figure 3.8. Unit cell parameter γ (°) plotted versus time (min) for (A) pH 4.7, (B) pH 6.7, (C) pH 8.9, and (D) pH 10.7.
90
Figure 3.9. Volume, V, (Å3) plotted as a function of time (min) fit with Eqn. 3.8 for (A) pH 4.7, (B) pH 6.7, (C) pH 8.9, and (D) pH 10.7.
91
Figure 3.10. The log of the initial rate plotted as a function of pH from three replicate reactions: (■) NSLS, March , (♦) APS, June , and (▲) APS, November 2011. Data from Lopano et al. ( ) is also plotted (●)
92
Figure 3.11. Plots of Cs adsorbed (wt%) from the flow-through solution as a function of time (min). Each plot represents a cation exchange reaction completed at a different pH: (A) pH 5, (B) pH 6.7, (C) pH 9, and (D) pH 10.
93
Figure 3.12. Na (ppm) released during the Cs cation exchange reaction plotted as a function of time (min). Each plot represents a cation exchange reaction completed at a different pH: (A) pH 5, (B) pH 6.7, (C) pH 9, and (D) pH 10.
94
Figure 3.13. Mn (ppm) released during the Cs cation exchange reaction plotted as a function of time (min). Each plot represents a cation exchange reaction completed at a different pH: (A) pH 5, (B) pH 6.7, (C) pH 9, and (D) pH 10.
95
2 1.8
1.6 1.4 1.2 1 0.8
0.6 Cs loading (mg) loading Cs 0.4 0.2 0 4 5 6 7 8 9 10 11 pH
Figure 3.14. Cs loading (mg) as a function of pH. Note that the amount of triclinic Na- birnessite is assumed to remain constant at 4 mg for each experiment.
96
0.6 pH 10
0.5
0.4
0.3
0.2 Cs loading (mg) loading Cs 0.1
0 0 100 200 300 400 500 600 Time (min)
Figure 3.15. Cs loading (mg) as a function of time (min) at pH 10. Cs is sorbed during two distinct time frames, which may indicate there is a second sorption site on birnessite at pH 10.
97
Figure 3.16. Na released (x 10-6 mol) from the triclinic birnessite as Cs (x 10-6 mol) is sequestered from solution. Each data point represents the mass of each element as the reaction proceeds forward in time as a function of pH: (A) pH 5, (B) pH 6.7, (C) pH 9, and (D) pH 10.
98
Figure 3.17. ASEM elemental maps showing the distribution of Mn, O, Na, and Cs on partially exchanged triclinic Na-birnessite at pH 7 (60 min).
99 Tables
Table 3.1. Final Rietveld refinement parameters for end-member triclinic Na-birnessite exchange with 0.001 M Cs+ at varying pHs.
pH 4.7 pH 6.7 pH 8.9 pH 10.7 Space group C -1 C -1 C -1 C -1
Unit cell a (Å) 5.14018(2) 5.12370(9) 5.12072(8) 5.14396(0) b (Å) 2.84144(1) 2.85385(4) 2.84774(9) 2.85042(2) c (Å) 7.43766(4) 7.49998(0) 7.44225(6) 7.44297(6) α (°) 89.653(6) 89.860(8) 89.522(6) 89.832(8) β (°) 101.011(8) 101.106(9) 100.419(8) 101.138(2) γ (°) 89.791(2) 89.620(7) 89.655(9) 90.323(9) V (Å3) 106.627(6) 107.609(8) 106.730(5) 107.074(8)
Refinement No. of diffraction points 1241 958 1241 1241 No. of reflections 74 267 73 73
Diffraction range ( θ) 11.5 -36.2 4.4 -23.8 11.5 -36.2 11.5 -36.2 No. of variables 31 20 30 28 R(F2) 0.0141 0.0425 0.0143 0.0167
Rwp 0.0119 0.0185 0.0154 0.0124 χ2 0.5119 0.4408 1.351 0.7557
100 Table 3.2. Atomic coordinates and isotropic displacement factors for Cs-exchanged triclinic Na-birnessite across a wide pH range.
Site U x 102 pH Atom x y z occupancy iso (Å2) factor
4.7 Mnoct 0 0 0 1.0 0.60(0)
Ooct 0.35021(4) 0.01272(5) 0.10981(5) 1.0 1.20(0)
Oint 0.29422(1) -0.15501(4) 0.42204(6) 0.525(8) 4.34(0)
Oint2 0.39534(8) 0.49750(8) 0.44864(7) 0.437(2) 1.07(0)
6.7 Mnoct 0 0 0 1.0 1.50(6)
Ooct 0.36701(6) -0.00110(5) 0.12074(3) 1.008(2) 1.68(3)
Oint 0.64905(2) 0.32700(2) 0.50365(0) 0.689(6) 10.11(2)
8.9 Mnoct 0 0 0 1.0 0.60(0)
Ooct 0.35345(1) 0.02738(6) 0.11292(5) 1.0 1.20(0)
Oint 0.33586(1) -0.19381(1) 0.42673(2) 0.544(1) 10.65(4)
Oint2 0.39142(2) 0.50799(5) 0.46156(0) 0.411(7) 2.53(3)
10.7 Mnoct 0 0 0 1.0 0.60(0)
Ooct 0.35242(2) 0.00092(5) 0.11061(4) 1.0 1.20(0)
Oint 0.34580(0) 0.542198 0.46524(1) 0.587(8) 9.20(6)
Oint2 0.727462 0.86199(7) 0.58063(5) 0.395(3) 2.53(3)
101 Table 3.3. Bond distance values for Cs-exchanged triclinic Na-birnessite as a function of pH.
Bond pH 4.7 pH 6.7 pH 8.9 pH 10.7
Mnoct-Mnoct 2.8414(10) x 2 2.8539(14) x 2 2.8477(14) x 2 2.8504(8) x 2
Mnoct-Mnoct 2.9412(8) x 2 2.9407(17) x 2 2.9371(15) x 2 2.9334(8) x 2
Mnoct-Mnoct 2.9321(8) x 2 2.9242(17) x 2 2.9222(15) x 2 2.9475(8) x 2
Mnoct-Ooct 1.830(5) x 2 1.9236(13) x 2 1.855(8) x 2 1.841(6) x 2
Mnoct-Ooct 1.847(9) x 2 1.8904(12) x 2 1.817(13) x 2 1.871(10) x 2
Mnoct-Ooct 1.905(8) x 2 1.8822(12) x 2 1.939(12) x 2 1.885(8) x 2
Ooct-Mnoct 1.830(5) 1.9236(13) 1.855(8) 1.841(6)
Ooct-Mnoct 1.905(8) 1.88229(12) 1.939(12) 1.885(8)
Ooct-Mnoct 1.847(9) 1.8904(12) 1.817(13) 1.871(10)
Oint-Oint 1.43(4) 1.80519(11) 2.15(11) 1.59(4)
Oint-Oint 1.1323(7) 1.55(12) 2.05(9)
Oint-Oint 1.690(34)
Oint-Oint2 1.11(4) 0.91(7) 1.24(4)
Oint-Oint2 1.93(4) 2.03(8) 1.76(4)
Oint-Oint2 1.95(4) 1.73(6) 1.25(4)
Oint-Oint2 2.14(4) 1.64(8)
Oint-Oint2 1.56(4)
Oint2-Oint 1.93(4) 2.03(8) 1.24(4)
Oint2-Oint 1.11(4) 0.91(7) 1.76(4)
Oint2-Oint 1.95(4) 1.73(6) 1.25(4)
Oint2-Oint 2.14(4) 1.64(8)
Oint2-Oint 1.56(4)
Oint2-Oint2 1.195(33) 1.155(32) 1.42(4)
Oint2-Oint2 2.15(10)
102 + Table 3.4. The first stage rate constant, k1, and initial rate, dV/dt, of Cs exchange into triclinic Na-birnessite as a function of pH for each replicate reaction.
-1 3 0903 NSLS pH k1 (min ) (Å /min)
6.7 0.0148 0.0447
9.1 0.0153 0.0329
11.0 0.0116 0.0191
-1 3 0906 APS pH k1 (min ) (Å /min)
4.7 0.0081 0.0237
5.8 0.0888 0.0808
8.9 0.0271 0.0482
10.7 0.0097 0.0177
-1 3 0911 APS pH k1 (min ) (Å /min)
5.0 0.0405 0.0408
6.5 0.0380 0.0590
9.0 0.0148 0.0399
10.0 0.0085 0.0036
103 Table 3.5. [Cs] in effluent as a function of pH and time.
104 Chapter 4 Exploring the phase transformation between triclinic and hexagonal birnessite at pH 3 and 13
Abstract
Birnessite-like phases are among the most commonly occurring manganese oxides in soils, and their structures can exhibit either hexagonal or triclinic symmetry. Triclinic birnessite has been shown to transform into hexagonal birnessite at low pH and in the presence of reducing bacteria, but no studies have demonstrated that the reverse transformation will occur at high pH or that any symmetry-changing transformation will proceed during a competing cation exchange reaction. Here we report the results of our study on the effect of pH and aqueous Cs on the phase transformation characteristics between triclinic and hexagonal birnessite in order to fully understand the 137Cs sequestration capacity of each birnessite phase in a range of groundwater chemistries.
Time-resolved X-ray diffraction (TR-XRD) and inductively coupled plasma-mass spectrometry (ICP-MS) were used to measure the kinetics of the phase transformation as a function of pH and aqueous Cs concentration. These data demonstrate that: (1) the reverse transformation from hexagonal to triclinic symmetry will proceed at high pH; (2) the triclinic hexagonal phase transformations occur in both directions even in the presence of aqueous Cs; (3) Cs cation exchange precedes the phase transformation; and
(4) the Cs+ remains in the interlayer region during both phase transformations.
105 Introduction
The global average Mn concentration is between 80 and 1300 ppm; McBride
(1994), Barnett et al. (2002), Fredrickson et al. (2004), and Taylor et al. (1964) report that the average soil contains less than 1 wt% Mn oxides. Although surface concentrations of
Mn are often low, Mn oxides commonly control heavy metal mobility in soils because many Mn oxide phases exhibit high sorption and cation exchange capacities for heavy metals (Burke 1970; Burns 1976; McKenzie 1977; McKenzie 1980; Eary and Ral 1987;
Fu et al. 1991; Le Goff et al. 1996; Al-Attar and Dyer 2002; Lopano et al. 2007; Kang et al. 2010; Cho et al. 2011). Layered Mn oxides, or phyllomanganates, are often the most reactive phases, and these include birnessite, ranciéite, and lithiophorite (Taylor et al.
1964; Lind and Hem 1993; Bilinski et al. 2002; Vaniman et al. 2002).
Birnessite-like phases are reported to occur with either triclinic or hexagonal symmetry. Synthetic triclinic Na-birnessite has the chemical formula
4+ 3+ Na0.56(Mn 1.44Mn 0.56)O4 • H2O (Post and Veblen 1990), and the Mn cations reside in
3+ 4+ octahedral MnO6 sheets (Fig. 4.1). The presence of Mn in place of Mn imparts a negative charge to the sheets, which are charge balanced by hydrated cations that reside between the sheets. In hexagonal birnessite, the octahedral MnO6 sheets are composed of
Mn4+ cations and Mn vacancies (□). For example, ranciéite, which is a naturally
4+ occurring birnessite-like phase, has the chemical formula Ca0.19K0.01(Mn 0.91
□0.09)O2•0.63H2O (Post et al. 2008). In synthetic hexagonal H-birnessite, the octahedral vacancies are charge balanced by interlayer Mn2+/3+ and protons (Fig. 4.2). These species reside within the interlayer in tridentate coordination with the octahedral vacancies
106 (Silvester et al. 1997). The primary distinction between triclinic and hexagonal symmetry is the large proportion of Mn3+ and lack of vacancies in the triclinic octahedral sheet, according to Silvester et al. (1997).
Early crystallographic studies of natural birnessite-like phases presumed a hexagonal symmetry for the mineral. Giovanoli et al. (1970) argued that a synthetic
Mn(III,IV) oxide identified as hexagonal -MnO2 exhibited an X-ray powder diffraction pattern that was similar to that of natural birnessite, and Glover (1977) likewise argued for a hexagonal structure for birnessite found in a sediment core from the Gulf of Mexico.
Chukhrov et al. (1985) and Bargar et al. (2009) characterized birnessite-like phases found in the Pacific Ocean and Pinal Creek, AZ with TEM, powder XRD, XRF, EXAFS, and
XANES, and they report a hexagonal or pseudo-hexagonal symmetry for the natural birnessite structures they examined. Frierdich et al. (2011) found that phyllomanganates deposited in caves were mixtures of hexagonal birnessite and vernadite (δ-MnO2) by conducting ICP-OES, SEM-EDS, and XRD experiments.
On the other hand, synchrotron-based X-ray absorption spectroscopy has led scientists to identify some natural birnessite-like phases as triclinic. Webb et al. (2005a) reported that microbial oxidation of Mn2+ initially precipitated nanocrystalline, one-layer thick phyllomanganate sheets with numerous vacancies. Cations in solution (e.g., Ca or
Na) drove the rapid ordering of single sheets into bulk hexagonal birnessite layers, followed by the transformation into pseudo-orthogonal triclinic birnessite. Tan et al.
(2010) collected soil samples from two sites in western PA and quantified the percentages of natural hexagonal and triclinic birnessite from each site with EXAFS.
107 Significant amounts of each birnessite phase were found at each site, indicating the importance of each phase in natural systems.
Webb et al. (2005a) report that biogenically precipitated hexagonal birnessite transforms to pseudo-orthogonal birnessite within days in soils, and they suggest several phase transformation mechanisms to account for the transition. They propose that surface redox reactions between Mn4+ and Mn2+ may produce Mn3+, but they discount this pathway on the presumption that the difference between hexagonal and pseudo- orthogonal birnessite centers on the ordering of Mn3+, not the net valence of Mn within the octahedral sheet. A second phase transformation mechanism they consider involves solid-state rearrangement of octahedral Mn. However, this behavior has only been observed with EXAFS during the transformation of pseudo-orthogonal birnessite to hexagonal birnessite (Silvester et al. 1997).
Lastly, these authors point to thermodynamic driving forces to stimulate symmetry-changing transitions in birnessite-like phases. Bargar et al. (2005) report that a biologically produced birnessite quickly transformed to fietknechtite in the presence of high Mn2+ concentrations in a batch reactor; however, when the Mn2+ concentration fell below 0.5 mM, the feitknechtite substrate transformed back to birnessite. Therefore,
Webb and others propose that the relative thermodynamic stabilities of hexagonal and pseudo-orthogonal birnessites differ with respect to particular cations in solution, and this difference may initiate the phase transformation (Webb et al. 2005b; Zhu et al. 2010).
Protons in solution may also alter the relative thermodynamic stabilities of birnessite-like phases with different symmetries. Hem (1978) was an early experimentalist to apply the Mn3+ disproportionation reaction to manganese oxides. Drits
108 et al. (1997) and Lanson et al. (2000) support that a Mn3+ disproportionation reaction occurs at low pH:
3+ 4+ 2+ 2Mn oct Mn oct + Mn interlayer □oct. (4.1)
Manceau et al. (1997) and Lanson et al. (2002) simultaneously monitored heavy metal cation exchange and the reaction of Na-buserite (a doubly hydrated 10-Å phyllomanganate whose symmetry is unknown) to 7-Å hexagonal H-birnessite at low pH with XRD and EXAFS. In each case, the transformation reaction mechanism was independent of the amount of heavy metal sorbed, indicating to these authors that other chemical influences (i.e., pH) must control these reactions.
Fischer et al. (in prep), however, have demonstrated that dissolved Cr3+ will catalyze the transition from triclinic to hexagonal birnessite, and in general the effects of dissolved metals and pH on the behavior of birnessite remain poorly constrained. As demonstrated in Chapters 2 and 3, even though the kinetics of metal uptake by birnessite show only a small dependence on pH, the amount of Cs sequestered by hexagonal H- birnessite is strongly determined by pH. Moreover, our work suggests that the hexagonal-triclinic transition itself influences the amount of Cs that can be removed from solution.
These processes may be particularly important at localities such as the DOE
Hanford Site in Richland, WA, where about 3.8 million L of Cs-rich high-level nuclear waste have leaked from single-shell storage tanks, in part due to the high pH of the waste
(pH 13-14). These caustic solutions have corroded the tanks, contaminating approximately 28,300 m3 of soil (McKinley et al. 2001; Gee et al. 2007). Although the pH of the contaminant plumes from these tanks initially is 14, the pH decreases with
109 increasing distance beneath the tanks as the solutions are buffered by soil minerals such as birnessite (Wan et al. 2004). In the present study, we determine the rate of transformation between triclinic and hexagonal birnessite phases as a function of pH, in the presence and absence of Cs. For the first time, we show that the transformation reaction is reversible as a function of pH.
Experimental section
Material synthesis
Triclinic Na-birnessite was synthesized according to the procedure described in
Golden et al. (1986). At room temperature, oxygen was bubbled (4.2 L/min) through a mixture of chilled 250 mL 5.5 M NaOH and 200 mL 0.5 M MnCl2 for five hours. The precipitate was divided evenly among 12 centrifuge tubes and centrifuged. The solution was poured off and replaced with fresh DI water; after five replicate rinsing cycles, the birnessite was stored in fresh DI water until needed for experiments. After aging at STP conditions, aliquots of the stored triclinic Na-birnessite were filtered through a μm hydrophilic polypropylene membrane filter (GH Polypro, Pall) and allowed to air dry.
Dry triclinic Na-birnessite (250 mg) was ground under acetone and reacted with 250 mL
0.001 M HCl for three hours. The resultant hexagonal H-birnessite was rinsed with DI water, filtered through a μm filter, and allowed to air dry The material was initially characterized with a conventional sealed Mo tube source on a Rigaku II D/MAX-RAPID
110 microdiffractometer (Materials Characterization Laboratory, Pennsylvania State
University).
Chemical analysis. Time-resolved X-ray diffraction (TR-XRD) experiments were completed at beam line 13-BM-C at the Advanced Photon Source (APS) at Argonne
National Laboratory (ANL). Approximately 4 mg of birnessite were packed between two cotton plugs in a 0.7 mm quartz capillary and attached to a flow-through apparatus as described in Wall et al. (2011) (Fig. 4.3). Triclinic Na-birnessite was reacted with separate solutions of 0.001 M HCl (pH 3) and 0.001 M CsCl (buffered to pH 3 with HCl).
Hexagonal H-birnessite was reacted with separate solutions of 0.1 M NaOH and 0.1 M
CsOH (pH 13). The average solution flow rate was 1.39 drop/min (~ 0.07 mL/min). The leachate was collected as a function of time: a stepping motor controlled a UR-150
Newport Rotary Stage with an Al tray capable of supporting 15 polypropylene vials. The stepping motor was programmed to rotate to the next vial every 20 min. The leachate was analyzed with an X-Series 2 Thermo Scientific quadrupole inductively coupled plasma-mass spectrometer (ICP-MS) in conjunction with Thermo Scientific PlasmaLab software (Materials Characterization Laboratory, Pennsylvania State University) to determine the change in Cs, Na, and Mn elemental concentrations.
Structural analysis
TR-XRD patterns of triclinic Na-birnessite were collected every 60 seconds for 9 hours and the patterns of hexagonal H-birnessite were collected every 30 seconds for 8.5 hours using a MAR165 CCD detector. Because the reaction of triclinic Na-birnessite
111 with 0.001 M CsCl at pH 3 was completed separately from the other three experiments, the parameters for this run were different from the others. In the 0.001 M CsCl experiment at pH 3, the sample-to-collector distance was 97.6 mm, and the wavelength was 0.8335 Å, allowing us to collect over a θ range of -50°. For the remaining three experiments, the sample-to-collector distance was 103.8 mm with a wavelength of 0.8321
, for a θ range of -40°. The sample was rotated through a 30° phi angle for all four experiments.
Rietveld refinements (Rietveld 1969) were completed for each TR-XRD pattern using the EXPGUI interface (Toby 2001) of the general structure analysis system
(GSAS) (Larson and Von Dreele 1994). The background intensities for the TR-XRD patterns were fit using up to 15 terms of a linear interpolation function. The peak profiles were modeled by a pseudo-Voigt profile function as parameterized by Thompson et al.
(1987), with asymmetry correction by Finger et al. (1994), and microstrain anisotropic broadening terms of Stephens (1999).
During initial cycles of refinement only the background, scale, peak profile, and unit-cell parameters were allowed to vary. The position of the O atom in the Mn-O sheet was then refined. Following refinements of the interlayer atom positions, occupancy factors and isotropic atomic displacement factors for the interlayer sites were allowed to vary. Once the interlayer atoms settled, the isotropic thermal parameters for the Mn and
O ions in the octahedral sheets, followed alternately with those for the interlayer atoms, were refined The final χ2 values for end-member birnessite structures ranged from
1.516-2.850 (Table 4.1). Soft bond constraints were applied to the Mn-O bond distances,
112 constrained at 1.90 Å with a standard deviation of 0.02 Å. The initial bond constraint weight was initially 5.0 and reduced to 2.0 over subsequent iterations.
Kinetic analysis
When triclinic and hexagonal phases of birnessite were both present, we were unable to complete a Rietveld refinement for either phase because the structure parameters were too similar. Instead, we quantified the changing ratio of the triclinic/hexagonal phases during the transition by selecting the refined structures for each phase nearest the transition and then modeling the XRD patterns of a 1:1 mixture of these structures using CrystalDiffract 1.4.0 (CrystalMaker, Inc). From this mixture, we extracted relative intensity ratios (RIR), which allowed us to determine the evolution of each phase fraction during the reaction. Since the densities of the triclinic and hexagonal phases were the same within error, a 1:1 mixture by volume was equivalent to a 1:1 mixture by mass. Consequently, the volume-based RIR ratios yielded by CrystalDiffract were not further corrected to achieve a conventional mass-based RIR. For each pattern collected during the symmetry-changing transitions, we decomposed overlapping (002) peaks for the triclinic and hexagonal phases using Jade 6.5.23 (Materials Data, Inc.), integrated the areas of the component peaks, and then calculated a mass ratio normalized by the RIR. This procedure yielded the relative mass abundance of the triclinic and hexagonal phases as a function of time. The rates of the phase transformations were modeled using standard first-order kinetic equations (House 1997).
113 Results
At pH 3, triclinic Na-birnessite transformed to hexagonal birnessite in solutions containing 0.001 M HCl or 0.001 M CsCl, as can be observed through the disappearance of diagnostic triclinic peaks and the emergence of peaks associated with the hexagonal structure in stacked TR-XRD patterns (Fig. 4.4 and 4.5). In contrast, the reverse reaction occurred at pH 13. As can be seen in the TR-XRD patterns in Figures 4.6 and 4.7, hexagonal H-birnessite transformed to triclinic Na-birnessite in the presence of 0.001 M
NaOH and to triclinic Cs-birnessite in 0.001 M CsOH. Refinement results for the end- member triclinic and hexagonal birnessite phases are compiled in Table 4.1. Atomic coordinates are listed in Table 4.2 and a summary of bond distances are shown in Table
4.3.
Discussion: Triclinic Na-birnessite reactions at pH 3
Rietveld analyses
When triclinic Na-birnessite reacted with 0.001 M HCl at pH 3, the unit-cell volume expanded from 105.5 Å3 to 106.5 Å3 within 35 min (Fig. 4.8A). After this expansion, diffraction peaks corresponding to hexagonal H-birnessite appeared. It was not possible to refine the structures of the individual phases when both triclinic and hexagonal birnessite were present because the unit-cell parameters were too similar.
Based on the disappearance of diffraction peaks, the triclinic Na-birnessite had completely transformed after 70 min of reaction. The final hexagonal H-birnessite
114 exhibited a unit-cell volume of 51.2 Å3 (or 102.4 Å3 in the corresponding triclinic setting). In the presence of 0.001 M CsCl at pH 3, the initial triclinic Na-birnessite unit- cell volume increased from 105.8 Å3 to 106.1 Å3 within the first 60 min (Fig. 4.8B). The triclinic structure completely transformed to hexagonal birnessite after 300 min of reaction, resulting in hexagonal Cs-birnessite with a unit-cell volume of 51.7 Å3 (or 103.4
Å3 in the corresponding triclinic setting). The first structural refinement of both hexagonal H-birnessite and hexagonal Cs-birnessite can be found in Chapter 2.
ICP-MS analyses
The concentrations of Cs, Na, and Mn in the effluent as a function of time were measured with ICP-MS for both reactions. In the absence of Cs, Na cations were exchanged out of the interlayer for protons, as revealed by the detection of 10 to 35 ppm
Na in the eluate for the first 140 min (Fig. 4.9). When aqueous Cs was present in the influent, the Cs was sequestered by the birnessite in the first 120 min while Na was concurrently released (Fig. 4.10). The phase transformation, as discussed in more detail below, occurred after cation exchange in both experiments at low pH. This sequence is evident from the XRD patterns, and it was indicated chemically by the loss of Mn to solution after Na was leached from the structure (80-300 min without Cs, 200-300 min with Cs).
115 Phase transformations
After the cation exchange reaction occurred, the phase transformation initiated.
Following the models from Hem (1978), Silvester et al. (1997), and Drits et al. (1997),
Mn3+ from the octahedral sheet is reduced by the disproportionation of two octahedral
Mn3+ at low pH. The energy associated with this reaction in birnessite may be estimated
3+ from the standard reduction potentials for the disproportionation of Mn when Mn2O3(s)
2+ reacts to form MnO2(s) and Mn (aq) (Harris 2003):
3+ + 4+ 2+ Mn 2O3 (s) + 2H (aq) Mn O2 (s) + Mn (aq) + H2O E˚ eV (4.2)
In birnessite, Mn4+ remains in the octahedral sheet and Mn2+ is leached into the interlayer and also into solution. In our experiments involving triclinic Na-birnessite at pH 3, the concentration of Mn in the eluate was maximized (6 ppm with no Cs and 10 ppm with
Cs) at the completion of the transformation – at 120 min in the absence of Cs and at 300 min in the presence of Cs (Figs. 4.9 and 4.10). The fact that Mn release peaked once the reaction concluded is strong evidence for loss of Mn2+ from the interlayer region. The leaching of Mn from the octahedral sheet to the interlayer is supported by the change in the occupancy of the octahedral Mn from 1.00 (triclinic) to 0.593 (hexagonal), as determined from Rietveld refinement (Table 4.2). Moreover, the Mn-O octahedral bond lengths in hexagonal H- and Cs-birnessite refined to ~1.89 Å, indicating that all of the octahedral Mn was tetravalent (Table 4.2).
116 The transformation of triclinic Na-birnessite to hexagonal birnessite occurred more slowly with 0.001 M CsCl. With Cs present, the transformation did not begin until
80 min after the start of solution flow, and it continued for 220 min. As shown by our
ICP-MS analyses (Fig. 4.10), the release of interlayer Na cations to the eluate coincided with the uptake of Cs cations from solution. Initially 100 wt% of the Cs in the influx was sequestered by the powder, and after 130 min of flow, 0 wt% was removed by the powder, such that the concentration of Cs in the effluent was equivalent to the Cs concentration in the influent.
When the transformation was between 20% and 70% of completion, we measured higher concentrations of Cs in the eluate than in the influent. Consequently, excess interlayer Cs leached from the two-phase birnessite mixture, and this leaching continued at low levels even after the transformation had terminated. We conclude that triclinic birnessite has a greater capacity for the sequestration of Cs than does hexagonal birnessite, as is consistent with our results from Chapters 2 and 3. The final volume of the hexagonal Cs-birnessite formed is larger than the final volume of hexagonal H- birnessite. Thus, despite the release of Cs to the solution during the phase transition, enough Cs+ is present in the interlayer region to prop the layers apart. The occupancy of the interlayer water/cation site is greater for hexagonal Cs-birnessite (0.738) than for hexagonal H-birnessite (0.674), and this higher electron density also supports the idea that some Cs substituted into the interlayer of hexagonal birnessite. We conclude that
Cs+ is the preferred cation in the interlayer region when both protons and Cs cations are present in equivalent concentrations (0.001 M).
117 Kinetics of the triclinic-hexagonal transformations
The concentrations of triclinic H- and Cs-birnessite decrease linearly with time as these phases transform into hexagonal H- and Cs-birnessite, respectively (Fig. 4.11).
Consequently, the rate at which the phase transformation occurs can be modeled using
-kt first-order kinetics of the form A = Aoe , where Ao is the initial concentration and k is the rate constant (Fig. 4.12). Therefore, the rate laws were determined from the dependence of the natural logarithm of the concentration of triclinic birnessite [ln (%tri)] on the time since the onset of fluid flow (t), and they were calculated as follows:
For 0.001 M HCl: [%tri] = 5062.9e-0.107t (4.3)
For 0.001 M HCl, 0.001 M Cs: [%tri] = 199.5e-0.0087t (4.4)
Thus, the rate constants for the phase transformation with H+ and with Cs+ in solution are
0.107 min-1 and 0.0087 min-1, respectively, indicating that the phase transformation occurs much more rapidly when no Cs is present. We propose that the slower rate of transformation is caused by an additional reaction step. With 0.001 M Cs, Cs cation exchange occurs before the birnessite phase transformation, delaying the phase transformation:
triclinic Na-birnessite(s) triclinic s -birnessite(s) hexagonal s -birnessite(s) (4.5)
118 The occurrence of triclinic Cs-birnessite as an intermediate phase is confirmed by our analyses of the eluate, which reveal the uptake of Cs before the onset of the transition.
As the transformation proceeded, some Cs was released from the interlayer region because triclinic birnessite has a greater capacity for Cs adsorption than does hexagonal birnessite.
Discussion: Hexagonal H-birnessite reactions at pH 13
Transformation to triclinic Na-birnessite
Prior to this study, the reversibility of the triclinic-to-hexagonal birnessite reaction had not been experimentally demonstrated, even though Webb et al. (2005a) propose that this reaction characterizes the diagenesis of biogenic hexagonal birnessite in soils. Here, our TR-XRD experiments have provided clear evidence that a structural change from hexagonal to triclinic symmetry can be effected at high pH. As can be seen in our TR-
XRD data (Fig. 4.6), when hexagonal H-birnessite was reacted with solutions containing
0.1 M NaOH (pH 13), the diffraction peaks corresponding to hexagonal birnessite disappeared as triclinic birnessite reflections emerged. Rietveld analyses revealed that the unit-cell volume of hexagonal H-birnessite increased very slightly before the phase transformation, as the Na cations exchanged with smaller interlayer protons before the phase transformation occurred (Fig. 4.13A).
We infer that the final reaction product was structurally similar to the starting triclinic Na-birnessite used for our low pH experiments. Unfortunately, our ICP-MS
119 analyses of the eluate did not afford sufficient time resolution to capture the uptake of Na
(Fig. 4.14). Based on our structure refinements, the substitution of Na+ into the interlayer occurred during the first 20 min of the experiment, and our eluate collection technique did not allow for a time resolution below 20 min. Therefore, the Na uptake and release stages of the reaction took place before our first measurement, and we could not measure the amount of Na adsorbed. Nevertheless, we conclude that the interlayer cationic sites of the triclinic product were occupied by Na+ because the final unit-cell volume was equivalent to the volume for our triclinic Na-birnessite as synthesized according to
Golden et al. (1986).
Interestingly, our Rietveld analyses revealed that vacancies decreased in concentration yet persisted in the octahedral sheet of triclinic Na-birnessite (Mnoct =
0.717), suggesting that the interlayer Mn2+ of the starting hexagonal H-birnessite may have exchanged for Na+. Consequently, Mn2+ was released to the solution rather than being oxidized to Mn3+,4+ and filling all the vacancies in the octahedral sheet. This interpretation is supported by our ICP-MS analyses of the eluate, for which the Mn concentrations were steady at ~10 ppm, consistent with the release of interlayer Mn2+.
Transformation to triclinic Cs-birnessite
The transformation of hexagonal H-birnessite to triclinic Cs-birnessite during reaction with 0.1 M CsOH generated a product with significantly sharper diffraction peaks than was the case without Cs, indicating a triclinic phase with greater crystallinity.
Moreover, the increase in the unit-cell volume was much larger in the presence of Cs,
120 signifying the formation of hexagonal Cs-birnessite before the phase transformation occurred (Fig. 4.13B). This conclusion is supported by our refined value for the final triclinic Cs-birnessite unit-cell volume (108.5 Å3), which closely corresponded to the volume of Cs-exchanged triclinic Na-birnessite (107.1 Å3) (Chapter 3). In addition, the
Oint and Oint2 occupancy factors were larger than those of triclinic Na-birnessite, demonstrating that cations with greater electron densities occupied the cationic interlayer sites (i.e., Cs cations).
Again, the rapidity of the Cs cation exchange reaction interfered with our ability to capture it through ICP-MS analyses of the eluate. As with Na, Cs uptake (and release) took place within the first 20 minutes, and we measured only negative values for Cs sequestration in our eluate (Fig. 4.15). Thus, the birnessite acted as a capacitor of sorts for Cs, taking up large concentrations immediately after the start of fluid flow and then releasing Cs at concentrations 20 wt% greater than was contained in the influent over the remainder of the experiment. Unlike the situation without Cs, no Mn was detected in the eluate, indicating that no dissolution of the octahedral sheet occurred and also that
2+ interlayer Mn did not leach into solution. In fact, the Mnoct occupancy factor increased to 1.0, leading us to conclude that all of the interlayer Mn2+ was oxidized to Mn3+,4+ and migrated to fill the octahedral vacancies (Table 4.2). We expected to observe this behavior, because the rate of Mn2+ oxidation increases with higher pH since the rate is second-order with respect to hydroxide concentration (Eary and Schramke 1990).
121 Kinetics of the hexagonal-triclinic transformations
Similar to the situation at low pH, the transformation from hexagonal to triclinic symmetry at pH 13 occurred more slowly in the presence of Cs, although the disparity in rates was less marked (Fig. 4.16). Without Cs, the onset of the phase transformation began 63 min after the onset of fluid flow and persisted for 335 min. In the presence of
0.1 M CsOH, the phase transformation began at 75 min after fluid flow and endured for
350 min. We again propose that cation exchange preceded the symmetry transition to yield an intermediate phase for each reaction:
hexagonal H-birnessite(s) hexagonal Na -birnessite(s) triclinic Na -birnessite(s). (4.6) hexagonal H-birnessite(s) hexagonal s -birnessite(s) triclinic s -birnessite(s). (4.7)
The rate laws determined from the dependence of the natural logarithm of the concentration of hexagonal birnessite [ln (%hex)] on time since the onset of fluid flow (t) are as follows (Fig. 4.17):
For 0.1 M NaOH: [%hex] = 221.8e-0.0148t (4.8)
For 0.1 M CsOH: [%hex] = 100.3e-0.007t (4.9)
Thus, the transformations of hexagonal H-birnessite to triclinic Na-birnessite and triclinic
Cs-birnessite yield rate constants of 0.0148 min-1 and 0.007 min-1, respectively. These rates are more similar to each other than are the corresponding rates determined at low
122 pH, likely because both of the high pH experiments entail the exchange of a larger cation for a smaller cation before initiation of the phase transformation. Of the four experiments included in this study, only the reaction in which triclinic Na-birnessite was exposed to a solution containing 0.001 M HCl (pH 3) represented a cation exchange in which a smaller cation replaced a larger interlayer cation (H+ for Na+), which proved the fastest reaction.
Acknowledgements
The author thanks Peter Eng, Joanne Stubbs, and Nancy Lazarz at APS. Funding for this research was provided by the NSF-DOE Center for Environmental Kinetics Analysis
(CEKA) Grant No. CHE-0431328 and NSF Grant No. EAR07-45374. Use of the
Advanced Photon Source was supported by the U. S. Department of Energy, Office of
Science, Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357.
References
Al-Attar, L., and Dyer, A. (2002) Sorption behaviour of uranium on birnessite, a layered manganese oxide. Journal of Materials Chemistry, 12(5), 1381-1386. Bargar, J.R., Fuller, C.C., Marcus, M.A., Brearley, A.J., and De la Rosa, M.P. (2009) Structural characterization of terrestrial microbial Mn oxides from Pinal Creek, AZ. Geochimica Et Cosmochimica Acta, 73, 889-910. Bargar, J.R., Tebo, B.M., Bergmann, U., Webb, S.M., Glatzel, P., Chiu, V.Q., and Villalobos, M. (2005) Biotic and abiotic products of Mn(II) oxidation by spores of the marine Bacillus sp. strain SG-1. American Mineralogist, 90(1), 143-154. Barnett, M.O., Jardine, P.M., and Brooks, S.C. (2002) U(VI) adsorption to heterogeneous subsurface media: Application of a surface complexation model. Environmental Science & Technology, 36(5), 937-942. Bilinski, H., Giovanoli, R., Usui, A., and Hanzel, D. (2002) Characterization of Mn oxides in cemented streambed crusts from Pinal Creek, Arizona, U.S.A., and in
123 hot-spring deposits from Yuno-Taki Falls, Hokkaido, Japan. American Mineralogist, 87, 580-591. Burke, K.E. (1970) Study of Scavenger Properties of Manganese(IV) Oxide with Atomic Absorption Spectrometry - Determination of Microgram Quantities of Antimony, Bismuth, Lead, and Tin in Nickel. Analytical Chemistry, 42(13), 1536-1540. Burns, R.G. (1976) Uptake of Cobalt into Ferromanganese Nodules, Soils, and Synthetic Manganese(Iv) Oxides. Geochimica Et Cosmochimica Acta, 40(1), 95-102. Cho, Y., Jang, S., Kim, Y., Komarneni, S., and Kim, S. (2011) Uptake of cadmium, copper, and lead by microporous synthetic Na-birnessite. Journal of Porous Materials, 18(1), 125-131. Chukhrov, F.V., Sakharov, B.A., Gorshkov, A.I., Drits, V.A., and Dikov, Y.P. (1985) Crystal structure of birnessite from the Pacific Ocean. International Geology Review, 27, 1082-1088. Drits, V.A., Silvester, E., Gorshkov, A.I., and Manceau, A. (1997) Structure of synthetic monoclinic Na-rich birnessite and hexagonal birnessite.1. Results from X-ray diffraction and selected-area electron diffraction. American Mineralogist, 82(9- 10), 946-961. Eary, L.E., and Ral, D. (1987) Kinetics of chromium (III) oxidation to chromium (VI) by reaction with manganese dioxide. Environmental Science & Technology, 21, 1187-1193. Eary, L.E., and Schramke, J.A. (1990) Rates of Inorganic Oxidation Reactions Involving Dissolved Oxygen. Chemical Modeling of Aqueous Systems II, 416, p. 379-396. American Chemical Society, Washington D.C. Finger, L.W., Cox, D.E., and Jephcoat, A.P. (1994) A Correction for Powder Diffraction Peak Asymmetry Due to Axial Divergence. Journal of Applied Crystallography, 27, 892-900. Fredrickson, J.K., Zachara, J.M., Kennedy, D.W., Kukkadapu, R.K., McKinley, J.P., Heald, S.M., Liu, C., and Plymale, A.E. (2004) Reduction of TcO4- by sediment- associated biogenic Fe(II). Geochimica et Cosmochimica Acta, 68(15), 3171- 3187. Frierdich, A.J., Hasenmueller, E.A., and Catalano, J.G. (2011) Composition and structure of nanocrystalline Fe and Mn oxide cave deposits: Implications for trace element mobility in karst systems. Chemical Geology, 284(1-2), 82-96. Fu, G., Allen, H.E., and Cowan, C.E. (1991) Adsorption of cadmium and copper by manganese oxide. Soil Science, 152(2), 72-81. Gee, G.W., Oostrom, M., Freshley, M.D., Rockhold, M.L., and Zachara, J.M. (2007) Hanford site vadose zone studies: An overview. Vadose Zone Journal, 6(4), 899- 905. Giovanoli, R., Stahli, E., and Feitknec.W. (1970) On Oxide-Hydroxides of Manganese (4) with Stratified Lattices. 2. Manganese (3)-Manganate(4). Helvetica Chimica Acta, 53(3), 453-464. Glover, E.D. (1977) Characterization of a Marine Birnessite. American Mineralogist, 62(3-4), 278-285. Golden, D.C., Chen, C.C., and Dixon, J.B. (1986) Synthesis of Todorokite. Science, 231(4739), 717-719.
124 Hem, J.D. (1978) Redox Processes at Surfaces of Manganese Oxide and Their Effects on Aqueous Metal-Ions. Chemical Geology, 21(3-4), 199-218. House, J.E. (1997) Principles of Chemical Kinetics. 244 p. Wm. C. Brown Publishers, Dubuque, IA. Kang, D.H., Schwab, A.P., Johnston, C.T., and Banks, M.K. (2010) Adsorption of iron cyanide complexes onto clay minerals, manganese oxide, and soil. Journal of Environmental Science and Health Part A-Toxic/Hazardous Substances & Environmental Engineering, 45(11), 1391-1396. Lanson, B., Drits, V.A., Gaillot, A.C., Silvester, E., Plancon, A., and Manceau, A. (2002) Structure of heavy-metal sorbed birnessite: Part 1. Results from X-ray diffraction. American Mineralogist, 87(11-12), 1631-1645. Lanson, B., Drits, V.A., Silvester, E., and Manceau, A. (2000) Structure of H-exchanged hexagonal birnessite and its mechanism of formation from Na-rich monoclinic buserite at low pH. American Mineralogist, 85(5-6), 826-838. Larson, A.C., and Von Dreele, R.B., 1994. GSAS-General Structure Analysis System. Los Alamos National Laboratory, Le Goff, P., Baffier, N., Bach, S., and Pereira-Ramos, J.P. (1996) Synthesis, ion exchange and electrochemical properties of lamellar phyllomanganates of the birnessite group. Materials Research Bulletin, 31(1), 63-75. Lind, C.J., and Hem, J.D. (1993) Manganese minerals and associated fine particulates in the streambed of Pinal Creek, Arizona, U.S.A.: a mining-related acid drainage problem. Applied Geochemistry, 8, 67-80. Lopano, C.L., Heaney, P.J., Post, J.E., Hanson, J., and Komarneni, S. (2007) Time- resolved structural analysis of K- and Ba-exchange reactions with synthetic Na- birnessite using synchrotron X-ray diffraction. American Mineralogist, 92(2-3), 380-387. Manceau, A., Drits, V.A., Silvester, E., Bartoli, C., and Lanson, B. (1997) Stuctural mechanism of Co2+ oxidation by the phyllomanganate buserite. McBride, M.B. (1994) Environmental Chemistry of Soils. Oxford University Press, New York. McKenzie, R.M. (1977) Manganese Oxides and Hydroxides. In J.B. Dixon, and S.B. Weed, Eds. Minerals in Soil Environments, p. 181-193. Soil Science Society of America, Madison, WI. -. (1980) Adsorption of Lead and Other Heavy-Metals on Oxides of Manganese and Iron. Australian Journal of Soil Research, 18(1), 61-73. McKinley, J.P., Zeissler, C.J., Zachara, J.M., Serne, R.J., Lindstrom, R.M., Schaef, H.T., and Orr, R.D. (2001) Distribution and retention of Cs-137 in sediments at the Hanford Site, Washington. Environmental Science & Technology, 35(17), 3433- 3441. Post, J.E., Heaney, P.J., and Ertl, A. (2008) Rietveld refinement of the rancieite structure using synchrotron powder diffraction data. Powder Diffraction, 23(1), 10-14. Post, J.E., Heaney, P.J., and Hanson, J. (2002) Rietveld refinement of a triclinic structure for synthetic Na-birnessite using synchrotron powder diffraction data. Powder Diffraction, 17(3), 218-221.
125 Rietveld, H.M. (1969) A Profile Refinement Method for Nuclear and Magnetic Structures. Journal of Applied Crystallography, 2, 65-71. Silvester, E., Manceau, A., and Drits, V.A. (1997) Structure of synthetic monoclinic Na- rich birnessite and hexagonal birnessite. 2. Results from chemical studies and EXAFS spectroscopy. American Mineralogist, 82(9-10), 962-978. Stephens, P.W. (1999) Phenomenological model of anisotropic peak broadening in powder diffraction. Journal of Applied Crystallography, 32, 281-289. Tan, H., Zhang, G.X., Heaney, P.J., Webb, S.M., and Burgos, W.D. (2010) Characterization of manganese oxide precipitates from Appalachian coal mine drainage treatment systems. Applied Geochemistry, 25(3), 389-399. Taylor, R.M., McKenzie, R.M., and Norrish, K. (1964) The mineralogy and chemistry of manganese in some Australian soils. Australian Journal of Soil Research, 2, 235- 248. Thompson, P., Cox, D.E., and Hastings, J.B. (1987) Rietveld Refinement of Debye- Scherrer Synchrotron X-Ray Data from Al2O3. Journal of Applied Crystallography, 20, 79-83. Toby, B.H. (2001) EXPGUI, a graphical user interface for GSAS. Journal of Applied Crystallography, 34, 210-213. Vaniman, D.T., Chipera, S.J., Bish, D.L., Duff, M.C., and Hunter, D.B. (2002) Crystal chemistry of clay-Mn oxide associations in soils, fractures, and matrix of the Bandelier Tuff, Pajarito Mesa, New Mexico. Geochimica Et Cosmochimica Acta, 66(8), 1349-1374. Wall, A.J., Heaney, P.J., Mathur, R., Post, J.E., Hanson, J.C., and Eng, P.J. (2011) A flow-through reaction cell that couples time-resolved X-ray diffraction with stable isotope analysis. Journal of Applied Crystallography, 44, 429-432. Wan, J.M., Larsen, J.T., Tokunaga, T.K., and Zheng, Z.P. (2004) pH neutralization and zonation in alkaline-saline tank waste plumes. Environmental Science & Technology, 38(5), 1321-1329. Webb, S.M., Tebo, B.M., and Bargar, J.R. (2005a) Structural characterization of biogenic Mn oxides produced in seawater by the marine bacillus sp strain SG-1. American Mineralogist, 90(8-9), 1342-1357. -. (2005b) Structural influences of sodium and calcium ions on the biogenic manganese oxides produced by the marine Bacillus sp., strain SG-1. Geomicrobiology Journal, 22(3-4), 181-193. Zhu, M.Q., Ginder-Vogel, M., Parikh, S.J., Feng, X.H., and Sparks, D.L. (2010) Cation Effects on the Layer Structure of Biogenic Mn-Oxides. Environmental Science & Technology, 44(12), 4465-4471.
126 Figures
Figure 4.1. The crystal structure of triclinic Na-birnessite with Mn3+/4+ in all octahedral positions with hydrated Na+ in the interlayer positions, adapted from Post et al. (2002).
127
Figure 4.2. The crystal structure of hexagonal H-birnessite with Mn4+ occupying most of 2+/3+ the octahedral positions. Mn and H(H2O)x are within the interlayer. Adapted from Silvester et al. (1997).
128
Figure 4.3. Flow-through TR-XRD experimental setup at beam line 13-BM-C at APS.
129
Figure 4.4. TR-XRD patterns of triclinic Na-birnessite transforming to hexagonal H- birnessite in 0.001 M HCl.
130
Figure 4.5. Triclinic Na-birnessite transforms to hexagonal Cs-birnessite in 0.001 M CsCl at pH 3.
131
Figure 4.6. Hexagonal H-birnessite begins to transform to triclinic Na-birnessite in 0.1 M NaOH.
132
Figure 4.7. Hexagonal H-birnessite transforms to triclinic Cs-birnessite in 0.1 M CsOH at pH 13.
133
Figure 4.8. Changes in volume during pH 3 reactions in (A) 0.001 M HCl and (B) 0.001 M CsCl and 0.001 M HCl.
134
Figure 4.9. Changes in Na and Mn concentration over the course of the reaction of triclinic Na-birnessite at pH 3.
135
Figure 4.10. Changes in Cs, Na, and Mn concentration over the course of the reaction of triclinic Na-birnessite exchanged with 0.001 M CsCl at pH 3.
136
Figure 4.11. The change in percentage of triclinic Na-birnessite present as a function of time at pH 3: (A) 0.001 M HCl and (B) 0.001 M HCl and 0.001 M CsCl.
137
Figure 4.12. The rate of triclinic Na-birnessite exchanged with (A) 0.001 M HCl and (B) 0.001 M CsCl and 0.001 M HCl, both buffered to pH 3, follow first-order kinetics.
138
Figure 4.13. Changes in volume during pH 13 reactions in (A) 0.1 M NaOH and (B) 0.1 M CsOH.
139
Figure 4.14. Changes in Na and Mn concentration over the course of the reaction of hexagonal H-birnessite exchange with 0.1 M NaOH at pH 13.
140
Figure 4.15. Percentage of Cs adsorbed to hexagonal H-birnessite exchanged with 0.1 M CsOH (pH 13). Na and Mn concentrations were below detection limits.
141
Figure 4.16. The change in percentage of hexagonal H-birnessite present as a function of time at pH 13: (A) 0.1 M NaOH and (B) 0.1 M CsOH.
142
Figure 4.17. The rate of hexagonal H-birnessite exchanged with (A) 0.1 M NaOH and (B) 0.1 M CsOH at pH 13 follow first-order kinetics.
143 Tables
Table 4.1. Rietveld refinement parameters for the final transformation structures.
hexagonal hexagonal triclinic Na- triclinic Cs- H-birnessite Cs-birnessite birnessite birnessite pH 3 3 13 13 Space group P -3 P -3 C -1 C -1
Un it cell a (Å) 2.84296(4) 2.84657(4) 4.96983(0) 5.04417(8) b (Å) 2.93450(8) 2.90777(4) c (Å) 7.31278(9) 7.38267(4) 7.16424(5) 7.45523(5) α (°) 88.158(9) 88.218(5) β (°) 95.960(1) 96.924(0) γ (°) 91.412(6) 89.479(8) V (Å3) 51.1866(1) 51.807(1) 103.841(1) 108.489(4)
Refinement No. of diffraction points 1606 1567 1607 1606 No. of reflections 27 27 86 75
Diffraction range (° θ) 5.5 -37.5 5.5 -37.5 5.5 -37.5 5.5 -37.5 No. of variables 30 27 33 34 R(F2) 0.0064 0.0092 0.0147 0.0115
Rwp 0.0201 0.019 0.0221 0.152 χ2 2.732 2.374 2.850 1.516
144 Table 4.2. Atomic coordinates and isotropic displacement factors for transformed birnessite in the absence and presence of Cs at pH 3 and 13.
Site U x 102 Structure Atom x y z occupancy iso (Å2) factor hexagonal H- Mn 0 0 0 0.593(6) 1.500 birnessite oct
pH 3 Ooct 0.333333 0.666667 0.12846(2) 1.00 2.477
Oint 0.666667 0.333333 0.47578(2) 0.674(0) 8.000
Mnint 0 0 0.72490(9) 0.181(8) 1.500
hexagonal Cs- Mn 0 0 0 0.823(3) 1.10(0) birnessite oct
pH 3 Ooct 0.3333 0.6667 0.12778(4) 1.00 1.97(2)
Oint 0.6667 0.3333 0.45601(0) 0.738(7) 17.65(8)
Mnint 0 0 0.71723(9) 0.127(1) 0.77(3)
triclinic Na- Mn 0 0 0 0.717(4) 2.12(8) birnessite oct
pH 13 Ooct 0.35442(4) -0.01201(3) 0.14681(6) 1.00 2.06(4)
Oint 0.62111(4) 0.16694(0) 0.63657(2) 0.696(1) 8.20(2)
Oint2 1.11292(3) 0.80758(5) 0.45330(1) 0.358(7) 0.600
triclinic Cs- Mn 0 0 0 0.995(4) 0.52(7) birnessite oct
pH 13 Ooct 0.35539(8) -0.04426(2) 0.13693(6) 1.00 0.83(8)
Oint 0.58149(5) 0.31583(2) 0.52689(1) 0.674(9) 6.14(6)
Oint2 0.14064(1) 0.23032(8) 0.57756(8) 0.726(9) 1.01(9)
145 Table 4.3. Bond distance values for transformed birnessite in the absence and presence of Cs at pH 3 and 13.
hexagonal H- hexagonal Cs- triclinic Na- triclinic Cs- Bond birnessite birnessite birnessite birnessite
Mnoct-Mnoct 2.8430(4) x 6 2.84657(28) x 6 2.9345(8) x 2 2.9226(15) x 2
Mnoct-Mnoct 2.8544(26) x 2 2.8997(15) x 2
Mnoct-Mnoct 2.9167(26) x 2 2.9078(9) x 2
Mnoct-Ooct 1.891(6) x 6 1.895(4) x 6 1.956(9) x 2 1.954(11) x 2
Mnoct-Ooct 1.972(11) x 2 1.891(9) x 2
Mnoct-Ooct 1.997(9) x 2 2.044(12 x 2
Mnoct-Mnint 2.012(14) x 2 2.088(14) x 2
Ooct-Mnint 1.961(4) x 3 2.003(6) x 3
Ooct-Oint 1.596(18) 2.340(18)
Oint-Oint 1.679(10) x 3 1.767(18) x 3
Oint-Mnint 2.452(18) x 3 2.534(26) x 3
Oint-Mnint 2.202(21) x 3 2.083(13) x 3
Oint-Oint2 1.360(28) 1.296(26)
Oint-Oint2 1.53(4) 1.301(25)
Oint-Oint2 1.878(29) 1.748(21)
Oint-Oint2 2.004(29) 2.13(5)
Oint2-Oint2 1.50(6) 1.69(5)
Oint2-Oint2 1.80(7) 2.21(4)
146 Chapter 5 Reduction of triclinic Na-birnessite by Mn2+ cation exchange
Abstract
Triclinic Na-birnessite transforms to hexagonal H-birnessite in an aqueous solution at pH
3 due to the reduction of octahedral Mn (Sylvester et al. 1997; Chapter 4 of this thesis), and Fischer et al. (in prep) have demonstrated that aqueous Cr3+ induces a transition from triclinic to hexagonal birnessite during the oxidation to Cr6+. Whether aqueous Mn2+ similarly promotes the reduction of Mn3+,4+ in birnessite was not determined by previous studies, but this reaction is important as it bears on the cycling of Mn in soils and on the reversibility of birnessite redox reactions. We used in situ time-resolved X-ray diffraction (TR-XRD) and inductively coupled plasma-mass spectrometry (ICP-MS) to characterize the phase transformation of triclinic Na-birnessite as a function of pH and
2+ 2+ [Mn (aq)]. These data demonstrate that Mn does initially exchange into the interlayer region of triclinic Na-birnessite at pH 3, 4.5, and 6. The rate of cation exchange is: 1)
2+ dependent on pH for 0.001 and 0.01 M but not 0.1 M MnCl2; and 2) dependent on [Mn ] at pH 6. Following cation exchange at pH 3, the aqueous Mn2+ catalyzed the transition from triclinic to hexagonal birnessite, but at pH 4.5 and 6, the octahedral Mn was reduced
3+ to Mn (s), instigating a structural breakdown of the triclinic birnessite.
147 Introduction
Although dissolved manganese occurs in low concentrations in most groundwaters, manganese oxides typically constitute 0.4 to 1.1 wt% of pedalfers, and they contribute strongly to soil ion exchange and redox processes (Jenne 1968; McKenzie
1977; Post 1999; O'Reilly and Hochella 2003; Weaver and Hochella 2003; Negra et al.
2005a; Negra et al. 2005b; Drits et al. 2007). Birnessite is the most commonly occurring layered manganese oxide, and it occurs as submicron grains with large surface areas
(Taylor et al. 1964; Barnett et al. 2002; Fredrickson et al. 2004). Synthetic birnessite exhibits either triclinic or hexagonal symmetry (Silvester et al. 1997; Lanson et al. 2002a;
Post et al. 2002), and natural birnessite-like phases have been reported with hexagonal or pseudo-orthogonal symmetry (Webb et al. 2005a; Webb et al. 2005b; Post et al. 2008;
Tan et al. 2010; Santelli et al. 2011). Triclinic Na-birnessite has the chemical formula
4+ 3+ 3+/4+ Na0.56(Mn 1.44Mn 0.56)O4• H2O and is composed of Mn O6 octahedral sheets that are negatively charged due to the presence of Mn3+ rather than Mn4+. The octahedral sheets are charge balanced by hydrated Na cations that are positioned halfway between the sheets, and these interlayer cations are highly exchangeable (Fig. 5.1) (Post et al.
2002).
The structure of hexagonal birnessite differs from that of triclinic birnessite in the near-absence of Mn3+ and the presence of Mn vacancies (□) in the octahedral sheet,
3+ 2+ 4+ 3+ resulting in the chemical formula H0.33Mn 0.111Mn 0.055(Mn 0.722Mn 0.111□0.167)O2 (Fig.
5.2) (Silvester et al. 1997). The presence of vacancies in place of Mn4+ results in negatively charged sheets, which are charge balanced by Mn2+ and hydrated cations
148 between the sheets. To the best of our knowledge, prior studies of cation exchange in the phyllomanganate system have focused on triclinic birnessite.
The chemical properties of birnessite are determined by its structural state, but studies over the last decade have demonstrated that birnessite will symmetrically transform depending on solution pH, ionic strength, and biological environment. For example, when Mn2+ was incubated with spores of the marine Bacillus sp., strain SG-1,
Na cations promoted the bacterial precipitation of hexagonal birnessite whereas Ca cations favored pseudo-orthogonal birnessite phases (Webb et al. 2005b). In a separate study, Webb et al. (2005a) showed that initially precipitated hexagonal birnessite transformed to pseudo-orthogonal birnessite within 3 days. Sylvester et al. (1997) first showed that at pH 3, triclinic birnessite transforms to hexagonal birnessite, and in
Chapter 4 of this dissertation, we demonstrated that the reverse reaction occurs at pH 13.
Here we explore a phase transformation instigated by Mn itself. Fischer et al. (in prep) have demonstrated that dissolved Cr3+ will strongly accelerate the transformation of triclinic Na-birnessite to hexagonal symmetry. The authors explained this transition through a coupled redox reaction: Cr3+ oxidizes to Cr6+ as octahedral Mn3+ in birnessite reduces to Mn2+, which moves from an octahedral site to the interlayer, leaving behind an octahedral vacancy. This observation raises the question of whether aqueous Mn2+ will similarly induce a symmetry-changing transformation of triclinic to hexagonal birnessite.
In the process, does Mn2+ oxidize to Mn3+/4+ and precipitate as triclinic birnessite, or does
Mn2+ simply exchange into the interlayer and stabilize hexagonal birnessite? Answers to these questions will assist scientists in understanding the cycling of Mn between soils and groundwaters.
149 Several geochemists have previously explored the exchange of dissolved Mn2+
2+ with Mn oxides (MnOx). Fendorf et al. (1992; 1993) monitored Mn exchange reactions with colloids that had nonspecific birnessite-like structures using in situ electron paramagnetic resonance microscopy (EPR). They concluded that even low concentrations of Mn2+ led to complete surface adsorption in the first 600 ms of the reaction. However, this method did not distinguish between surface adsorption and cation exchange reactions, and no solid state characterization of the final birnessite structure was performed. Novikov et al. (2006) applied TEM, XRD, and IR to monitor
2+ changes in crystal structure during Mn uptake by a variety of MnO2 structures, and they found that Mn2+ does constitute an “exchangeable complex” in birnessite and that more
Mn is exchanged as Mn(aq) concentration increases. Again, however, the final crystal structures of the birnessite phases were not determined after the cation exchange reaction.
Long-term experiments that targeted the effects of dissolved Mn2+ on hexagonal birnessite revealed a wholesale structural transformation to a variety of Mn oxy- hydroxides, whose identity depended on the starting solution pH and Mn concentration
(Tu et al. 1994; Elzinga 2011). At neutral pH, the hexagonal birnessite transformed into feitknechtite (-MnOOH) before evolving into manganite (-MnOOH).
For our study, we pursued flow-through in situ Mn2+-exchange reactions with triclinic Na-birnessite, and we continuously analyzed the solid and aqueous states using time-resolved synchrotron X-ray diffraction (TR-XRD) and inductively coupled plasma- mass spectrometry (ICP-MS). The flow-through reactions occurred at a rate where the initial stages of phase transformation could be monitored, and we were able to capture the
Mn2+ exchange before the triclinic birnessite began to transform. This investigation
150 provides deeper insights into the mechanisms of birnessite phase transformations and the dependence of Mn oxide phases on the local chemical environment.
Experimental section
Material synthesis
Triclinic Na-birnessite was synthesized according to the procedure described in
Golden et al. (1986). At room temperature, oxygen was bubbled (4.2 L/min) through a mixture of chilled 250 mL 5.5 M NaOH and 200 mL 0.5 M MnCl2 for five hours. The precipitate was divided evenly between 12 centrifuge tubes and centrifuged. The solution was poured off and replaced with fresh DI water; after five replicate rinsing cycles, the birnessite was stored in fresh DI water until needed for experiments. Aliquots of the stored triclinic Na-birnessite were filtered through a μm hydrophilic polypropylene membrane filter (GH Polypro, Pall) and allowed to air dry. The dry triclinic Na- birnessite (250 mg) was ground under acetone and initially characterized with a conventional sealed Mo tube source on a Rigaku II D/MAX-RAPID microdiffractometer
(Materials Characterization Laboratory, Pennsylvania State University).
Chemical analysis
TR-XRD experiments were completed at beam line 13-BM-C at the Advanced
Photon Source (APS) at Argonne National Laboratory (ANL). Approximately 4 mg of triclinic Na-birnessite were packed between two cotton plugs in a 0.7 mm quartz capillary
151 and attached to a flow-through apparatus (Fig. 5.3) (Wall et al. 2011). The triclinic Na- birnessite interacted with nine different manganese (II) chloride solutions that varied in pH and [Mn2+] (pH 3, 4.5, and 6; [Mn2+] = 0.001 M, 0.01 M, and 0.1 M). The solutions were buffered to pH 3, 4.5, and 6 with 0.1 M HCl or 0.1 M NaOH. A batch reaction of
100 mL 0.1 M MnCl2 buffered to pH 6 and 100 mg triclinic Na-birnessite was reacted for
24 hr and also characterized with a conventional sealed Mo tube source on a Rigaku II
D/MAX-RAPID microdiffractometer.
TR-XRD patterns were collected every 60 seconds for 5-9 hours using a MAR165
CCD detector. The solution flow rate was 1 drop/40 sec (~ 0.075 mL/min). The sample- to-collector distance was mm, allowing us to collect over a θ range of -38°.
The wavelength was 0.8321 Å. Rietveld refinements (Rietveld 1969) were completed for each TR-XRD pattern using the EXPGUI interface (Toby 2001) of the general structure analysis system (GSAS) (Larson and Von Dreele 1994). The background intensities for the TR-XRD patterns were fit using up to 12 terms of a shifted Chebyschev function.
The peak profiles were modeled by a pseudo-Voigt profile function as parameterized by
Thompson et al. (1987), with asymmetry correction by Finger et al. (1994), and microstrain anisotropic broadening terms of Stephens (1999).
During initial cycles of refinement only the background, scale, peak profile, and unit-cell parameters were allowed to vary. The position of the O atom in the Mn-O sheet was then refined. Following refinements of the interlayer atom position, the occupancy factor and isotropic atomic displacement factor for the interlayer site were allowed to vary. Once the interlayer atom was settled, the isotropic thermal parameters for the Mn and O in the octahedral sheets, followed alternately with that for the interlayer atom, were
152 refined. The final χ2 values for the Mn end-member birnessite refinements before structural breakdown occurred ranged from 0.9526-2.794 (Table 5.1).
The leachate was collected as a function of time for the cation exchange experiments: a stepping motor controlled a UR-150 Newport Rotary Stage with an Al tray capable of supporting 15 polypropylene vials. The stepping motor was programmed to rotate to the next vial every 20 min. The leachate was analyzed with an X-Series 2
Thermo Scientific quadrupole inductively coupled plasma-mass spectrometer (ICP-MS) in conjunction with Thermo Scientific PlasmaLab software (Materials Characterization
Laboratory, Pennsylvania State University) to determine the change in Na and Mn elemental concentrations.
Cation exchange coefficients
The adsorption and desorption of Mn and Na are best described by a cation exchange coefficient (Kex) which takes into account the mass of initial triclinic Na- birnessite within the capillary:
(5.1)
where
- - (5.2)
- - . (5.3)
[Mn]i and [Na]i are the concentrations of Mn and Na in ppm at time = i min (ti), respectively. [Mn]i+1 and [Na]i+1 are the concentrations of Mn and Na in ppm in the
153 proceeding time fractionation (ti+1), respectively, while Q is the flow rate in mL/min and g is the mass of triclinic birnessite packed within the capillary. After appropriate unit conversions (i.e., division by 106), the mass of Mn or Na (g) is then divided by the respective atomic molar mass to achieve mole amounts of each cation. Σ Mn and Σ Na represent the net mole amounts of Mn and Na that are adsorbed (Σ > 0) or desorbed (Σ
< 0) over the course of each reaction. The desorption of Na is used to monitor the amount of Mn exchanged into the interlayer since these cations undergo cation exchange in the interlayer region of triclinic Na-birnessite (Morales et al. 1990). A cation exchange
2+ reaction would ideally result in Kex = 0.5 because 1 mol Mn would adsorb for every 2 mol Na+ released to maintain charge neutrality.
Kinetic analysis
One-stage cation exchange reactions
The volume of the unit cell is used as a proxy for the extent of cation exchange based on Vegard’s Law Vegard’s Law (Denton and Ashcroft 1991) states that the total volume of the unit cell (Vtot) is the summation of the individual phase fractions (Xx) each multiplied by their respective unit cell volumes, q and s:
(5.4) where XA is the phase fraction of initial triclinic Na-birnessite, A, and XC is the phase fraction of triclinic M-birnessite, C. The rate of a one-stage reaction is determined with standard first-order reaction kinetics:
154
A (5.5)
- e (5.6)
where k is the rate constant, t is time, and is the initial phase fraction of starting material, A. From mass balance,
(5.7) and simplifying Eqn. 5.6 and 5.7:
- - e (5.8) we can invoke Vegard’s Law (Eqn ) to determine the dependence of Vtot on t:
- - e ( - e ) (5.9) which simplifies to
- - e . (5.10)
The dependence of volume versus time for each pH is fit with Eqn. 5.10 using Origin 6.1, solving for q, s, and k. These three variables were allowed to iterate simultaneously using the Levenberg-Marquardt algorithm until the reduced χ2 was minimized. The initial rate in terms of XA, XC, and k can be defined with standard first-order reversible kinetics as
(Eqn. 5.5) (Lasaga 1981)
- (5.11)
(5.12)
where n is the reaction order Taking the derivative of Vegard’s Law (Eqn ) with respect to time,
, (5.13)
155
and substituting Eqns. 5.11 and 5.12 into Eqn. 5.13, the rate in is min
lim - (5.14) , or
- . (5.15)
The log of the rate is plotted versus pH, resulting in a linear regression where the negative slope is the reaction order, n, and the y-intercept is the log of the initial rate
constant, w, with units M min :
log - log , (5.16)
which simplifies to
. (5.17)
Two-stage cation exchange reactions
Lopano et al. (2009) have shown that Cs+ exchange into triclinic Na-birnessite occurs as a two-stage linear reaction where the first stage involves the complete exchange of cations and the second stage is the reordering of cations within the interlayer to more thermodynamically stable positions. The cation exchange reaction still can be modeled using Vegard’s Law (Eqn. 5.4), which is then written as a summation of three phases and their respective unit cell volumes, q, r, and s:
(5.18)
156 where XA is the phase fraction of initial triclinic Na-birnessite, XB is the phase fraction of disordered triclinic M-birnessite, and XC is the phase fraction of ordered triclinic M- birnessite. The rate constant of each reaction stage (k1 and k2, respectively, with units min-1) is determined according to standard kinetic equations for a first order compound reaction (Fromherz 1964):
A (5.19)
- e (5.20)
- - (e -e ) (5.21) -
- - e e - - . (5.22) -
According to mass balance:
. (5.23)
Eqns. 5.20- are substituted into Vegard’s Law (Eqn ):
e- e- - - - e e -e - - . (5.24) - -
The dependence of volume versus time for each pH is fit with Eqn. 5.24 using Origin 6.1, solving for q, r, s, k1, and k2. These five variables were allowed to iterate simultaneously using the Levenberg-Marquardt algorithm until the reduced χ2 was minimized. At this time, we are most interested in the initial exchange of cations when XA = 1 and t = 0, and not the reordering of cations; therefore, the initial rate of the cation exchange reaction is determined using the first-stage rate constant (k1) and phase fractions XA and XB. The initial rate in terms of XA, XB, and k1 can be defined with standard first-order reversible kinetics as (Lasaga 1981)
157
- (5.25)
. (5.26)
Taking the derivative of Vegard’s Law (Eqn ) in terms of the initial exchange of cations with respect to time,
, (5.27)
and substituting Eqns. 5.25 and 5.26 into Eqn. 5.27, the initial rate in is min
lim - (5.28) , or
- . (5.29)
The initial rate is then plotted against pH according to Eqns. 5.16 and 5.17, allowing us to determine how the solution pH affects the rate of Mn cation exchange measured from volume changes of the triclinic Na-birnessite unit cell. The effect of [Mn] on the exchange rate is also determined with Eqns 5.16 and 5.17, where log(dV/dt) is plotted against -log([Mn]), or pMn. A more in-depth discussion of the kinetic models can be found in Lopano et al. (2011).
Phase transformation reactions
When triclinic and hexagonal phases of birnessite were both present, we were unable to complete a Rietveld refinement for either phase because the structure parameters were too similar. Instead, we quantified the changing ratio of the
158 triclinic/hexagonal phases during the transition by selecting the refined structures for each phase nearest the transition and then modeling the XRD patterns of a 1:1 mixture of these structures using CrystalDiffract 1.4.0 (CrystalMaker, Inc). From this mixture, we extracted relative intensity ratios (RIR), which allowed us to determine the evolution of each phase fraction during the reaction. Since the densities of the triclinic and hexagonal phases were the same within error, a 1:1 mixture by volume was equivalent to a 1:1 mixture by mass. Consequently, the volume-based RIR ratios yielded by CrystalDiffract were not further corrected to achieve a conventional mass-based RIR. For each pattern collected during the symmetry-changing transitions, we decomposed overlapping (002) peaks for the triclinic and hexagonal phases using Jade 6.5.23 (Materials Data, Inc.), integrated the areas of the component peaks, and then calculated a mass ratio normalized by the RIR. This procedure yielded the relative mass abundance of the triclinic and hexagonal phases as a function of time. The rates of the phase transformations were modeled using standard zero- or first-order kinetic equations (House 1997).
Results and Discussion
As can be observed from the stacked TR-XRD patterns, triclinic Na-birnessite transformed to hexagonal H-birnessite at all Mn2+ concentrations at pH 3 (Fig. 5.4) and at
2+ 0.01 M MnCl2 at pH 4.5. The remaining solutions of varying [Mn ] at pH 4.5 and pH 6 induced a partial structural transformation of the initial triclinic Na-birnessite after Mn2+ cation exchange occurred (Figs. 5.5 and 5.6). Refinement results for the end-member
159 triclinic and hexagonal birnessite phases are compiled in Table 5.1. Atomic coordinates are listed in Table 5.2 and a summary of bond distances are shown in Table 5.3.
Unit-cell variations in triclinic birnessite
The unit-cell volume of triclinic Na-birnessite decreased for all pH values and Mn concentrations as the reaction proceeded (Fig. 5.7-5.9). The ionic radius of Mn2+ is 0.89
Å while that for Na+ is 1.16 Å, suggesting that the substitution of Mn2+ for interlayer Na+ should shorten the c-axis. Moreover, Morales et al. (1990) have reported that the reverse reaction, in which interlayer Mn2+ was replaced by Na+, caused an increase in unit-cell volume. Therefore, we argue that a decrease in unit-cell volume is evidence that Mn2+ has exchanged with interlayer Na+. At pH 3, the triclinic unit-cell volume decreased from
3 3 3 106.3 ± 0.27 Å to 104.0 ± 0.12 Å at 0.001 and 0.01 M MnCl2, and 104.6 ± 0.02 Å at
0.1 M MnCl2 before a hexagonal birnessite phase appeared (Fig. 5.7). When the Mn solution was buffered to pH 4.5 or 6, the triclinic unit-cell volume diminished even more, to between 102 and 103 Å3 (Figs. 5.8 and 5.9).
Kinetic analyses based on unit-cell variations
Our kinetic analyses suggest that Mn2+ and H+ actively competed for interlayer sites within triclinic Na-birnessite. Moreover, we conclude that the experimental parameters selected for this study extended from systems that were H-controlled to those that were Mn-controlled. We base this interpretation on the dependence of the initial
160 reaction rates with pH and Mn concentration. As noted above, we used the change in unit-cell volume as a proxy for the degree of reaction progress – in this case, the degree of substitution of Mn or H into the interlayer. Our rate constants were derived from kinetic modeling of the total change in unit-cell for each experimental condition. All reactions but one were best modeled as a one-stage cation exchange reaction; the only exception was the reaction that involved 0.1 M MnCl2 at pH 6, which was best modeled as a two-stage cation exchange reaction.
Based on previous cation exchange studies presented in this dissertation and by prior researchers (Lopano et al. 2011; Fischer et al. in prep), the rate of cation exchange within birnessite is strikingly well-captured as a simple first-order reaction: dV/dt = w[H+]n, so that log(dV/dt) varies linearly with pH (or -log[Mn2+]) (Eqns. 5.16 and 5.17).
In this work (Fig. 5.10), we observed this linear dependence to be obeyed in two sets of experiments: 1) those in which low concentrations (0.001 and 0.01 M) of Mn2+ were examined as a function of pH, a condition that we designate as H-controlled; and 2) those in which a low concentration of H+ (pH 6) was examined as a function of [Mn2+], a condition that we designate as Mn-controlled. In the remaining systems, the ratios
[Mn2+]:[H+] were intermediate between these extremes, and the dependence of log(dV/dt) on -log[Mn2+] or pH departed sharply from linearity (Table 5.4).
For those lower concentrations of Mn2+ (0.001 M and 0.01 M) that fell within the
H-controlled regime, the derived rate laws were as follows (Fig. 5.10A):
dV 2 0.001 M MnCl2: H R = 0.995 (5.30) dt
161
dV 2 0.01 M MnCl2: H R = 0.991 (5.31) dt
For the low concentration of H+ (pH 6) that fell within the Mn-controlled regime, the calculated rate law indicated that the cation exchange rate increased as [Mn2+] increased
(Fig. 5.10B):
dV pH 6: Mn R2 = 0.996 (5.32) dt
At the highest Mn concentration (0.1 M), Mn outpopulated the dissolved protons by at least two orders of magnitude. This difference may have been so great that the effect of protons on the cation exchange rate was undetectable or even non-existent. For those systems that were intermediate in Mn2+ and H+ concentrations, more complex rate relationships must be invoked in order to capture the competition between these species.
Kinetic and mechanistic insights from eluate measurements
The Mn and Na concentrations in the eluate were measured with a time resolution of 20 min for each reaction using ICP-MS. When the solution was heavily dominated by
+ + H (0.001 M MnCl2 and pH 3), the behavior of Mn in the eluate paralleled that of Cs exchanged into triclinic Na-birnessite at pH 3 (Chapter 4): the cation in the influent was first heavily adsorbed and then sparingly desorbed during the initial stages of phase transformation (Fig. 5.11A). For 0.001 M MnCl2, ~ 70 wt% Mn was adsorbed to the
162 triclinic Na-birnessite in the first 20 min, followed by a brief release of Mn into solution before the concentration returned to the initial concentration of Mn in the influent. We conclude that a cation exchange reaction occurred in the first 50 minutes of this experiment because the initial adsorption of Mn corresponded exactly with the desorption of Na, which resided in the interlayer region of the starting triclinic Na-birnessite (Fig.
5.12A). Our time resolution of 20 min prevents us from drawing a similar conclusion with respect to the adsorption of Mn for concentrations of 0.01 and 0.1 M MnCl2. We suspect that because of the faster reaction rates at these higher Mn concentrations, the
[Mn] in the effluent returned to that of the influent before our first measurement when
[Mn] = 0.01 M (Fig. 5.11B); at [Mn]= 0.1 M, we infer that dissolution of the birnessite structure occurred due to the large amounts of Mn detected in the eluate (Fig. 5.11C).
However, the amount of Na released into the eluate decreased with increasing Mn in the influent, indicating that more Mn was exchanged into the interlayer with decreasing [Mn]
(Fig. 5.12).
When pH was 4.5, Mn adsorption followed the same trends as was observed at pH
3: 100 wt% of the Mn was adsorbed in the first 50 min for 0.001 M MnCl2, and Mn sorption decreased as [Mn] increased (Fig. 5.13). Again, the cation exchange reaction occurred slower at the lowest Mn concentration and we could detect the complete adsorption of Mn, whereas the cation exchange rate for the experiment with 0.1 M MnCl2 was faster than the time resolution of our eluate detection method. Na leaching again decreased as [Mn] increased, which supports the idea that less Mn is adsorbed as [Mn] increases (Fig. 5.14). Very similar trends were also apparent at pH 6 (Fig. 5.15 and
5.16).
163
The Kex values are only determined at 0.001 M MnCl2 because the cation exchange reaction occurred too quickly at higher [Mn] (Fig. 5.17). Both Mn and Na are released from the solid at pH 3, resulting in a negative slope, or negative Kex. From our
ICP-MS analysis, Mn is initially adsorbed to the birnessite, followed by a slight desorption of Mn before its concentration returns to that of the starting solution. The release of interlayer Mn2+ occurs during the triclinic-to-hexagonal phase transformation due to the lower sequestration capacity of hexagonal birnessite with respect to triclinic birnessite. As mentioned previously, this trend was also seen during Cs+ exchange into triclinic Na-birnessite where Cs+ was desorbed during the phase transformation (Chapter
4). At pH 4.5 and 6, Kex is positive which represents the concurrent Mn sorption and Na desorption from the triclinic birnessite. The cation exchange reaction becomes more ideal (1 mol Mn adsorbed:2 mol Na desorbed) as pH increases, shown by the approach of
Kex to its ideal value of 0.5. Kex also fluctuates over the course of each the reaction and remains most consistent and ideal at pH 6 due to the decrease in interference from protons in solution (Fig. 5.18).
Phase transformation reactions and kinetics
It is well-established that at pH 4 and below, triclinic Na-birnessite will spontaneously transform to hexagonal H-birnessite, and the rate of the transformation increases with decreasing pH (Lanson et al. 2002b). As noted in the introductory section, the transition at low pH is driven by the reduction of octahedral Mn3+ to Mn2+, which translocates into the interlayer, leaving behind an octahedral vacancy. Fisher et al. (in
164 prep) have observed that aqueous Cr3+ catalyzes this transition, presumably due to the oxidation of Cr3+ to Cr6+.
Our results demonstrate unambiguously that Mn2+ will likewise catalyze the transformation of triclinic to hexagonal birnessite at pH 3, and by analogy with Cr, we suggest catalysis occurs through the concomitant oxidation of Mn2+ to Mn3+,4+. At pH 3, the transformation to hexagonal H-birnessite went to completion (Figs. 5.4 and 5.19), whereas at pH 4.5 and 6, the onset of transformation was evident (Fig. 5.5) but the products were poorly crystalline. The refinements of the final products from these reactions indicate that the occupancy of the Mn octahedral sites decreased over the course of all reactions (Table 5.2). The initial occupancy for octahedral Mn refined to 1.0 for all cases. When well-crystalline hexagonal birnessite was the final reaction product (at pH 3 and sometimes at pH 4.5), the octahedral occupancies refined to as low as ~0.60. When the reaction had not gone to completion and triclinic birnessite remained detectable, the octahedral Mn occupancy in triclinic birnessite decreased to as low as ~0.75. These results support the idea that octahedral Mn3+/4+ in the starting birnessite was reduced by dissolved Mn2+, which is unstable in octahedral coordination.
As further evidence for the reduction of octahedral Mn, the Mnoct-Ooct bond distances increased from an average 1.97 Å (Chapter 4) to an average of 2.01 Å, possibly indicating that the oxidation state of octahedral Mn was transitioning from a mixture of
Mn3+/4+ to Mn3+ (Table 5.3). Elzinga (2011) examined the Mn2+ mediated reduction of hexagonal birnessite in batch reaction over much longer periods than were employed in the current study, and he reports the transformation of birnessite into completely different
Mn3+-rich phases, such as manganite (γ-MnOOH). Similarly, when we completed a
165 batch reaction containing 0.1 M MnCl2 at pH 6 for 24 hr, we observed a mix of triclinic birnessite, hausmannite (Mn3O4), groutite (α-MnOOH), manganite (γ-MnOOH), and feitknechtite (β-MnOOH) (Fig. 5.20), revealing that Mn2+ is a potent reducing agent that can transform triclinic birnessite beyond the hexagonal phase.
In Chapter 4, we quantified the rates at which triclinic Na-birnessite transforms to hexagonal H-birnessite at pH 3 in the presence of aqueous Cs. The rate constants of those reactions and from Fischer et al. (in prep) are listed with the rate constants determined from this study in Table 5.5. As [Mn] increased, the rate of the phase transformation with respect to [triclinic birnessite] increased and the reaction order transitioned from first-order ([Mn2+] = 0.001 and 0.01 M) to zero-order ([Mn2+] = 0.1 M)
(Fig. 5.21). Similarly, for the reaction at pH 4.5 in which triclinic birnessite transformed to the hexagonal phase ([Mn2+] = 0.01 M), our analysis revealed a zero-order dependence on the concentration of triclinic Na-birnessite (Fig. 5.22). Consequently, when the ratio of Mn:H exceeds a certain threshold, the rate of transformation is no longer dependent on the ionic strength. These results indicate that dissolved cations are in constant competition with protons, and subtle changes in the balance of power can exert dramatic differences in the ability of birnessite to engage in cation exchange, redox, and phase transformation reactions.
Acknowledgements
The author thanks Peter Eng, Joanne Stubbs, and Nancy Lazarz at APS. Funding for this research was provided by NSF Grant No. EAR07-45374. Use of the Advanced Photon
166 Source was supported by the U. S. Department of Energy, Office of Science, Office of
Basic Energy Sciences, under Contract No. DE-AC02-06CH11357.
References
Barnett, M.O., Jardine, P.M., and Brooks, S.C. (2002) U(VI) adsorption to heterogeneous subsurface media: Application of a surface complexation model. Environmental Science & Technology, 36(5), 937-942. Denton, A.R., and Ashcroft, N.W. (1991) Vegard's law. Physical Review A, 43(6), 3161. Drits, V.A., Lanson, B., and Gaillot, A.C. (2007) Birnessite polytype systematics and identification by powder X-ray diffraction. American Mineralogist, 92(5-6), 771- 788. Elzinga, E.J. (2011) Reductive Transformation of Birnessite by Aqueous Mn(II). Environmental Science & Technology, 45(15), 6366-6372. Fendorf, S.E., Sparks, D.L., Franz, J.A., and Camaioni, D.M. (1993) Electron- Paramagnetic Resonance Stopped-Flow Kinetic-Study of Manganese(Ii) Sorption Desorption on Birnessite. Soil Science Society of America Journal, 57(1), 57-62. Fendorf, S.E., and Zasoski, R.J. (1992) Chromium(III) Oxidation by Delta-MnO2.1. Characterization. Environmental Science & Technology, 26(1), 79-85. Finger, L.W., Cox, D.E., and Jephcoat, A.P. (1994) A Correction for Powder Diffraction Peak Asymmetry Due to Axial Divergence. Journal of Applied Crystallography, 27, 892-900. Fischer, T.B., Heaney, P.J., and Post, J.E. Time-resolved XRD of synthetic birnessite reduction by trivalent chromium, in prep. Fredrickson, J.K., Zachara, J.M., Kennedy, D.W., Kukkadapu, R.K., McKinley, J.P., Heald, S.M., Liu, C., and Plymale, A.E. (2004) Reduction of TcO4- by sediment- associated biogenic Fe(II). Geochimica et Cosmochimica Acta, 68(15), 3171- 3187. Fromherz, H. (1964) Physico-chemical calculations in science and industry. Butterworth & Co., London. Golden, D.C., Chen, C.C., and Dixon, J.B. (1986) Synthesis of Todorokite. Science, 231(4739), 717-719. House, J.E. (1997) Principles of Chemical Kinetics. 244 p. Wm. C. Brown Publishers, Dubuque, IA. Jenne, E.A. (1968) Controls on Mn, Fe, Co, Ni, Cu, and Zn Concentrations in Soils and Water - Significant Role of Hydrous Mn and Fe Oxides. Advances in Chemistry Series(73), 337-387. Lanson, B., Drits, V.A., Feng, Q., and Manceau, A. (2002a) Structure of synthetic Na- birnessite: Evidence for a triclinic one-layer unit cell. American Mineralogist, 87(11-12), 1662-1671.
167 Lanson, B., Drits, V.A., Gaillot, A.C., Silvester, E., Plancon, A., and Manceau, A. (2002b) Structure of heavy-metal sorbed birnessite: Part 1. Results from X-ray diffraction. American Mineralogist, 87(11-12), 1631-1645. Larson, A.C., and Von Dreele, R.B., 1994. GSAS-General Structure Analysis System. Los Alamos National Laboratory, Lasaga, A.C. (1981) Rate Laws of Chemical Reactions. In A.C. Lasaga, and R.J. Kirkpatrick, Eds. Kinetics of Geochemical Processes, 8, p. 1-68. Mineralogical Society of America, Chelsea, MI. Lopano, C.L., Heaney, P.J., Bandstra, J.Z., Post, J.E., and Brantley, S.L. (2011) Kinetic analysis of cation exchange in birnessite using time-resolved synchrotron X-ray diffraction. Geochimica Et Cosmochimica Acta, 75(14), 3973-3981. Lopano, C.L., Heaney, P.J., and Post, J.E. (2009) Cs-exchange in birnessite: Reaction mechanisms inferred from time-resolved X-ray diffraction and transmission electron microscopy. American Mineralogist, 94(5-6), 816-826. McKenzie, R.M. (1977) Manganese Oxides and Hydroxides. In J.B. Dixon, and S.B. Weed, Eds. Minerals in Soil Environments, p. 181-193. Soil Science Society of America, Madison, WI. Morales, J., Navas, J.J., and Tirado, J.L. (1990) Low-Temperature Hydrothermal Formation and Ion-Exchange of Hydrated Sodium Manganates. Solid State Ionics, 44(1-2), 125-130. Negra, C., Ross, D.S., and Lanzirotti, A. (2005a) Oxidizing behavior of soil manganese: Interactions among abundance, oxidation state, and pH. Soil Science Society of America Journal, 69(1), 87-95. -. (2005b) Soil manganese oxides and trace metals: Competitive sorption and microfocused synchrotron X-ray fluorescence mapping. Soil Science Society of America Journal, 69(2), 353-361. Novikov, G.V., Kulikova, L.N., Bogdanova, O.Y., and Sychkova, G.I. (2006) Layered hydrous manganese dioxide saturated with alkaline-earth cations: Synthesis and sorption properties. Russian Journal of Inorganic Chemistry, 51(2), 177-188. O'Reilly, S.E., and Hochella, M.F. (2003) Lead sorption efficiencies of natural and synthetic Mn and Fe-oxides. Geochimica Et Cosmochimica Acta, 67(23), 4471- 4487. Post, J.E. (1999) Manganese oxide minerals: Crystal structures and economic and environmental significance. Proceedings of the National Academy of Sciences of the United States of America, 96(7), 3447-3454. Post, J.E., Heaney, P.J., and Ertl, A. (2008) Rietveld refinement of the rancieite structure using synchrotron powder diffraction data. Powder Diffraction, 23(1), 10-14. Post, J.E., Heaney, P.J., and Hanson, J. (2002) Rietveld refinement of a triclinic structure for synthetic Na-birnessite using synchrotron powder diffraction data. Powder Diffraction, 17(3), 218-221. Rietveld, H.M. (1969) A Profile Refinement Method for Nuclear and Magnetic Structures. Journal of Applied Crystallography, 2, 65-71. Santelli, C.M., Webb, S.M., Dohnalkova, A.C., and Hansel, C.M. (2011) Diversity of Mn oxides produced by Mn(II)-oxidizing fungi. Geochimica Et Cosmochimica Acta, 75(10), 2762-2776.
168 Silvester, E., Manceau, A., and Drits, V.A. (1997) Structure of synthetic monoclinic Na- rich birnessite and hexagonal birnessite. 2. Results from chemical studies and EXAFS spectroscopy. American Mineralogist, 82(9-10), 962-978. Stephens, P.W. (1999) Phenomenological model of anisotropic peak broadening in powder diffraction. Journal of Applied Crystallography, 32, 281-289. Tan, H., Zhang, G.X., Heaney, P.J., Webb, S.M., and Burgos, W.D. (2010) Characterization of manganese oxide precipitates from Appalachian coal mine drainage treatment systems. Applied Geochemistry, 25(3), 389-399. Taylor, R.M., McKenzie, R.M., and Norrish, K. (1964) The mineralogy and chemistry of manganese in some Australian soils. Australian Journal of Soil Research, 2, 235- 248. Thompson, P., Cox, D.E., and Hastings, J.B. (1987) Rietveld Refinement of Debye- Scherrer Synchrotron X-Ray Data from Al2O3. Journal of Applied Crystallography, 20, 79-83. Toby, B.H. (2001) EXPGUI, a graphical user interface for GSAS. Journal of Applied Crystallography, 34, 210-213. Tu, S.H., Racz, G.J., and Goh, T.B. (1994) Transformations of Synthetic Birnessite as Affected by pH and Manganese Concentration. Clays and Clay Minerals, 42(3), 321-330. Wall, A.J., Heaney, P.J., Mathur, R., Post, J.E., Hanson, J.C., and Eng, P.J. (2011) A flow-through reaction cell that couples time-resolved X-ray diffraction with stable isotope analysis. Journal of Applied Crystallography, 44, 429-432. Weaver, R.M., and Hochella, M.F. (2003) The reactivity of seven Mn-oxides with Cr- aq(3+): A comparative analysis of a complex, aq environmentally important redox reaction. American Mineralogist, 88(11-12), 2016-2027. Webb, S.M., Tebo, B.M., and Bargar, J.R. (2005a) Structural characterization of biogenic Mn oxides produced in seawater by the marine bacillus sp strain SG-1. American Mineralogist, 90(8-9), 1342-1357. -. (2005b) Structural influences of sodium and calcium ions on the biogenic manganese oxides produced by the marine Bacillus sp., strain SG-1. Geomicrobiology Journal, 22(3-4), 181-193.
169 Figures
Figure 5.1. The crystal structure of triclinic Na-birnessite with Mn3+/4+ in all octahedral positions with hydrated Na+ in the interlayer positions, adapted from Post et al. (2002).
170
Figure 5.2. The crystal structure of hexagonal H-birnessite with Mn4+ and vacancies in the octahedral positions with proton/water complexes in the interlayer and Mn2+ on either side of the vacancies.
171
Figure 5.3. Flow-through TR-XRD experimental setup at beam line 13-BM-C at APS.
172
Figure 5.4. Stacked TR-XRD patterns of triclinic Na-birnessite exchanged with 0.001 M MnCl2 at pH , with θ (°) along the x-axis, intensity along the y-axis, and time (min) on the z-axis where each pattern represents 2 min in time.
173
Figure 5.5. Stacked TR-XRD patterns of triclinic Na-birnessite exchanged with 0.001 M MnCl2 at pH , with θ (°) along the x-axis, intensity along the y-axis, and time (min) on the z-axis where each pattern represents 5 min in time.
174
Figure 5.6. Stacked TR-XRD patterns of triclinic Na-birnessite exchanged with 0.001 M MnCl2 at pH , with θ (°) along the x-axis, intensity along the y-axis, and time (min) on the z-axis where each pattern represents 5 min in time.
175
Figure 5.7. Unit-cell volume changes of triclinic Na-birnessite exchanged at pH 3 with (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2.
176
Figure 5.8. Unit-cell volume changes of triclinic Na-birnessite exchanged at pH 4.5 with (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2.
177
Figure 5.9. Unit-cell volume changes of triclinic Na-birnessite exchanged at pH 6 with (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2.
178
Figure 5.10. The log of the initial cation exchange rate (dV/dt) plotted as a function of (A) pH and (B) -log([Mn2+]).
179
Figure 5.11. Plots of Mn adsorbed (wt%) from the flow-through solution as a function of time (min) at pH 3. Each plot represents a cation exchange reaction completed at (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2.
180
Figure 5.12. Na (ppm) released during the Mn cation exchange reaction at pH 3 plotted as a function of time (min): (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2.
181
Figure 5.13. Plots of Mn adsorbed (wt%) from the flow-through solution as a function of time (min) at pH 4.5. Each plot represents a cation exchange reaction completed at (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2.
182
Figure 5.14. Na (ppm) released during the Mn cation exchange reaction at pH 4.5 plotted as a function of time (min): (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2.
183
Figure 5.15. Plots of Mn adsorbed (wt%) from the flow-through solution as a function of time (min) at pH 6. Each plot represents a cation exchange reaction completed at (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2.
184
Figure 5.16. Na (ppm) released during the Mn cation exchange reaction at pH 6 plotted as a function of time (min): (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2.
185
Figure 5.17. Mn (x 10-6 mol) adsorbed onto triclinic Na-birnessite as Na (x 10-6 mol) is released from the birnessite interlayer during Mn cation exchange in 0.001 M MnCl2 at (A) pH 3.0, (B) pH 4.5, and (C) pH 6.0. Each data point represents the summation of Mn loaded and Na released from the birnessite as time progresses; therefore, the slope corresponds to the overall Kex value.
186
Figure 5.18. Kex plotted against time (min) at (A) pH 3.0, (B) pH 4.5, and (C) pH 6.0. Each data point represents the Kex value associated with each individual fractionated solution.
187
Figure 5.19. The change in percentage of triclinic Na-birnessite as a function of time at pH 3: (A) 0.001 M MnCl2, (B) 0.01 M MnCl2, and (C) 0.1 M MnCl2.
188
Figure 5.20. A single XRD pattern of triclinic Na-birnessite reacted with 0.1 M MnCl2 at pH 6 for 24 hr as a batch reaction (λ ).
189
Figure 5.21. The rate of phase transformation of triclinic birnessite to hexagonal birnessite at pH 3. First-order kinetics are followed at (A) 0.001 M MnCl2 and (B) 0.01 M MnCl2. Zero-order kinetics are followed at (C) 0.1 M MnCl2.
190
Figure 5.22. (A) The change in percentage of triclinic Na-birnessite as a function of time at 0.01 M MnCl2, pH 4.5. (B) The rate of the phase transformation of triclinic birnessite to hexagonal birnessite at 0.01 M MnCl2, pH 4.5.
191 Tables
Table 5.1. Final Rietveld refinement parameters for end-member triclinic Na-birnessite exchanged with Mn2+ at varying [Mn] and pH.
192 Table 5.2. Atomic coordinates and isotropic displacement factors for Mn-exchanged triclinic Na-birnessite as a function of [Mn] and pH.
193 Table 5.3. Bond distance values for Mn-exchanged triclinic Na-birnessite as a function of [Mn] and pH.
194
Table 5.4. The first-stage rate constant, k1, and initial rate, dV/dt, of the Mn-exchange reaction with triclinic Na-birnessite as a function of pH and [Mn].
pH [Mn2+] k (min-1) (Å3/min) 1 3 0.001 0.0179 0.0760
3 0.01 0.0631 0.2496
3 0.1 0.1167 0.1561
4.5 0.001 0.0123 0.0519
4.5 0.01 0.0453 0.1411
4.5 0.1 0.0457 0.1136
6 0.001 0.0098 0.0310
6 0.01 0.0224 0.0637
6 0.1 0.0789 0.1558
195 Table 5.5. Zero-order and first-order rate constants, k, with respect to [triclinic birnessite] of the triclinic Na-birnessite phase transformation to hexagonal birnessite under varying experimental conditions.
Reaction order and Rate Constant (wrt triclinic birnessite) Solution Composition Source Zero order, k First order, (% tri birn·min-1) k (min-1) 0.001 M HCl (pH 3) - 0.107 Chapter 4
0.001 M CsCl (pH 3) - 0.0087 Chapter 4
0.050 M Cr3+ (pH 3.2) -0.253 - Fischer et al. (in prep)
0.010 M Cr3+ (pH 3.2) -0.0947 - Fischer et al. (in prep)
0.001 M Cr3+ (pH 3.2) -0.0086 - Fischer et al. (in prep)
0.050 M Cr3+ (pH 4.4) -0.0481 - Fischer et al. (in prep)
0.010 M Cr3+ (pH 4.4) -0.0187 - Fischer et al. (in prep)
0.001 M Cr3+ (pH 4.4) -0.0094 - Fischer et al. (in prep)
0.010 M Cr3+ (pH 5.2) -0.0106 - Fischer et al. (in prep)
0.001 M Cr3+ (pH 5.2) -0.0064 - Fischer et al. (in prep)
0.1 M MnCl2 (pH 3) -6.54 - this study
0.01 M MnCl2 (pH 4.5) -1.45 - this study
0.01 M MnCl2 (pH 3) - 0.236 this study
0.001 M MnCl2 (pH 3) - 0.0986 this study
VITA
Claire R. Fleeger
Department of Geosciences, Pennsylvania State University [email protected], 724-504-2227
Education Ph.D., Geosciences, Pennsylvania State University, University Park, PA 5/12 . Dissertation title: Contaminant sequestration and phase transformation properties of birnessite-like phases (δ-MnO2).
M.S., Chemistry, Pennsylvania State University, University Park, PA 8/08 . Thesis title: Characterization of cesium sites within zeolites using solid state nuclear magnetic resonance.
B.S., Chemistry, Messiah College, Grantham, PA 5/06 . Research title: Synthesis of silver oxides in a glow discharge chamber.
Selected Presentations American Chemical Society National Meeting, Anaheim CA, March 28, 2011: Determining the rate of Cs sequestration by hexagonal H-birnessite and triclinic Na- birnessite as a function of pH; Claire R. Fleeger, Peter J. Heaney, and Jeffrey E. Post.
Goldschmidt 2010, Knoxville TN, June 18, 2010: Sequestration of Cs by H-birnessite from pH 3 to 10 as Measured with Time-resolved Synchrotron X-ray Diffraction; Claire R. Fleeger, Peter J. Heaney, and Jeffrey E. Post.
American Chemical Society National Meeting, Philadelphia PA, August 20, 2008: Identification of Unique Zeolitic Cesium Sites by Using 133Cs MAS NMR; Claire R. Fleeger, David E.W. Vaughan, Karl T. Mueller.
Honors and Awards . First Place – Environmental Sciences Oral Session, 14th Annual Environmental Chemistry Student Symposium, Pennsylvania State University 2011 . Second Prize Award – Oral Presentation, 43rd Annual Graduate Student Colloquium, Dept. of Geosciences, Pennsylvania State University 2011 . Second Prize Award – Poster Presentation, 41st Annual Graduate Student Colloquium, Dept of Geosciences, Pennsylvania State University 2009 . Charles E. Knopf, Sr. Memorial Teaching Scholarship, Penn State 2008 . Incoming Graduate Student Award, Pennsylvania State University 2007-2008 . Dan H. Waugh Memorial Teaching Award, Pennsylvania State University 2007
Services Chair of 42nd & 43rd Annual Graduate Student Colloquium, Penn State 2010 & 2011 Chair of the GREATT Summer Teacher Workshop, Penn State 2008 . Graduate Research and Education in Advanced Transportation Technologies (GREATT)