<%BANNER%>

Thermodynamics of Ion Exchange and Water Molecule Behavior in Zeolites

Permanent Link: http://ufdc.ufl.edu/UFE0041104/00001

Material Information

Title: Thermodynamics of Ion Exchange and Water Molecule Behavior in Zeolites
Physical Description: 1 online resource (162 p.)
Language: english
Creator: Wang, Jie
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: calorimetry, cation, hydration, mordenite, thermodynamics, thermogravimetry, zeolite
Geological Sciences -- Dissertations, Academic -- UF
Genre: Geology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Zeolites are among the most important rock-forming and environmental minerals in the surface and near-surface regions of the Earth. The open structures of zeolite frameworks give rise to high ion-exchange capacities in many zeolites that play an important role in rock-fluid interaction. Review of previous studies on the thermodynamics of ion-exchanged zeolties illuminates that water molecule behavior is an essential aspect of the ion exchange reactions in zeolites. Mordenite, one of the most common rock-forming zeolites exhibiting exchangeable ion solutions independent of Si/Al ratio, was used as the primary material for the experimental portions of this study. Thermogravimetric and calorimetric observations of mordenite solid solutions reveal that there are two energetically distinct sets of water molecules in mordenite: W1 (relatively less energetic) and W2 (relatively high energetic). The water content of W2 is independent of cation composition, whereas the occupancy of W1 decreases as the mole fraction of K+ increases. The enthalpy of hydration of W1 appears to be essentially independent of cation contents, but W2 hydration becomes less exothermic with increasing mole fraction of K+. Binary ion exchange experiments involving Na+, K+, and Ca2+ were performed using endmember Na- and K-mordenite at 298.15 K. Ion exchange involving Ca2+ in mordenite is incomplete; maximum Ca2+ mole fraction of ~ 0.44 was observed, which is consistent with natural mordente sample compositions. Mordenite exhibits preference for K+ over Na+ and Ca2+, but the selectivity between Na+ and Ca2+ varies with the equivalent fraction of Ca in the aqueous phase. A thermodynamic model was built for the temperature dependence of ion exchange in mordenite, the predictions by the model is consistent with the experimental results at elevated temperatures. In addition, thermodynamic fits to the binary isotherm data were used to develop models of mordenite compositions in equilibrium with aqueous solutions. These calculations suggest that the extraframework cation compositions of mordenites observed in marine settings are consistent with equilibration close to dilluted seawater, whereas mordenites formed in alkaline lakes and meteoric hydrothermal systems exhibit different cation contents reflecting the fluid compositions in these environments.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jie Wang.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Martin, Jonathan B.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041104:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041104/00001

Material Information

Title: Thermodynamics of Ion Exchange and Water Molecule Behavior in Zeolites
Physical Description: 1 online resource (162 p.)
Language: english
Creator: Wang, Jie
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: calorimetry, cation, hydration, mordenite, thermodynamics, thermogravimetry, zeolite
Geological Sciences -- Dissertations, Academic -- UF
Genre: Geology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Zeolites are among the most important rock-forming and environmental minerals in the surface and near-surface regions of the Earth. The open structures of zeolite frameworks give rise to high ion-exchange capacities in many zeolites that play an important role in rock-fluid interaction. Review of previous studies on the thermodynamics of ion-exchanged zeolties illuminates that water molecule behavior is an essential aspect of the ion exchange reactions in zeolites. Mordenite, one of the most common rock-forming zeolites exhibiting exchangeable ion solutions independent of Si/Al ratio, was used as the primary material for the experimental portions of this study. Thermogravimetric and calorimetric observations of mordenite solid solutions reveal that there are two energetically distinct sets of water molecules in mordenite: W1 (relatively less energetic) and W2 (relatively high energetic). The water content of W2 is independent of cation composition, whereas the occupancy of W1 decreases as the mole fraction of K+ increases. The enthalpy of hydration of W1 appears to be essentially independent of cation contents, but W2 hydration becomes less exothermic with increasing mole fraction of K+. Binary ion exchange experiments involving Na+, K+, and Ca2+ were performed using endmember Na- and K-mordenite at 298.15 K. Ion exchange involving Ca2+ in mordenite is incomplete; maximum Ca2+ mole fraction of ~ 0.44 was observed, which is consistent with natural mordente sample compositions. Mordenite exhibits preference for K+ over Na+ and Ca2+, but the selectivity between Na+ and Ca2+ varies with the equivalent fraction of Ca in the aqueous phase. A thermodynamic model was built for the temperature dependence of ion exchange in mordenite, the predictions by the model is consistent with the experimental results at elevated temperatures. In addition, thermodynamic fits to the binary isotherm data were used to develop models of mordenite compositions in equilibrium with aqueous solutions. These calculations suggest that the extraframework cation compositions of mordenites observed in marine settings are consistent with equilibration close to dilluted seawater, whereas mordenites formed in alkaline lakes and meteoric hydrothermal systems exhibit different cation contents reflecting the fluid compositions in these environments.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jie Wang.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Martin, Jonathan B.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041104:00001


This item has the following downloads:


Full Text
xml version 1.0 standalone yes
Volume_Errors
Unscanned
PreviousPageID P2



PAGE 1

THERMODYNAMICS OF ION EXCHANGE AND WATER MOLECULE BEHAVIOR IN ZEOLITES By JIE WANG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORID A IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010 1

PAGE 2

2010 Jie Wang 2

PAGE 3

To my Dad and Mom for their eternal love and support; to Yingxia for her encouragement and care about me; to Dr. Neuhoff for his advice and generous assistance; and to all who helped me on my research 3

PAGE 4

ACKNOWLEDGMENTS I would like to thank foremost, my advis or, Dr. Philip Neuhoff, for his guidance and generous assistance throughout my graduate ca reer. His patience and advice helped me overcome a lot of difficultie s and finally finish this projec t. I am extremel y grateful to the members of my committee, Dr. Jon Martin, Dr. Guerry McClellan, Dr. Mike Perfit, and Dr. Yasu Takano, for their valuable i nput and constructive comments that helped shape the dissertaion. Special thanks go to Gokce Atalan and Laura Ruhl their assistance with sample preparation and instrum ent operation; to Dr. Jon Martin, Dylan Miner, P.J. Moore, and Jason Gulley in the Hydogeochemistry Lab for their help with ion chromatography measurements; to Yu W ang and Jie Gao for ICP-AES analyses on my samples. Many thanks to Dr. Guy Hovis at the Lafayette Co llege for HF solution caloric analyses and Dr. Lindsay McHenry at Univer sity of Wisconsin-Milwaukee for XRF data. I would also acknowledge the graduate st udents and other staff members in the Department of Geological Scienc es for their help during my five years graduate study. Finally, I would like to thank my par ents and Yingxia for all their love and encouragement, which helped me to focus on my research. This work would not have been possible without their support. This project was supported in part by the U.S. National Science Foundation (grant EAR-0819769) to Dr. Philip Neuhoff and research grants by GSA and CMS to Jie Wang at the University of Florida. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDG MENTS..................................................................................................4 LIST OF TABLES............................................................................................................7 LIST OF FI GURES..........................................................................................................9 LIST OF ABBR EVIATIONS...........................................................................................11 ABSTRACT ...................................................................................................................13 CHAPTER 1 INTRODUC TION....................................................................................................15 Occurrence and Industrial Appl ications of Zeolit es.................................................16 Mineralogical Nature of Zeo lites .............................................................................18 Crystal Chemistry of Mor denite...............................................................................20 Contributions of Present Study...............................................................................21 2 THERMODYNAMIC CONSTRAINTS ON THE ION EXCHANGE OF ZEOLITEs...24 Introducti on.............................................................................................................24 Crystal Chemistry of I on-exchanged Z eolites.........................................................25 Thermodynamic Formalism....................................................................................28 Summary and Critique of the Thermoynam ic Properties of Ion Exchange in Zeolites ...................................................................................................................30 Enthalpy of I on Excha nge.................................................................................30 Entropy of I on Exchan ge..................................................................................31 Heat Capacity of Ion Exc hange........................................................................33 Volume of I on Exchan ge..................................................................................34 Temperature and Pressure Dependence of Reacti on.............................................35 Geological Imp licatio ns...........................................................................................37 3 SYNTHESIS, COMPOSITIONAL VARIAT ION, AND MIXING PROPERTIES IN NA-K MORDENITE SO LID SOLUT IONS...............................................................49 Introducti on.............................................................................................................49 Experimental Methods............................................................................................51 Sample and Char acterization...........................................................................51 Thermograv imetry............................................................................................52 Calorime try.......................................................................................................54 Result s....................................................................................................................56 Chemical Compositio n of Mord enite.................................................................56 Dehydration of Mordenite.................................................................................58 5

PAGE 6

Hydration of Mordenite.....................................................................................58 Energetics of W1 and W2.................................................................................60 Mordenite-H2O Equilib rium...............................................................................61 Enthalpy of Formati on......................................................................................62 Discussio n..............................................................................................................65 Comparison with Structur e Refinement Results...............................................65 Hydration Thermody namics of W1...................................................................66 Cation Composition and Wate r Molecule Behavio r..........................................70 Applicat ions............................................................................................................71 4 THERMODYNAMICS OF NA-K ION EXCHANGE IN MORDENITE......................92 Introducti on.............................................................................................................92 Experimental Methods............................................................................................94 Samples and Char acterization.........................................................................94 Homoionic Ex change.......................................................................................95 Binary Ion Ex change........................................................................................95 Result s....................................................................................................................97 Discussio n..............................................................................................................99 Thermodynamics of Na-K Exc hange................................................................99 Thermodynamic Analysis of Temper ature Dependence of Ion Exchange......104 Geological Imp lications .........................................................................................106 5 EXPERIMENTAL INVESTIGATION OF CA-BEARING SOLID SOLUTIONS IN MORDENITE ........................................................................................................119 Introducti on...........................................................................................................119 Materials and Me thodology ...................................................................................121 Ca Exchange of Mordenite...................................................................................123 Water Molecules in Ionexchanged Mo rdenite......................................................128 Controls on Natural Mordenite Compositions.......................................................129 6 CONCLUSION S...................................................................................................147 LIST OF REFE RENCES.............................................................................................149 BIOGRAPHICAL SKETCH ..........................................................................................162 6

PAGE 7

LIST OF TABLES Table page 2-1 Literature data for the thermodynamic properties of the ion-exchanged zeolite.41 2-2 Mair and Kelley coefficients describ ing the temperatur e dependence of heat capacity for zeolit es............................................................................................42 2-3 Thermodynamic properties of ion exchange in zeolit es......................................43 3-1 Chemical compositions of anhydrous mordenite samples..................................73 3-2 Chemical compositions of hydrous mordenite....................................................74 3-3 Integral enthalpy of hydrat ion of mordenite at 25 C...........................................75 3-4 Thermodynamic properties of hydration for energetically distinct water sites in mor denite ........................................................................................................76 3-5 Solution calorimetric data for mordenite.............................................................77 3-6 Thermodynamic properties used in thermochemical cycles and calculated enthalpies of formation for mo rdenite.................................................................78 4-1 Experimental data for binary ion exchange between Na+ and K+ in mordenite with 0.1 N chloride solu tions at 298. 15 K..........................................................108 4-2 Experimental data for binary ion exchange between Na+ and K+ in mordenite with 0.1 N chloride solu tions at 323. 15 K..........................................................109 4-3 Experimental data for binary ion exchange between Na+ and K+ in mordenite with 0.1 N chloride solu tions at 348. 15 K..........................................................110 4-4 Thermodynamic properties of Na -K ion exchange in mordenite.......................111 4-5 Margules parameters and excess properties of mor denite solid solution with 0.5 mole fraction of K+......................................................................................112 5-1 Compositions of dehydrated ion-exchanged mordeni te samples analyzed by XRF..................................................................................................................137 5-2 Experimental data for bi nary ion exchange involving Ca2+ and Na+ in mordenite with 0.1 N chlori de solutions at 25 C..............................................138 5-3 Experimental data for bi nary ion exchange involving Ca2+ and K+ in mordenite with 0.1 N chlori de solutions at 25 C..............................................139 7

PAGE 8

5-4 Thermodynamic properties of ion exchange between CaCl2 solution and Na-, K-mordenite......................................................................................................140 8

PAGE 9

LIST OF FIGURES Figure page 1-1 View of the crystal structure of a synthetic Na-mo rdenite projected along the c axis. .................................................................................................................23 2-1 The water contents of ion-exchaged zeolites: clinoptilo lite, chabazite, merlinoite, zeolite-Y, natro lite-mesolite-scolecite................................................44 2-2 The ln K of ion exchange in c linoptilolite as a func tion of tem perature................45 2-3 The ln K of ion exchange in chabazite as a function of temperature. See the caption of 2-2 for ex planation of symbols. ..........................................................46 2-4 Temperature and pr essure dependence of ln K for ion exchange in merlinoite...47 2-5 The behavior of ln K with temperature for ion ex change in clinopt ilolite..............48 3-1 Unit cell parameters and molar volume of mordenite solid solutions as a function of XK+.....................................................................................................79 3-2 The dehydration behavior of Na-MOR, NaK-MOR3, and K-MOR between 25 and 650 C at dry conditi on................................................................................80 3-3 The dehydration behavior of partia lly rehydrated Naand K-mordenite.............81 3-4 Isothermal hydration experiments on Naand K-mordenite at 25 C and ~ 40% RH. .............................................................................................................82 3-5 The heat of hydration of mordenite solid solutions at 25 C as a function of XK+......................................................................................................................83 3-6 Partial molar enthalpy of hydration as a function of water content for Naand K-mordenite........................................................................................................84 3-7 Cumulative heat evolved during absorption of water into Naand Kmordenite as a functi on of mass absorbed. ........................................................85 3-8 Water content as a function of PH2O and temperatures for Naand Kmordenite ...........................................................................................................86 3-9 Maximum water contents of mordenite at saturated PH2O...................................87 3-10 Enthalpy of formation of fully hydrated mordenite as a function of XK+...............88 3-11 Isothermal results of absorption in W1 for Naand K-mordenite in terms of ln KLangmuir as a function of XW1............................................................................89 9

PAGE 10

3-12 Prediction for the water content of mordenite as a function of PH2O. The points represent the experimental data for Na-MOR, Na-MOR3, and K-MOR...90 3-13 Fractional occupancy of W1 in mordeni te as a function of temperature at PH2O of 1 bar and 1 mbar ....................................................................................91 4-1 Isotherm for binary ion exchange of K+/Na+ in mordenite with 0.1 N chloride solutions at 298.15K. ........................................................................................113 4-2 Isotherm for binary ion exchange of K+/Na+ in mordenite with 0.1 N chloride soluctions at 323.15 K. .....................................................................................114 4-3 Isotherm for binary ion exchange of K+/Na+ in mordenite with 0.1 N chloride soluctions at 348.15 K. .....................................................................................115 4-4 Hydration states of Naand K-mordeni te as a function of temperature during ion exchan ge....................................................................................................116 4-5 The variation of equilibrium const ant for reaction 4-4 as a function of temperatur e......................................................................................................117 4-6 Excess Gibbs free energy of mordenite solid solution as a function of XK+ at different temperatures......................................................................................118 5-1 Extraframework cation contents repor ted for natural mordenites from a variety of different parageneses along with the composition of the maximum Ca-bearing mordenite generat ed in this study..................................................141 5-2 Binary ion exchange isotherm for Ca2+/Na+ in Na-mordenite with 0.1 N chloride solutions at 25 C ................................................................................142 5-3 Binary ion exchange isotherm for Ca2+/K+ in K-mordenite with 0.1 N chloride solutions at 25 C.............................................................................................143 5-4 The dehydration behavior of Na-M ordenite and CaNa-Mordenite between 25 and 650 C. ......................................................................................................144 5-5 The dehydration behavior of K-Mordenite and CaK-Mordenite between 25 and 650 C. ......................................................................................................145 5-6 Phase diagram illustrating the extraf ramework cation contents of mordenite as a function of solution composition at 298.15 K, 1 bar..................................146 10

PAGE 11

LIST OF ABBREVIATIONS a Activity CP Heat capacity CP,R Heat capacity of reaction H2O Fugascity of water G Gibbs free energy Gex Excess Gibbs free energy GR Gibbs free energy of reaction Hex Excess enthalpy Hf Enthalpy of formation HR Enthalpy of reaction HHyd Enthalpy of hydration HydH Intergral enthalpy of hydration HydH Partial molar enthalpy of hydration K Equilibrium constant of reaction KLangmuir Langmuir (ideal) equilibrium constant of reaction Kv Vanselow coefficient m Concentration of cations n Number of water molecules per framework P Pressure PH2O Water vapor pressure R Gas constant S Entropy Sex Excess entropy SR Entropy of reaction 11

PAGE 12

T Temperature V Volume VR Volume of reaction W Margules parameters X Mole fraction Activity coefficient of mordenite 12

PAGE 13

Abstract of Dissertation Pr esented to the Graduate School of the University of Florida in Partial Fulf illment of the Requirements for t he Degree of Doctor of Philosophy THERMODYNAMICS OF ION EXCHANGE AND WATER MOLECULE BEHAVIOR IN ZEOLITES By Jie Wang August 2010 Chair: Jonathan B. Martin Major: Geology Zeolites are among the most important rock-forming and environmental minerals in the surface and near-surface regions of t he Earth. The open stru ctures of zeolite frameworks give rise to high ion-exchange capacities in many zeolites that play an important role in rock-fluid interacti on. Review of previous studies on the thermodynamics of ion-exchang ed zeolties illuminates that water molecule behavior is an essential aspect of the ion exchange reac tions in zeolites. Mordenite, one of the most common rock-forming zeolites exhibi ting exchangeable ion solutions independent of Si/Al ratio, was used as the primary mate rial for the experimental portions of this study. Thermogravimetric and calorimetric ob servations of mordenite solid solutions reveal that there are two energetically distinct sets of water molecules in mordenite: W1 (relatively less energetic) and W2 (relatively high energetic). T he water content of W2 is independent of cation composit ion, whereas the occupancy of W1 decreases as the mole fraction of K+ increases. The enthalpy of hydrati on of W1 appears to be essentially independent of cation content s, but W2 hydration becomes less exothermic with increasing mole fraction of K+. Binary ion exchange experiments involving Na+, K+, and Ca2+ were performed using endmember Na and K-mordenite at 298.15 K. Ion 13

PAGE 14

14 exchange involving Ca2+ in mordenite is incomplete; maximum Ca2+ mole fraction of ~ 0.44 was observed, which is consistent wit h natural mordenite sample compositions. Mordenite exhibits preference for K+ over Na+ and Ca2+, but the selectivity between Na+ and Ca2+ varies with the equivalent fraction of Ca2+ in the aqueous phase. A thermodynamic model was built for the temperature dependence of ion exchange in mordenite. The predictions by the model is consistent with the ex perimental results at elevated temperatures. In addition, thermodynamic fits to the binary isotherm data were used to develop models of mordenite co mpositions in equilib rium with aqueous solutions. These calculations suggest that the extraframework cation compositions of mordenites observed in marine settings are c onsistent with equilibration close to dilluted seawater, whereas mordenites formed in alkaline lakes and meteoric hydrothermal systems exhibit different cation contents refl ecting the fluid compositions in these environments.

PAGE 15

CHAPTER 1 INTRODUCTION Zeolites are aluminosilicates whose structures are characterized by a tectosilicate framework of Siand Al-cent ered tetrahedra arranged to fo rm 2-10 channels that contain water molecules and extraframework cations (Gottardi and Galli 1985). The open structure of the zeolite framework and t he requirement of charge balance provide many zeolites with high ion-exchange capaci ties. The abundance of zeolites in Earths crust and the generally rapid kinetics of ion exchange reactions make this process an essential aspect of fluid-rock interaction. Be cause of the common occurrence of zeolites in low-grade metamorphic and sedimentar y environments with aqueous solutions, ion exchange of zeolites plays an im portant role in controlling t he chemical compositions of rocks and natural fluids both in geologic syst em (e.g., Sturchio et al. 1989) and in passive engineering applications like at Yucca Mountain (e.g., Broxton et al. 1987; Bish 1984; Pabalan 1994). For example, in zeolitebearing, volcanic aquif ers the effects of ion exchange have been shown to deplete groundwaters in Ca2+, Sr2+, Mg2+ relative to Na+ and K+ (e.g., White et al. 1980; Broxton et al. 1987; Vaniman et al. 2001) and geothermal solutions in Cs+, Ra+ and Ba2+ relative to Na+, K+, and Ca2+ (e.g., Keith et al. 1983; Sturchio et al. 1989). Similarly, zeolite compositions in these systems often reflect modification through ion exch ange (e.g., Ogihara1996; Vaniman et al. 2001). As a consequence, description and prediction of t he fate of geochemical processes during mineral-water interaction requires detailed knowledge of the behavi or of ion exchange processes acting in concert with other homogeneous (i.e., aqueous phase) and heterogeneous reactions in fluid-rock systems (e.g., dissolution, precipitation, surface adsorption). 15

PAGE 16

Considerable experimental data are availabl e for assessing partitioning of elements during ion exchange reactions at room te mperature, but the understanding of the temperature dependence of i on-exchange equilibria and the nature of mixing between ion-exchanged forms of zeolites is very limit ed. Previous thermodynamic studies of ion exchange in zeolites focused on derivation of isothermal, is obaric (usually 298.15 K, 1 bar) equilibrium constants by applying st andard aqueous-phase activity coefficient calculations (e.g., extended Debye-Hckel or Pitzer formulations) and assuming regular solution behavior for the mixing between zeolite components (e.g., Ames 1964a, 1964b; Fletcher et al. 1984; Shi bue 1998; Pabalan 1994; Fridri ksson et al. 2004). However, zeolites generally occur at temperatures above 298.15 K and some under pressures greater than 1 bar. In a situat ion like at Yucca Mountain, the zeolites are exposed to considerably higher temperatures. Extension of previous isothermal studies to geologic and engineered environments thus requires rigorous thermodynamic models that provides for quantitative asse ssment of the thermodynamic properties of ion exchange of zeolites as a function of temperature and pressure. As the conditions for reaction of zeolites with aqueous solutions are not cons tant, accurate prediction of temperature dependence of ion exchange in zeolites can be ke y to understanding their solid solution and selectivity sequences in diffe rent geologic environments. Occurrence and Industrial Applications of Zeolites Zeolites occur in a wide variety of env ironments, including two major types of occurrences: 1) macroscopic and microscopic crystals, often in veins, fractures, and vugs within plutonic and volcanic rocks and their metamorphosed equivalents; 2) submicroscopic crystals, commonly distribut ed in vitroclastic sediments which have undergone diagenetic or low-gr ade metamorphic processes (Passaglia and Sheppard 16

PAGE 17

2001). Zeolites can also occur during reacti ons of aqueous fluids with marine sediments (e.g., Boles and Coombs 1977), saline lake sediments (e.g., Hay and Moiola 1963), volcanic tuffs (e.g., Hay and Sheppard 2001) and soils (e.g., Baldar and Whittig 1968). Hydrothermal systems often produce zeolites as well with complex parageneses caused by overprinting of diagenetic and metamorp hic occurrences (e.g., Gottardi 1989; Neuhoff et al. 1999). Zeolites are widely used in industry because of their high ion-exchange capacities and water sorption capabilities. For instance, the favorable ion-exc hange selectivity of natural zeolites for certain cations such as Ca2+, Cs+, Sr2+, NH4 + have been applied in water softening, catalysis, and wastewater treatment (Mumpton 1977). The relatively large energy storage density of zeolites make s them useful in energy storage and heat pump technologies, where their use instead of activated alumina or silica gel can result in significant reduction of storage weight (e.g ., Shigeishi et al. 1979; Scarmozzino et al. 1980; Gopal et al. 1982; Selvidge and Miaoulis 1990). In agriculture, natural zeolites have been used as soil conditioners, carr iers for insecticides and herbicides, remediation agents in contaminated soils slow-release fertilizers, and dietary supplements in animal nutrition because of their capability of cation exchange, adsorption and their abundance in near surface, sedimentary deposits (Ming 1987; Ming and Mumpton 1989; Boettinger et al. 1995). The surface area of a zeolite-rich rock is ~10 m2/g, much bigger than that of sand (~0.01 m2/g), thus the filtration efficiency of a sand bed can be increased by mixing porous zeolit ic rock with it (e.g., Grigorieva et al. 1988; Galindo et al. 2000). Zeolites and a ssociated authigenic clay minerals can significantly reduce the porosity and permeabi lity of hydrocarbon reservoir rocks, and 17

PAGE 18

their presence is often regarded as an economic basement of explorat ion for oil and gas (e.g., McCulloh et al. 1973; McCulloh and Stew art 1980). In addition, they are frequently considered as passive barriers in radioactive waste repositories both as sorptive barriers to radionuclide mi gration and consumption of thermal energy (e.g., Carey and Bish 1996; Bish et al. 2003). Zeolites are being used in more and more new technologies, and these potential applications provide numerous possibilities to improve the environment. Mineralogical Nature of Zeolites The general chemical formula for natural zeolites is (Li, Na, K)a(Mg, Ca, Sr, Ba)d[Al(a+2d)Sin-(a+2d)O2n]m H2O where the portion in square brackets repres ents the framework and rest of the species reside within the channels, a and d are the stoichiometries of monovalent and divalent cations respectively, and m is the number of water molecu les in the unit cell (Gottardi and Gallli 1985). While modification of the framework compos ition requires dissolution and precipitation of the mineral, the extraf ramework cations are readily exchangeable (e.g., Newell and Rees 1983) and water molecules can be reversibly removed from the structure at elevated temperatures (e.g., van Reeuwijk 1974). Three distinct, yet coupled, types of solid solution are present in zeolites (e.g., Passaglia and Sheppard 2001; Armbruster and Gunter 2001; Neuhoff and Ruhl 2006). Many zeolites exhibit variable Si and Al contents, with Si/Al ratios generally ranging from 1 to 5 (e.g., Passaglia and Sheppard 20 01). The amount of extraframework cations is determined by the Al content of the framework, which effectively fixes the theoretical ion exchange capacity (e.g., Pabalan and Bertetti 2001). Although a few zeolite species exhibit coupled NaSi-CaAl substitution as obs erved in plagioclase feldspars (notably 18

PAGE 19

phillipsite and thomsonite; Galli and Ghittoni 1972; Ross et al. 1992), in most cases extraframework cations variation is independ ent of Si-Al substitution (Neuhoff and Ruhl 2006). Water contents often co-vary with both Si-Al and extraframework cation substitutions (in addition to being sensitive to temperature, pre ssure, and humidity), although this variation is generally not well c haracterized. As a general rule, however, water contents increase with increasing Si/Al ratio, increasing extraframework cation charge, and decreasing cation size (e.g., Neuhoff and Ruhl 2006). Outside of the study by Hovis et al. (2002) of Rb-Na and Cs-Na se ries in analcime (pollucite), there have been no careful studies of the water content across an ion exch ange series in a zeolite. The ease of ion exchange and sele ctivity for specific ions in zeolites is controlled by a variety of factors in cluding the framework topology (channel configuration and dimensions), ion size and shape (polarizability), c harge density of the ani onic framework, ionic charge, Si/Al ordering, and concentrati on of the external electrolyte solution (Barrer 1978). Thermal and pressure effe cts on the zeolite lattice, as well as macroscopic thermodynamic differences in the enthalpies and molar volumes of ions between the zeolitic and aqueous states may significantly a ffect these factors. For instance, high temperatures can offset the effect of large ionic radius and facilitate replacement of Na+ in zeolites by large cations like K+, Ag+, Tl+ and NH4 + (Vaughan 1978). This is borne out by the limited expe rimental observations of zeolite ionexchange equilibria at variable te mperatures (e.g., Shibue 1981; Fridriksson et al. 2004) which show a temperature effect on the cation selectivity and ion-exchange kinetics. In addition, differences in the hydration stat e as a function of ex changeable ion content may significantly influence ion exchange ener getics (e.g., Fletcher et al. 1984; Carey 19

PAGE 20

and Bish 1996; Yang et al 2001). Charge an d size differences between competing cations can lead to marked variations in water content of natural and ion-exchanged zeolites with large-cation forms being generally less hydrous and endmembers with divalent cations being more hydrous than monovalent forms (e.g ., Neuhoff and Ruhl 2006). Crystal Chemistry of Mordenite Mordenite, with ideal formula (Ca,Na2,K2)Al2Si10O24(H2O), often coexists with other zeolites in vugs of volcanic (andesite, basalt, dolerite, porphy rite, and rhyolite) and intrusive (granite) rocks (e.g., Sukheswala et al. 1974; Vezzalini et al. 1994). It also occurs as a diagenetic product of silicic tu ff, pitchstone and volcanic marine sediments (Passaglia and Sheppard 2001). Mordenite shows a high Si/Al ratio from 4.2 to 5.9, and has high thermal stability (Passaglia 1975). The framework structure of mordenite exhibits little change after dehydration, and is retained upon heating up to 1173 K (Tsitsishvili et al. 1992). Crystal chemistr y studies on numerous natural mordenite indicate that the extraframework cations in mordenite are mainly Na+, Ca2+, and K+, and the content is generally Na+>Ca2+>K+ (Passaglia and Sheppard 2001). The results of structure refinement for mo rdenite are shown in Figure 1-1. The mordenite structure can be envisioned as puckered sheets composed of six-membered rings of tetrahedra, which are parallel to (100) (e.g., Meie r 1978; Armbruster and Gunter 2001; Simoncic and Armbruster 2004) These sheets are connected along b by fourmembered ring pillars so that large channel s of twelve-membered rings and strongly compressed eight-membered rings are formed along c (e.g., Armbruster and Gunter 2001; Simoncic and Armbruster 2004). Mortie r (1982) suggested that there were three major cation sites A, D, and E in the extraf ramework of mordenite, and water molecules 20

PAGE 21

could be accommodated into sites B, C, D, E, G, H, and J. Previous results indicated that site A was occupied only by small cations like Na+ (Schlenker et al. 1979) and Ca2+ (Mortier et al. 1978), and it became empty when mordenite was exchanged with large cations like K+ (e.g., Mortier et al. 1978). Si te B may be co-occupied by K+ and H2O, and the mixed occupancy is also possible fo r site C when mordenite contains Ca2+ (Martucci et al. 2003). In addition, Martu cci et al (2003) also found t hat in mordenite containing Ca2+ (i.e., CaNa-mordenite) only site B c ould not be completely dehydrated above 903.15 K. Contributions of Present Study The present study focuses on thermodynam ic behavior of ion exchange processes in zeolites, with a specific em phasis on elucidating the nature of solid solutions that lead to this behavior. This task is accomplished th rough theoretical, calorimetric, equilibrium, and crystallographic methods that are int egrated to develop rigorous, consistent thermodynamic models of both ion exchange reactions and exchangeable ion solutions in zeolites. The zeolite used in this study is mordenite, one of the most common rockforming zeolites exhibiting exchangeable ion so lutions independent of Si/Al ratio (e.g., Neuhoff and Ruhl 2006). The thermodynamic models for the ion exchange and solid solutions of mordenite serves as natural co mponent to the earlier studies, and facilitates evaluation of chemical and t hermodynamic consequences of the whole range of solid solutions observed in zeolites. At the sa me time, this study greatly facilitate understanding of water behavior in zeolite solid solutions and and temperature dependence of ion exchange reactions, which have received relatively little rigorous attention. This study consists of six chapte rs: Chapter 1 (this chapter) provides relevant background material; Chapter 2 reviews the thermodynamic constraints of ion exchange 21

PAGE 22

in zeolites based on the liter ature results; Chapter 3 dem onstrates the relationship between the water molecule behavior and cation composition in mordenite; Chapter 4 presents thermodynamics of ion exchange and the mixing proper ties of mordenite solid solutions; Chapter 5 shows the ex perimental investigation of Ca2+-bearing solid solutions in mordenite and its controls on nat ural mordenite compositions; Chapter 6 is the concluding remarks of this study and pr oposes some problems fo r future research. 22

PAGE 23

23 Figure 1-1. View of the cryst al structure of a synthetic Na -mordenite projected along the c axis (modified from Martu cci et al. 2003; Simoncic and Armbruster 2004). The gray boxes delineate the unit ce ll boundaries. The tetrahedra contain Si or Al surrounded by four oxygens. The big and samll spheres denote the positions of Na+ and water molecules, respectively.

PAGE 24

CHAPTER 2 THERMODYNAMIC CONSTRAINTS ON THE ION EXCHANGE OF ZEOLITES Introduction Ion exchange reactions are an integral aspect of fluid-rock interaction in many geologic and engineered environ ments. Zeolites, characterized by unique sorption capabilities and high ion-exch ange capacities, are among the most common rockforming minerals readily participating in i on exchange reactions with aqueous solutions in natural systems (e.g., Helfferich 1962; Sposito 1981). The abundance of zeolites in the Earths crust and the generally rapid ki netics of ion exchange reactions (e.g. Rao and Rees 1966; Yang et al. 1997) make this process quantitatively important in many geochemical systems. For instance, because of the common occurrence of zeolites in low-grade metamorphic rocks and a variety of sedimentary ro cks associated with aqueous solutions, ion exchange of zeolites plays an important role in controlling groundwater chemistry both in natural systems (e.g., Sturchio et al 1989) and in passive engineering applicati ons like Yucca Mountain (e.g., Bish 1984; Broxton et al. 1987; Pabalan 1994). Thus, description and predi ction of the geochemical processes in the geochemical systems requires detailed knowledge of the behavior of ion exchange between zeolites and aqueous solutions. Considerable experimental data are available for assessing partitioning of elements during ion exchange reactions at room temperature, but the understanding of the temperature and pressure dependence of ion-exchange equilibria and the behavior of water molecules in zeolites during the i on exchange reaction is very limited. Previous thermodynamic studies of ion exchange in zeo lites focused on derivation of isothermal, isobaric (usually 298.15 K, 1 bar) equilibrium constants by applying standard aqueous24

PAGE 25

phase activity coefficient calculations (e.g., extended Debye-Hckel or Pitzer formulations) and assuming regular solution behavior for the mixing between zeolite components (e.g., Ames 1964a, 1964b; Fletcher et al. 1984; Shibue 1998; Pabalan 1994; Fridriksson et al. 2004). However, zeolit es generally occur at temperatures above 298.15 K and pressures greater than 1 bar. Fu rthermore, most thermodynamic studies of ion exchange in zeolites ig nored the change of water content in zeolites across the reaction, which may lead to large errors in calculating the therm odynamic properties of reaction. Consequently, extension of previous isothermal studies to geologic and engineered environments require s rigorous thermodynamic models that provides for quantitative assessment of the th ermodynamic properties of i on exchange of zeolites as a function of temperature and pressure. The objective of this study is to illum inate the significance of temperature and pressure dependence of ion exchange in zeolit es. Literature data pertaining to the compositions, volumes, and thermodynamic pr operties of ion-exchanged zeolites were compiled and synthesized to assess the effe ct of temperature and pressure on ion exchange equilibria and selectivit ies. Explicit in this analysis are the consequences of the variation of water content with ex changeable ion content in zeolties. Crystal Chemistry of Ion-exchanged Zeolites Zeolites are tectosilicates whose structures are characterized by frameworks of Siand Al-centered tetrahedra surrounding 2-10 c hannels that contain water molecules and cations (e.g., Gottardi and Galli 1985; Armbruster a nd Gunter 2001). The compositions of zeolites can be represent ed by the general chemical formula (Az+)y/zAlySixO2(x+y) n H2O 25

PAGE 26

where Az+ represents extraframework cations of charge z+ (generally alkali or alkaline earth cations), n is the number of moles of in terchannel water molecules, and x and y are stoichiometric coefficients for Si and Al respectively. The ex traframework cations are generally coordinated by a combination of framework oxygens and water molecules and serve to balance the net negative c harge brought about by the presence of tetrahedral sites (e.g., Armbruster and Gunter 2001). Substitution of Fe3+ or Be3+ for Al3+ occurs in the tetrahedral sites of some zeo lites, but is ignored here as the extent of these subititutions tends to be limited in rock-forming zeolites. The relatively weak, largely ionic, bonds between extraframework cations and framework water oxygens, along with the porous nature of zeolite framew orks, result in the ready cation exchange properties of zeolites. The extent of framework and ex traframework solid solution in zeolites is limited by charge balance and structural constrains, and the substitutions between framework and extraframework sites follow some discrete rules (Neuhoff and Ruhl 2006). Many zeolites exhibit variable Si and Al cont ents, with Si/Al in natural materials generally ranging from 1 to 5 (e.g., Passaglia and Sheppard 2001). The amount of extraframework cations is determined by the Al content of the framework, which effectiv ely fixes the theoretical ion exchange capacity (e.g., Pabalan and Bertetti 2001). Although a few zeolite species exhibit coupled NaSi-CaAl substitution as obs erved in plagioclase feldspars (notably phillipsite and thomsonite; Galli and Ghittoni 1972; Ross et al. 1992), in most cases extraframework cations variation is independ ent of Si-Al substitution (Neuhoff and Ruhl 2006). In the case of ion exchan ge in zeolites, there is no variation of Si or Al in the 26

PAGE 27

aqueous solution, and the Si/Al ratio is const ant for the zeolite en dmembers during the ion exchange. Ion exchange in zeolites should have signi ficant impact on the water content of the minerals due to the similar occupancies and limited space in the extraframework sites. Zeolites exhibiting induced ion exchange ha ve variable water stoichiometries due to cation loading and cation size. For instance, the different form s of ion exchanged clinoptilolite (Na-, K-, and Ca -) show various maximum H2O-sorption capacities (Carey and Bish 1996). Substitution of large cations such as K+ into extraframework sites often leads to reduced water content (e.g., Fridrikss on et al. 2001; Yang et al. 2001; Simoncic et al. 2004). An extreme example of this is provided by polluciteanalcime solid solutions, in which the water content is in versely related to Cs content because Cs occupies the water site and not the Na site in this solution (e.g., Armbruster and Gunter 2001; Hovis et al. 2002). Generally, water cont ents of zeolites increase with increasing cation charge and decreasing cation size. The relationship described above between extraframework cations and water molecules necessitates a modification of c hemical equations used to describe ion exchange reactions in zeolites. Most of previous thermodynamic studies of ion exchange in zeolites assumed constant water contents in zeolites during the reaction to reduce the number of species in the reaction (e.g., Ames 1964a, 1964b; Fletcher et al 1984; Shibue 1998; Pabalan 1994). However, as ntoed above, ion exchange leads to a partial de/hydration in zeolites, which is an essential part in the t hermodynamics of this reaction. Because the exchange generally occurs between the cations with large difference in effective size per mole of charge, the amount of H2O participating in the 27

PAGE 28

reaction is non-negligible. The thermodynamic properties of H2O are in the same magnitude as those of the other species in the reaction. I gnoring this species not only violates the first law of thermodynamics, but also causes a significant error in determining the thermodynamic properties of reaction. Consequently, exchange of ions (Aa+ and Bb+) between a zeolite and an aqueous solution should be expressed as 1/ 2 1/ 2 211 BAZHOABZHO()HOba abnm n ba m (2-1) where a and b are the valences of the respective cations, Z is defined as a portion of zeolite framework holding un it negative charge. Thus, A1/aZ n H2O and B1/bZ m H2O can be considered as the homoionic exchangeable cation endmem bers of the zeolite solid solution, and n m refer to the number of water mole cules of the two cation endmembers, respectively, at a given temperature. Standar d states for the homoionic zeolite forms and liquid water are unit activity of the pur e phase at all temperatures and pressures. For the aqueous ions, the standard state corresponds to unit activity of a hypothetical one molar solution of the ion referenced to in finite dilution at all temperatures and pressures. Thermodynamic Formalism Temperature ( T ) and pressure ( P ) dependence of ion exchange in zeolties was interpreted based on the thermodynamic properti es of reaction 2-1, including enthalpy of reaction (), entropy of reaction ( ), Gibbs free energy of reaction ( ), heat capacity of reaction (), and volume of reaction (). However, the thermodynamic properties of ion exchan ge in zeolites have not been systematically o R, T,PH o R, T,PS Ro R, T,PG o P,C o R, T,PV 28

PAGE 29

studied so far, the properties of reaction () thus could only be calculated i from those of individual species in the reaction using the equation o R, T,PoCdTndirectly dPo R,=o T,P jj,T,P j (2-2) where the summation is over all species j j is the stoichiometric reaction coefficient for species j (taken to be positive for produc ts and negative for reactants) and is the corresponding standard state pro perty of the substance: t he standard state enthalpy of formation ( ), absolute entropy ( ), Gibbs energy of formation ( ), heat capacity ( ), or volume ( ) of the substance at T and P. Literature data of the thermodynamic properties for some common zeolites and cations are summarized and the ionexchange properties are then inte rpreted in the following par ts. Due to the different materials and experimental methods, the qualit y of the literature data is not reviewed. The derived thermodynamic properties of ion exchange in zeolites are systematically studied. o R, T,Po fHoSo fGo PCoVThe equilibrium constant (K) for reaction 2-1 were ca lculated as a function of temperature and pressure from the synt hesized thermodynamic properties of ion exchange. With known properties at Tr and Pr (i.e., ) and can be calculated as a function of tem perature and pressure for some zeolites via the relationship rro R, T,PHrro R, T,PSo P, RCo R, T,PGrr rr rrooo oo R, R, R, r P, R P, R R() lnTTP T,PT,PT,P TTPGGSTTTCdTV (2-3) And K is related to at the same T and P by o R, T,PG 29

PAGE 30

o R,lnT,PGRT K (2-4) where R is the gas constant (8.314 J/mol K). Summary and Critique of the Thermoyn amic Properties of Ion Exchange in Zeolites Enthalpy of Ion Exchange The of ion exchange in zeolites we re calculated from the standard enthalpy of formation ( ) of the individual species (i ncluding zeolite endmembers, cations, and H2O). The available data of o R, T,PHo fH o fH for some natural and synthetic zeolties in different cation forms (Table 2-1) were used with those of the common cations to calculate the of ion exchange in several zeolites (Table 2-3). Some of dehydrated zeolites were also collected for the investigation of the ion exchange without H2O. Comparing the chemical compositions of the zeolite samples shows that the water content in zeolite of different cation forms wa s variable. This observation is coincident with the assumption that ion exchange can caus e partial de/hydration of zeolites. It can be seen that zeolites with large size of ca tions basically have less water contents than those in the form of small cations (Figur e 2-1). For instance, the water content in clinoptilolite increases with decreasing cation size (K+ > Na+ > Ca2+). o R,HT,Po fHThe disparities between the of ion exchange in hydrated and dehydrated zeolites indicates that H2O has a significant impact on the thermodynamics of reaction. The of ion exchange in zeolites is consi dered to include two components: the heat of ion exchange and the heat of partial de/hydration, which appear to be able to offset one another. This offset effect is obviously observed in ion exchange of the zeolites having relatively large extent of partial de/hydration associated with cation o R, T,PHo R, T,PH 30

PAGE 31

substitution. This offset effect makes of ion exchange in most zeolites become relatively low energetic (Table 2-3). Because K of reaction 2-1 is related to as by the vant Hoff equation: o R, T,PHo R, T,PH o R, 2lnT,PH dK dTRT (2-5) the smaller the magnitude of the less temperature dependence of K. As a consequence, the offset effect reduces t he temperature dependence of ion exchange in zeolites. In the case of dehydrated zeolit es, without the heat of partial de/hydration during the ion exchan ge, the values of show large discrepancy from those of ion exchange in zeolites (Table 2-3). Some of the are different in magnitude, some are even reversed (the exchange between t he same cation pair is exothermal in hydrated zeolites but endothermal in dehydrat ed zeolites). Consequently, the change of water content in zeolites across the ion exchange is of significance for the thermodynamics of reaction. o R, T,PHo R, T,HPo R, T,PHEntropy of Ion Exchange The discussion of of ion exchange in this study was restricted due to the limited literature data S for ion-exchanged zeolites. The at reference temperature (Tr=298.15 K) and pressure (Pr=1 bar) is represented by the expression o R, T,PSo oS rr298.15 o oo oSP CONPHASE 0 T,PC Sd TS T (2-6) where is the configurational entropy whic h results from mixing on exchangeable cations with the same energy; and is the entropy change associated with any o CONSo PHASES 31

PAGE 32

phase transition between 0 and 298.15 K. No di scontinuities have been observed in the temperature dependence of of zeolties below 298.15 K, and thus are equal to zero for these minerals. The part integrated from is thermal entropy at 298.15 K ( ). The may be calculated from crystal stru cture determiniation and can vary substantially between individual spec imens of a given zeolite. The is independent of temperature, making the contribution from S dominate the calculation of as a function of temperature. o PCo 298.15o 298.15o R, T,P o PHASESo CONo 29So PC8.15Co 298.15So CONS S15o 29o R, T,PGo 298.15So R, T,PSDirect calculation of S is difficult because of most zeolites are unknown. In order to overcome this limitation, of ion-exchanged zeolite endmembers were estimated using the algorithms s uggested by Neuhoff (2000). Comparing of ionexchanged zeolites (e.g., clinoptilolite and me rlinolite, Table 21) indicates that increases with decreasing cation size (e.g., K+ < Na+ < Ca2+). The data of natrolite, mesolite, scolecite, and merlinoite show that the difference of S among the ionexchanged zeolites is relatively sma ll compared to that of cations and H2O. The of ion exchange in zeolites is thus dependent on the of cation pairs and H2O. However, the variations of S in these two components are usually in opposite direction. For instance, the subs titution of a larger cation like K+ for a smaller one like Na+ in the merlinoite reduces whereas the partial dehydration associated with the ion exchange leads to the c onversion of water molecules in the merlinoite to liquid phase, which increases For the exchange occurring between the common o P o 29So 298.15So R, T,P8.15o 298. 8.15S S 32

PAGE 33

cation pairs in zeolites, the contribution of H2O to is generally smaller than that of cation pair. As a consequence, the substitu tion of large cations for the small cations in zeolites usually results in a negative o R, T,PS o R,Hzeoliteo R, T,PSo PCo P,HCHeat Capacity of Ion Exchange The of the phases in the reaction as a function of temperature are required in order to calculate and ultimately evaluate the tem perature integrals in Eq. 2-3. The numeric consequences of are most pronounced at elevated temperatures as this function is integrated over temperature to derive and Thus, the assumptions concer ning the magnitude of can have an important effect on the magnitudes of other therm odynamic properties regress ed from phase equilibrium observations, particularly those made at elevated temperatures. o PCo P, RCo PCT,Po R, T,PS o R, T,PGo PCThe temperature dependence of above 298.15 K is typically represented by empiracle polynomial expressions such as the Maier-Kelley (1932) function o PC= a + bT + cT -2 (2-7) where a, b, and c are empirical fitting coefficients and the unit of T is K. The difference between of hydrated and dehydrated zeolites can be considered as the heat capacity of water molecules in the zeolties ( ), which varies somewhat between zeolties. The reported can be used to assess the of the other zeolites by assuming that the is negligible for a constructive reaction between structurally related zeolites and the oxides (cf. Eq. 2-2; Helgeson et al. 1978). For hydrated zeolites, the reliability of estimated generated using this method (as well as other published o PC2O2zeoliteo P,HOCo P, R o PCo PCC 33

PAGE 34

estimation algorithms; Berman and Brown 1985; Robinson and Haas 1983) is dependent on the similarity of between the minerals participating in the algorithm. Using two zeolites with significantly different as structural analogs for each other would lead to large uncertainties in estimates. This source of error is generally minimized when this method is used fo r the structurally similar zeolites. The estimated coefficients for of the ion-exchang ed zeolites (except of merlinoite, the coefficients for which were regre ssed from the directly measured at different temperatures by Donahoe et al. 1990) are listed in Table 2-2. The of cations and H2O were calculated with the aid of the co mputer code SUPCRT92 (Johnson et al. 1992). Knowing the of all the species in reaction 2-1, of the ion exchange were then calculated via equation 2-2 (Table 2-3). 2zeoliteo P,HOC V2zeoliteo P,HOCo P, RCo PCo PCo PCo Po PC V Co PCoVVolume of Ion Exchange Zeolites tend to be well crystallized, a ffording reliable determination of the standard volumes ( ) by x-ray diffraction (XRD) measur ements of unit cell parameters. The of some ion-exchanged zeolites were calculated from the unit cell parameters (Table 2-1). The data for clinoptilolite were actually the difference of total space occupied by both cations and water molecu les between the Na-c linoptilolite and the other cation forms of clinopt ilolite (Yang et al. 2001). These values were estimated by the moles of water molecules and the cation radii in a six-fold coordination in the clinoptilolite, their errors are thus relatively greater than thos e of the other. Note that the cation exchange within the extraframework of zeolites usually cause insignificant variations in It thus appears that depends primarily on the of cations and oVoVo R, T,P o 34

PAGE 35

H2O. The results of of cations and H2O as a function of temperature and pressure computed by SUPCRT92 indicate that they ar e insensitive to temperature. An offset effect similar to that in has also been observed in For instance, the substitution of K+ for Na+ in merlinoite causes partial dehydration. The change of from K+ to Na+ is negative, whereas that from H2O is positive. Although these two parts of depend on the size difference of cation pair as well as the framework (Si/Al ratio) of zeolites (as shown by the merlinoite samples with different Si/Al ratios), their contributions to the become rather small due to t he offset effect, and have the same order of magnitude as the volume cha nge of ion-exchanged zeolites. As a result, the of ion exchange in zeolites are gener ally very small. Because the ionexchanged zeolites of different charge cati ons have greater volume change than those of same charge cations, the of the former is correspond ingly larger than that of the latter. oVo R,Vo R, T,PHT,P o R, T,PVoVoVo R, T,PVo R, T,PVTemperature and Pressure Dependence of Reaction The lnK for ion exchange in clinoptilolite, c habazite, and merlinoite were calculated as a function of temperature (F igure 2-2, 2-3, and 2-4). The re sults illustrate a significant temperature dependence of the reaction. The exchange bet ween different cation pairs in the same zeolite also behaves quite differ ently with temperature. The selectivity of zeolites on certain cations can even be reve rsed with increasing tem perature, such as Ca2+/K+ in clinoptilolite (Figure 2-2) and K+/Na+ in chabazite (Figure 2-3). Comparison of the lnK of merlinoite having different Si/Al ratio (Figure 2-4) indicates that Si/Al ratio can influence the selectivity of zeolites, as well as the behavior of lnK with temperature. It 35

PAGE 36

seems that the sample with lower Si/Al rati o (P-9) is more sensitive to the one with higher Si/Al ratio (P-8). The temperature dependence of i on exchange in clinopt ilolite was also interpreted from the solubility data of the clinoptilolite end members. The solubilities of Naand K-clinoptilolite were determined by Wilk in and Barnes (1998) in dilute solutions under 573.15 K, and that of Ca -clinoptilolite was measured by Benning et al. (2000) between 298.15 and 548.15 K. The equi librium constant of solubility was determined as a function of temperature for thes e homoionic clinoptilolite. The lnK of ion exchange in clinoptilolite (K+/Na+, Ca2+/Na+, and Ca2+/K+) could thus be calculated as a function of temperature from the equilibrium constants of solubility of the clinoptilolite endmembers (Figure 2-5). These thermodynamic results, particularly for K+/Na+ are generally in agreement with those calculat ed from the thermodynamic properties of ion exchange (Figure 2-3). However, the lnK of Ca2+/Na+ and Ca2+/K+ calculated from solubilities change more dramatically with temperature than those from thermodynamic properties. This discrepancy may result from several aspe cts. The different chemical compositions (including the Si/Al ratio and water content) between the clinoptiolite samples in these studies can cause the different behaviors of ion exchange. The Ca-c linoptilolite studied by Yang et al. (2001) contains Mg2+, which was involved in the thermodynamic calculations, whereas the minor cations were not taken into considerations in the solubility studies. Inspection of solubility study for Ca-clinoptilolite indicated that its equilibrium constant of solubi lity as a function of temperat ure was regressed from the data at limited number of diffe rent temperatures (only tw o for the temperature below 373.15 K). Large uncertainties might exist in the regression of lim ited data points. In 36

PAGE 37

addition, the of the clinoptilolite e ndmembers in this study were not experimental results, which could to some extent affe ct the thermodynamic calculations of lnK. o P, RC The pressure dependence of ion exch ange in zeolites is relevent to and the results are shown in terms of lnK at different pressures (Figure 2-2, 2-4). When is positive, the higher P leads to a lower lnK; and a negative results in a greater lnK at higher P. As discussed above, the of ion exchange in zeolites is generally small and insensitive to the tem perature and pressure. The little change in volume during the reaction leads to a negligible impact of P on the ion exchange of zeolites. However, for some zeolites like clinoptilolite the may be relatively large (although with uncertainties), so the high pressure can cause relatively big change in lnK. o R, T,PV o R, T,PV o R, T,PVo R, T,PVo R, T,PVGeological Implications Zeolites are common authigenic minerals in deep oceanic sediments, and believed to play an important role in the intera ction between the sediments and hydrothermal fluids because of their high ion-exchange ca pacities. Most zeolites in the deep sea are formed by hydrothermal alteration of volcani c glass, and are commonly zoned vertically with depth, temperature, and interacting fluids (Alt et al. 1986). For instance, James and Silver (1988) studied the cores and cutting samples from the Cajon Pass Deep Drillhole, and found that zeolites occurr ed most intensely in faulted and fractured zones. This observation was consistent with the suggestion that the San Andreas fault was marked by fractured rocks with abundant high-temperature geothermal fluids permitting extensive zeolitic alteration (Wang 1984; Wang et al. 1986). Evans and Chester (1995) 37

PAGE 38

investigated the variations of fluid-rock interactions in ro ck at different sites on the San Gabriel fault of the San Andreas system, and the results showed that the cataclasite evolved with fluid-rock interaction during slip contained clays and zeolite veins which were enriched in Fe, Mg, Mn, and Ti relative to the average co mposition of the protolith. Authigenic formation of zeolites were al so observed in the basal sediments on the Eastern Flank of the Juan de Fuca Ridge, and vertical advection of basement fluid through the sediment column was required to produce the zeolitic alteration (Buatier et al. 2001). The thermodynamic models of ion exchan ge in zeolites as a function of temperature and pressure can help understand the zonation of zeolites enriched in different cations. The temperature and pressu re varies in the hydrothermal occurrence of zeolites, and the impact of these factors on the cation selectiv ity of zeolites has to be considered in the studies of zeolite-fluid in teraction. For example, Ogihara and Iijima (1989) investigated the chemis try of zeolite zones in Japanese arc-related basins and found several trends of clinoptilolite composition with depth: in some areas NaKclinoptilolite was gradually converted to Na -clinoptilolite with increasing depth whereas in others Na-clinoptilolite was converted to Ca-clinoptilolite with depth. Similarly, petrologic studies at Yucca Mountain by Broxton et al. (1987) demonstrated that clinoptilolites were generally K-rich at sha llow levels but became more Naor Ca-rich with depth. Broxton et al. ( 1987) hypothesized that the exchangeable cation contents were influenced by the differences in the composition of groundwaters and/or the minerals cation exchange properties. However, Ogihara (2000) found that the clinoptilolite in calcite-cemented tuff at 1100 m depth in the offshore MITI-Somaoki 38

PAGE 39

borehole was not Ca-rich, although it had ce rtainly been in contact with Ca-rich pore water during calcite cementat ion. The discrepancies of these observations can be explained by the temperature and pressure dependence of cation selectivity in clinoptilolite. The pressure for the above form ations of clinoptilolit e is generally lower than 1 kbar, consequently the te mperature dominates the i on exchange of clinoptilolite with depth. It can be seen in Figure 2-2 that clinoptilolite has a cation selectivity of K+>Ca2+>Na+ at 298.15 K. However, with increas ing depth the temperature goes up, clinoptilolite becomes more selective for Na+ than Ca2+, and its preference for K+ to Na+ decreases. This interpretation is cons istent with Ogihara and Iijima (1989)s observations in Japanese arc-related basins. In the Yucca Mountain the present temperature of Ca-clinoptilo lite zone is no higher than 328.15 K (Broxton et al. 1987), and clinoptilolite should have great preference for K+ rather than Ca2+ under this condition (Figure 2-3). It is believed t hat the present diagenet ic zones at Yucca Mountain were established during an earlier period of higher geothermal gradient when the upper crustal magma chambers replaced t he Timber Mountain-Oasis Valley caldera complex to the north (Byers et al. 1976). T he much higher temperature in that period might lead to the enrichment of Ca in clinoptilo lite. In the case of the clinoptilolite in the offshore MITI-Somaoki borehole, the temper ature at the depth w here clinoptlolite appears is 318.15 K. Therefore, even if the pore water was Ca -rich during the calcite cementation, if the concentration of K+ reached an appropriate le vel, clinoptilolite could become K-rich due to the high preference for K+ at low temperatures (Figure 2-2). In addition, the large proportion of Na+ in seawater might have constrained the absorption of Ca2+ into the clinoptilolite. 39

PAGE 40

40 Another significant applicatio n of the thermodynamic models might be tracing the geochemistry of the latest por e fluid that had contact with the sediments. The high ion exchange capacities make zeolites easily absorb certain cations from hydrothermal fluid in a short period. Spooner et al. (1977) investigated the st rontium isotopic contamination during ocean floor hydrothermal metamorphi sm of the ophiolitic rocks of the Troodos Massif, Cyprus. Among the strontium is otope analyses of the hydrothermally metamorphosed basalts and dolerites, the 87Sr/86Sr ratio in the zeolite sample was closest to that in the source of contam ination (Campanian sea water). Besides, the Rb/Sr ratio in the zeolite sample was 0.013, indicating that the zeolite sample had great selectivity of Sr to Rb. These observations demonstrate that zeo lites can be good media to record the isotopic signatures of the hy drothermal fluid during water-rock interaction. The thermodynamic models of ion exchange in zeolties can be used to determine the cation selectivity of zeolites at certain te mperature and pressure, which makes it clear that which isotopes and trace elements c an be absorbed by zeolites during the hydrothermal alteration. In addition, the geoc hemistry of the fluid can be reconstructed based on the isotopic analyses of the zeolite samples.

PAGE 41

Table 2-1. Literature data for the thermody namic properties of the ion-exchanged zeolite Zeolite Chemical composition o f,HyHe (kJ/mol) o f,AnhyHf (kJ/mol) o HydHg (kJ/mol-H2O)o 298.15S (J/mol/K ) oV (cm3/mol) Natroltiea Na2Al2Si3O10H2O -5718.6.0 359.730.72 1353.77 Mesolitea Na0.67Ca0.67Al2Si3O10.67H2O -5947.1.4 363 4107.10 Scolecitea CaAl2Si3O10H2O -6049.0 367.420.72 1378.47 Na-Clinoptiloliteb Na0.182Al0.182Si0.818O2.572H2O -1130.05.00 -949.25 .60 -30.24 80.46.5 10.80h NaK-Clinoptiloliteb Na0.097K0.085Al0.182Si0.818O2.495H2O -1109.49.04 -953.99 .20 -28.30 75.91.5 9.74h K-Clinoptiloliteb K K K K K0.182Al0.182Si0.818O2.433H2O -1094.21.12 -957.06 .32 -30.91 72.47.5 9.01h CaClinoptiloliteb Ca0.084Mg0.008Al0.182Si0.818O2.646H2O -1153.78.07 -951.52.6 0 -27.25 82.16.5 11.88h Na-Chabazitec Na3.33K0.01Si8.69Al3.31O24.69H2O -15235.87.9 -11624.2927.9 -23.1 K-Chabazitec 3.23Si8.69Al3.31O2 8.9H2O -14529.82.2 -11770.4422.2 -24.2 Li-Chabazitec Li3.75Na0.15K0.02Si7.11Al4.89O24.52H2O -15929.49.8 -12021.4119.8 -26.3 K-Chabazitec 4.79Si7.11Al4.89O24.63H2O -15439.89.0 -12053.7219.0 -32.7 NaK-Merlinoited Na0.81K0.19AlSi1.94O5.88.13H2O -3591.2.9 282.4.4 1204.43 KNa-Merlinoited 0.8Na0.2AlSi1.94O5.88.81H2O -3519.9 276.6.4 1217.68 K-Merlinoited KAlSi1.94O5.88.69H2O -3481.8.0 274.3.4 1209.25 NaK-Merlinoited Na0.81K0.19AlSi1.81O5.62.18H2O -3488.3.8 274.6.4 1229.12 KNa-Merlinoited 0.91Na0.09AlSi1.81O5.62.79H2O -3387.3.8 260.5.3 1211.66 K-Merlinoited KAlSi1.81O5.62.69H2O -3360.8 259.7.3 1198.41 aJohnson et al. (1983). bYang et al. (2001). cShim et al. (1999). dDonahoe et al. (1990). e o fH of hydrous zeolites. f o fH of anhydrous zeolites. gEnthalpy of hdyration. hThe overall volume occupied by water molecules and cations. 41

PAGE 42

Table 2-2. Mair and Kelley coefficients (Eq. 2-4) describi ng the temperature dependence of heat capacity for zeolites Zeolite a b3 c-5 Tmax (K) Natroltie 279.69 452.8 -46.0513 403 Mesolite 201.21 660.3 -20.3342 473 Scolecite 151.17 801.6 -6.1505 473 Na-Clinoptilolite 83.76 80.9 -19.6075 498 K-Clinoptilolite 77. 55 67.9 -18.5218 498 Ca-Clinoptilolite 81.54 86.3 -18.7859 498 Na-Chabazite 1191.58 877.9 -143.8627 313 K-Chabazite 976.51 1434.3 -255.7210 313 Li-Chabazite 1242.41 1834.5 -348.1166 313 K-Chabazite 1230.76 1656.2 -304.9686 313 NaK-Merlinoitea -6.39 1068.1 -0.0004 318 KNa-Merlinoitea 4.35 942.4 -0.0037 343 K-Merlinoitea 5.68 920.0 -0.0040 315 NaK-Merlinoiteb -7.36 1050.4 0.0000 313 KNa-Merlinoiteb 5.60 872.0 -0.0039 316 K-Merlinoiteb 7.55 850.1 -0.0045 320 aMerlinoite with Si/Al rations of 1.94 (P-9). bMerlinoite with Si/Al rations of 1.81 (P-8). 42

PAGE 43

Table 2-3. Thermodynamic properti es of ion exchange in zeolites o P, RC (J/mol K) Ion exchange reaction (Bb+/Aa+)a o R, T,PH (kJ/mol) o R, T,PS (J/mol K) o R, T,PG (kJ/mol) o R, T,PV (cm3/mol)a b3 c-5 Natrolite (Ca2+/Na+) Mesolite 2.78 56.48 -14.06 0.76 70.43 62.6 -37.9751 Natrolite (Ca2+/Na+) Scolecite 8.88 54.13 -7. 26 0.51 65.03 81.4 -37.3125 Chabazite (K+/Na+) -15.68 -30.13 -6.69 162.63 -141.8 -101.8404 Chabazite (K+/Li+) 17.57 105.58 -13.91 9.17 -33.8 18.3434 Dehydrated Chabazite (K+/Na+) -32.09 -85.18 -6.69 170.59 -52.3 -112.5329 Dehydrated Chabazite (K+/Li+) -33.22 -64.77 -13.91 13.77 18.0 12.1587 Clinoptilolite (K+/Na+) -9.58 -3.63 -8.50 -6.25 191.02 -381.1 -62.4065 Clinoptilolite (Ca2+/Na+) 16.64 61.12 -1.58 6.14 95.41 -17.3 -43.0188 Clinoptilolite (Ca2+/ K+) 26.22 64.74 6.92 12.38 95.60 363.6 19.3873 Dehydrated Clinoptilolite (K+/Na+) -31.11 -33.35 -8.50 193.20 -315.0 -63.6713 Dehydrated Clinoptilolite (Ca2+/Na+) 18.33 71.30 -1.58 94.25 -52.5 -42.3455 Dehydrated Clinoptilolite (Ca2+/ K+) 49.44 104.65 6.92 -131.69 247.1 19.3873 Merlinoite(P-9) (K+/Na+) -8.41 -14.77 -4.14 0.06 233.88 -495.6 -68.7417 Merlinoite(P-8) (K+/Na+) -2.72 -18.85 2.95 -2.63 239.11 -559.3 -68.6406 aBb+/Aa+ means the substitution of Bb+ for Aa+ in zeolites. 43

PAGE 44

Figure 2-1. The water contents of ion-exchaged zeolites: clinopt ilolite (Yang et al. 2001), chabazite (Shim et al. 1999), merlinoi te (Donahoe et al. 1990), zeolite-Y (Yang and Navrotsky 2000), natrolite-meso lite-scolecite (Johnson et al. 1983). 44

PAGE 45

Figure 2-2. The lnK of ion exchange in clinoptilolite as a function of temperature. The cation before the slash substitutes the one after it. The solid curves represent the reactions under 1 bar, and the dashed ones represent those under 3 kbar. 45

PAGE 46

Figure 2-3. The lnK of ion exchange in chabazite as a function of temperature. See the caption of Figure 2-2 for explanation of symbols. 46

PAGE 47

Figure 2-4. Temperature and pressure dependence of lnK for ion exchange in merlinoite. See figure captions of Figure 2-2 for explanation of symbols and curves. 47

PAGE 48

48 Figure 2-5. The behavior of lnK with temperature for ion ex change in clinoptilolite. The curves were calculated from the solu bility data of clinoptilolite endmembers (Wilkin and Barnes 1998; Benning et al. 2000).

PAGE 49

CHAPTER 3 SYNTHESIS, COMPOSITIONAL VARIATION, AND MIXING PROPERTIES IN NA-K MORDENITE SOLID SOLUTIONS Introduction The energetics of water within zeolites ex erts a profound influenc e on their stability and behavior in natural and engineered systems. Due to the high water content in zeolites, the equilibria between zeolites and water are important to their stability in diagenetic and low-grade metamorphic environm ents, which may be used to determine the conditions of metamorphism (Carey and Bish 1996). For instance, the partial dehydration of laumontite during metamorphism and diagenesis leads to the formation of leonhardite. Consideration of leonhardite as a mineral species distinct from laumontite facilitates predicti on of relative stability of zeolites in geologic systems (Neuhoff and Bird 2001). In addition, the hydrat ion capacity and ener getics of zeolites are integral to the heat evol ution at radioactive waste repository like Yucca Mountain. The hydration state of zeolit es in response to the thermal energy consumption plays a large role in water heat budgets after wast e burial (e.g., Bish et al. 2003; Long and Ewing 2004; Wang and Neuhoff 2008b). Besides the environmental conditions, the properties of zeolites themselves such as chemical composition can also cause large difference in the water behavior. Previous studies on the thermodynamic proper ties of zeolties indicate that the amount and behavior of zeolitic water vary due to the extraframework cation compositions in zeolites. Carey and Bi sh (1996) performed clinoptilolite-water equilibrium experiments, and the results showed that th e maximum water sorption capacities for the Na-, K-, and Ca-exchanged c linoptilolite samples were different. Yang et al. (2001) also studied the thermodynam ics of natural and ion-exchanged clinoptilolite 49

PAGE 50

by high-temperature calorimetry. They suggested that the average ionic potential of the extraframework cations not only affects the hy dration capacity of cli noptilolite, but also has significant impact on its energetics. Anot her example of the relationship between cations and water molecules in zeolites is provided by natrolite and scolecite (e.g., Johnson et al. 1983; Passaglia and Sheppard 2001). These two natural zeolites have identical framework, but the different extraf ramework cations lead to distinct water contents and energetics. Most of previous studies about the thermodynamics of ionexchanged zeolites focused on the endmembers in terms of the cati ons. However, the fact is that most rock-fo rming zeolites are solid solutions of mixing cations, and the energetics of water are heterogeneous and vary with the cation composition. As a consequence, detailed information for t he relationship between cations and water molecules in the zeolite solid solutions is thus required for better understanding the stability of zeolites in the geologic systems. The present study investigates the relationship between the cation mixing and water content and behavior in mordenite. Mo rdenite is one of the most common rockforming zeolites exhibiting exchangeable ion so lutions independent of Si/Al ratio (e.g., Neuhoff and Ruhl 2006). Mordenite solid solutions involving Na+ and K+ were generated by binary ion exchange experiments. The th ermodynamic properties and water contents of mordenite solid solutions between homoi onic Na and K-forms were determined to interpret the effect of ca tion mixing on the energetics and site occupancy of the water molecules. In addition, enthalpies of formation ( Hf) for mordenite endmembers and solid solutions were determined to assess t he mixing properties of mordenite solid solutions. 50

PAGE 51

Experimental Methods Sample and Characterization The zeolite sample used in this study was synthetic Na-mordenite powder (CBV 10A) obtained from Zeolyst In ternational. X-ray powder diffr action (XRD) data collection of mordenite sample was performed at 25 C on a Rigaku Ultima IV X-ray diffractometer with Ni-filtered CuK radiation generated at 40 kV and 44 mA. Sample purity and phase identity were confirmed by the XRD data. Homoionic K-mordenite was produced by immersing ~ 50 g of Na-mordenite in 500 ml of 1 N KCl solution. High-density polyethylene (HDPE) bottles containing Na-mordenite powder and KCl solution was kept in an oven at 80 C. Bottles were shaken frequently to ensure mixing between the powder and solution. The KCl solution was replaced every week, and the progress of the homoionic exchange was m onitored by inspecting the Na+ concentrations in the replaced chloride solutions using a Denver Instrument sodium i on selective electrode (ISE) and a Denver Instrument electrochemistry meter 250. Large interferences caused by the high concentrations of K+ in the replaced solutions were detected in the ISE measurements. However, the concentration of Na+ was found to decrease in each replaced solution. After twelve weeks, t he accomplishment of t he homoionic exchange was suggested by nearly constant values of Na+ concentrations in three or more replaced chloride solutions and ultimately confirmed by analysis of the resultant mordenite samples (see below). Finally, the ion-exchanged mordenite powder was separated from the solution by centrifuge, then washed with deionized water and dried naturally at room temperature. Mordenite solid solutions of mixed Na+ and K+ content were generated by binary ion exchange of Na-mordenite with 0.1 N chloride solutions containing Na+ and K+ in a 51

PAGE 52

known equivalent concentration ratio. Well-mi xed suspensions of these solutions and Na-mordenite powder in 250 mL HDPE bottles we re kept in a climate-controlled lab with constant room temperature at 25 C and frequently shaken for at least one week. The consistent results of reversed experiments starting with K-mordenite indicated that the binary ion exchange experiments had reached equilibrium. The mass of Namordenite and the equivalent concentration ratios of Na+ and K+ were adjusted to yield five mordenite solid solutions of different mole fraction of K+ (XK+). The cation concentrations of the chloride solutions a fter ion exchange were analyzed on a Dionex ion chromatography (IC) system in the hydroc hemistry lab of the Un iversity of Florida and subsequently used to determine (by ma ss balance) the compositions of the resultant mordenite solid solutions. Mordenite endmembers and solid solutions were analyzed using a battery of analytical methods to assess sample purity, composition, and structure. The samples of mordenite endmembers and solid solutions we re analyzed by XRD to verify sample purity (absence of newly-formed phases) and determine unit cell parameters. Bulk compositions of dehydrated mordenite endmembers and solid solutions were determined by x-ray fluorescence (XRF) at t he Department of Geociences, University of Wisconsin-Milwaukee (e.g., McHenry 2009). Thermogravimetry The thermogravimetric analyses (TGA) for the mordenite samples were conducted on a Netzsch STA 449C Jupiter simult aneous thermal analysis system and a TA Instruments Q5000 sorption analyzer in the physi cal geochemistry lab of the University of Florida. The former instrument c an monitor the mass change and differential scanning calorimetric (DSC) signal of a sample simultaneously in a wide range of 52

PAGE 53

temperatures (up to 675 C), and the latte r can be used to measure the mass of a sample over the full range of relative humid ity (RH) from 0 to 98% between 5 and 85 C. Temperature and caloric calibrations for Netzsch STA 449C Jupiter simultaneous thermal analysis system were performed using data based on the DSC response of standard materials. A multipoint temperatur e calibration curve was developed using the melting points of H2O, Ga, In, Sn, Bi, Zn, and Al along with the solid-solid transition points of CsCl and quartz (Cammenga et al. 1993; Gmelin and Sarge 2000; Hohne et al. 1990; Sabbah et al. 1999). Because these mate rials are incompatible with the Pt-Rh crucible used in the experiments, temperat ure calibration was conducted in identical crucibles lined with a sub-millimeter thick inse rt of alumina. Caloric calibration was accomplished by the heat-flow rate met hod (Gmelin and Sarge 2000) using the DSC response of synthetic sapphire (Gmelin and Sarge 2000; Sabbah et al. 1999; Sarge et al. 1994; Stolen et al. 1996). The backgroundcorrected DSC response of a synthetic sapphire disk similar in mass to the expe rimental charges was measured at heating rates of 5, 10, 15, and 20 K/mi n over the range of temperatures encountered in this study. Caloric calibration factors calculat ed from results at each heating rate agreed within 1% and were a nearly lin ear function of temperature. A gravimetric measurement of deliquescent salts (NaBr in this study) was used to verify the humidity control at 25 C over the range of 5 to 98%RH for Q5000 sorption analyzer (Greenspan 1977). The RH for NaBr was decreased linearly from 98 to 5% at a humidity ramp rate of 0.2%RH/min. The calibration was accomplished by obtaini ng a deliquescence point (defined as the percent relative humidity wher e the derivative of mass change with respect to the RH is zero) of NaBr that was consistent wit h the theoretical value (57.6%RH). 53

PAGE 54

Water contents of mordenite were found to be sensitive to the RH between 0 and 98%RH at 25 C. Two types of experiments were thus conducted. First, complete dehydration of mordenite was conducted on the Jupiter system by heating 20-30 mg of sample to 650 C (e.g., Neuhoff and Wang 2007a, 2007b; Wang and Neuhoff 2008b) after the sample was first equilibrated at ~ 40%RH (the maximum humidity conditions obtainable in the Jupiter syst em at room temperature). Second, the mass change of the same sample was measured as a function of RH using the TA sorption analyzer at room temperature to determine the maximum sorption capacity at 25 C. Mordenite-H2O equilibrium experiments were conducted on the TA instrument sorption analyzer. The RH of the cham ber was provided by mixing dry N2 gas with water-saturated N2, and it was adjusted by the autom ated RH control. The experiments were performed based on the methods designed by Carey and Bish (1996). For each experiment, 5-10 mg of sample was placed in a metal-coated quartz crucible, and equilibrated at 40%RH and experimental tem perature. Then the RH was decreased to 0% and the sample was stabilized until no mass change was detected. The RH was then stepped up to 95%RH and then down back to 0%RH with an interval of 5%RH. The sample was allowed to reach equilibrium at each RH, and the mass was monitored during the whole experiment. Each sample wa s run at five different temperatures to investigate the effect of te mperature on the equilibrium. Calorimetry The enthalpy of hydration (HHyd) of mordenite was determined using an isothermal DSC-based immersion tec hnique (Neuhoff and Wang, 2007b) using the Netzsch STA 449C Jupiter device previously described. This technique included two parts: complete dehydration by scanning heating and then rehydrat ion under isothermal 54

PAGE 55

conditions. Caloric and the rmal calibration of the instrument is described above. For each experiment, 20-30 mg of mordenite sample was placed into a Pt-Rh crucible with unsealed, perforated lids. During the dehydrat ion part, the system was evacuated under vacuum and then kept dry under ultra-pure N2 with a constant flowing rate of 100 mL/min. The sample was fully dehydrated by scanning heating to 650 C at the rate of 15 K/min and then allowed to cool to the 25 C. After instrument signa ls stabilized at 25 C, the gas stream was changed to humidified N2 which was generated by bubbling ultra-pure N2 gas through a saturated NaCl solution (note that a dry protective gas stream was maintained at all times flowi ng into the base of the sample chamber to protect the balance; this gas mixes with the humidified gas resulting in a lower RH than that fixed by the solution). The flow ra te of humidified gas was maintained at 20 mL/min which resulted in an average water vapor pressure (PH2O) of 13-15 mbar in the sample chamber during hydration steps (RH was monitored continuously on the gas stream exiting the system using a flow-t hrough humidity meter manufactured by Sable Systems). Under this condition the sample was allowed to react until the DSC readings stabilized. The mass change during rehydrati on was exactly the same as that of dehydration, and repeated exper iments on one mordenite sample gave virtually identical results, indicating that the sorp tion capacity and behavior of mordenite was not affected by dehydratio n and rehydration. The enthalpy of solution (HSol,50) in hydrofluoric acid (HF) of endmember and intermediate mordenite compositions was determined at Lafayette College (cf. Hovis and Roux 1993; Hovis et al. 1998). Sample weights for individual calorimetric experiments in the present study ranged from 70 to 100 mg. Each sample was 55

PAGE 56

dissolved in 910.1 g (about one liter) of 20.1 weight percent hydrofluoric acid (HF) at 50 0.05 C, under isoperibolic conditions (meaning that the temperatur e of the medium surrounding the calorimeter was held constant ) utilizing an internal sample container (Waldbaum and Robie 1970). Either one or tw o dissolution experiments were performed in each liter of acid. Multiple experiments in the same solution had no detectable effect on the data, because of the high dilution of di ssolved ions in the acid. Due to the high water content and sensitivity of mordenite to relative humidity, these samples were not dried in a dessicator before the experiments. The samples in this study were finegrained powders that dissolved r apidly, but it is not known if such small grain sizes produce heat effects related to surface area (Nitkiewicz et al. 1983). Although the calorimetric experiments were conduct ed at 50 C and 100%RH, sample mass was measured at 25 C and 55-66%RH. Even t hough the difference between the mass at these two conditions was included in the calculations of HSol,50 and Hf, the variation of surface water content adds uncertain ty to the experimental results. Results Chemical Composition of Mordenite The XRF results for major elements for all mordenite samples are reported in Table 3-1. Compositions are normalized to LOI-free 100 wt.%, and then converted to the number of cations per 24 fr amework O (Table 3-1). Note that the concentrations of Mg, P, and Ca are so small that they ar e negligible. The amounts of Ti and Fe are relatively small compared to those of Na and K, and they are generally constant in all samples. Because Ti and Fe are rarely observ ed in the framework of mordenite, the two oxides of these two elements indicated in XRF results were believed to be outside impurities. Therefore, Ti and Fe were not included in the formula calculation of 56

PAGE 57

mordenite. The results for the mordenite endmembers show the minor cation (K+ in Namordenite, Na+ in K-mordenite) is negligible compar ed to the major cation, indicating that both mordenite endm embers are homoionic. The XK+ of mordenite solid solutions calculated from XRF data were compared to those from IC analyses (Table 3-1), and both results show great agreement within the uncertainties. The average values of the two groups of results were used to calculat e the chemical formulae for mordenite solid solutions (Table 3-2). X-ray diffraction data were used to determine unit-cell dimensions and volumes. The XRD pattern of mordenite was recorded between 5 and 65 (2 ) in continuous scanning mode with a scanning rate of 2/min Unit-cell parameters refinement for all samples were carried out with the program us ing JADE 9.0. A model with orthorhombic Cmc21 structure (Alberti et al. 1986) wa s chosen for the refinement. Unit-cell dimensions for all mordenite samples are r eported in Table 3-2, and they are within the range of previous results: a = 18.052-18.168, b = 20.404-20.527, c = 7.501-7.537 (Passaglia 1975). Mordenite is the only natur al zeolite with a lattice rigid enough to withstand significant deformation during deh ydration (van Reeuwijk 1974). Therefore, the unit cell determinations are not significantly impacted by the hy dration stat e of the mordenite samples. It can be seen that the uni t cell dimensions in mordenite vary nonlinearly with cation composition (Figure 3-1) The three unit cell parameters generally decrease and then incr ease with increasing XK+. The variation of parameter c is smaller than the other two. The molar volume of mordenite solid so lution also shows nonlinearity with cation composition, indicating the existence of excess volume of mixing (Vex) in mordenite solid solutions. The Vex is negative, suggesti ng that mixing of Na+ and 57

PAGE 58

K+ in extraframework tends to reduce the volume of mor denite. The maximum Vex is estimated to be ~ -10 cm3/mol for mordentie solid solution. Dehydration of Mordenite The dehydration behavior of mordenite is illustrated in Figure 3-2. The broad endotherm apparent in the DSC signal reflects the heat of dehydrati on. The behavior of the DSC signal mirrors that of the first derivat ive of TGA (dTGA) curve almost exactly. It can be seen that besides the major peak there is another small one shown by both DSC and dTGA curves at ~ 300 C. The small peak in dTGA of mordenite suggests an abrupt change in the rate of dehydration. The two peaks shown by the DSC curve suggest that there are two energetic ally distinct sets of wate r molecules in mordenite: 1) a relatively less energetic set of water mole cules (W1) resulting in the endotherm at ~ 120 C; and 2) a relatively more energetic set of water molecules (W2), the dehydration of which is related to the endotherm at ~ 300 C. The peak center around ~ 300 C becomes less pronounc ed with increasing XK+ in mordenite. In order to better demonstrate the existence of W2, dehydration experiments were performed on the partially rehydrated Naand K-mordenite, which were generated by allowing the dehydrated samples to absorb only a small am ount of water. Unlike the fully hydrated mordenite, the partially rehydrated samp les appeared to be very stable and no mass change was observed at 25 C under dry condi tion (nearly 0%RH). The DSC curves for the dehydration of partially rehydrated samples both show only one obvious peak (Figure 3-3), indicating that the early rehydrati on only occurs at W2. Hydration of Mordenite Both dehydrated Naand K-mordenite show a strong water sorption capability in a humid environment, but they behaved somew hat differently during hydration. 58

PAGE 59

Dehydrated mordenite samples could becom e easily hydrated at 25 C and ~ 40%RH, and repeated cycling of the same sample thr ough the dehydration and rehydration steps of this procedure did not affect the sorp tion capacity of mordenite. The hydration behaviors of Naand K-mordenite are presented in Figure 3-4. The dTGA curve of Kmordenite shows that its hydration rate is rela tively constant for the majority of reaction (near-zero order) after initial exposure to humid gas until the final portion of the reaction, where the rate of reaction decreases expo nentially. As for Na-mordenite, the dTGA curve shows a general constant hydration rate during the first half of major portion of reaction, but the abrupt increase indicates that the second half has a higher rate of hydration than the first half. It thus appears that the difference bet ween the energetics of W1 and W2 is greater in Na-mordenite than in K-mordenite, which is consistent with the dehydration results presented above. The heat flow during hydration of mordenite is proportional to the area under the DSC curve, which allows a calculation of the integral enthalpy of hydration ( ) by the equation: HydH Hyd gain18.015U H km (3-1) where U is the area under the DSC curve, mgain is the mass gained in the rehydration and k is the caloric calibration factor. Calculated of mordenite (Table 3-3) from the results shown in Figure 3-4 (as we ll as those not shown for intermediate compositions) is a linear function of XK+ (Figure 3-5). The result of for Namordenite is within error of prev ious literature data (Table 3-3). HydH HydH 59

PAGE 60

Energetics of W1 and W2 The different behaviors of W1 and W2 obs erved in both dehydr ation and hydration of mordentie can also be reflected by partial molar enthalpy of hydration ( HydH ) (e.g., Fialips et al. 2005; Neuhoff and Wang 2007b), which can be determined by taking finite difference derivatives of t he TGA curve and the area under t he DSC curve in Figure 3-4 (Neuhoff and Wang, 2007b). Partial molar enthalpy of hydration was thus calculated as a function of water content for mordenite (Figure 3-6), which clearly indicates the presence of energetically distinct water type s in mordenite. The data demonstrate that the water is absorbed into W2 and W1 s equentially. The first occurs between water contents of 0 and ~ 3 H2O per formula unit (pfu) of mordenite, for which HydH is generally constant, and this part corresponds to the absorption in W2. Between water contents of ~ 3 and ~ 6 H2O pfu, HydH increases gradually and then becomes relatively constant until reaction ceases. The scatter in the data at the highest water contents is a consequence of the relatively large errors as sociated with integrating and differentiating the small DSC and dTGA signals, respectively, at the end of the experiment. The high degree part of hydration is relatively less energetic and represents the absorption in W1. In addition, the results shown in Figure 3-6 (as well as those not shown for mordenite solid solutions) suggest that t he amount of water in W2 is in sensitive to cation content at 3 H2O pfu. The total heat evolved during the course of hydration was plott ed as a function of mass absorbed for both Naand K-mordenite (F igure 3-7). It can be seen that the cumulative heat of Na-mordenite appears to in clude two different linear parts (Figure 37A), whereas the difference between the slopes of the two linear parts in K-mordenite is 60

PAGE 61

relatively less obvious (Figure 3-7B). In comparison with HydHdata (Figure 3-6), the two linear parts (low-dgree and high-degree hydrat ion) represent water absorption in W2 and W1, respectively. And the average HHdy of W1 ( HHyd,W1) and W2 ( HHyd,W2) can be calculated from the slopes of linear regressions for high-degree and low-degree hydration, respectively (Table 3-4). Mordenite-H2O Equilibrium Isothermal TGA data for Naand K-mordenite were collected at 25, 35, 50, and 60 C as a function of PH2O (Figure 3-8) to investigate the role of PH2O in the mordenite-H2O equilibrium. The results show that the water contents of Naand K-mordenite never reached a level below 3 H2O pfu in these experiments. Therefore, only water molecules in W1 exhibited de/hydrati on under the conditions of t hese experiments. Figure 3-8 shows that when the saturation vapor pr essure was approached, the apparent water content of mordenite increased dramatically at 25 C, and to a lesser extent at 35 and 50 C. However, the mass of mordenite had a more remarkable increase at 60 C than at the lower temperatures, and could not stabi lize readily. The increase in water content at high vapor pressures likely reflects intercrystalline condensation and not water molecules absorbed into mordenite (e.g., Carey and Bish 1996). Hysteretic sorption behavior was observed at relatively high PH2O in all experiments, suggesting that part of W1 in mordenite is similar to the water in nm-scale pores and clay interlayers. The comparison between the results of Naand K-mordenite shows that the behaviors of W1 in these two endmembers are quite similar to one another in response to PH2O, and the only difference is that the water content of W1 in Na-mordenite is systematically greater than that in K-mordentie. 61

PAGE 62

The maximum water content (nH2O) of mordenite endmember s and solid solutions at saturated PH2O were extrapolated from th e isotherms of mordenite-H2O equilibrium at 25 C. Because the increase of water content at relatively high PH2O (e.g., > 22.13 mbar for Na-mordenite) was caused by the interc rystalline condensation effect, it was necessary to exclude this part of data for extrapolation. A power function regression was performed for the dat a at relatively low PH2O, and the function was used to calculate nH2O with saturated PH2O. The results of nH2O for the mordenite endmembers and solid solutions are listed in Table 3-2 and shown as a function of XK+ in Figure 3-9. An excellent linear relationship has been observed between nH2O and XK+ within the uncertainties. As a consequence, the nH2O of mordenite were adjusted in accordance with the linear regression in Figure 3-9 fo r the future thermodynamic analysis of mordenite solid solutions (Table 3-2). Enthalpy of Formation HF solution calorimetric data for all sample s are reported in Tabl e 3-5. In order to gain a sense for calorimetric precisi on, twice the standard deviation of the HSol,50 for all experiments on each sample were computed, and then divided by the heat of solution for each sample (average of HSol,50 calculated from the heat capacity before and after dissolution of the sample). Calculated in th is way the spread in calorimetric data for each of the various samples ranged from about 0.26 to 0.87% of the mean heat-ofsolution values, and averaged 0.51% among all sa mples. This average error constitutes a high degree of calorimetric precision that enabled the detection of energy differences associated with variations in cation compos ition among the samples. The results show that HSol,50 of mordenite generally becomes less exothermic with increasing XK+. 62

PAGE 63

Enthalpy of formation of mordenite endmembers and solid solutions were calculated by deriving thermo chemical cycles based on the HSol,50 of various minerals with established heats of formation. The following reaction scheme was used to compute the Hf of mordenite: (1.67-x)NaAlSi3O8 + xKAlSi3O8 + 5.32SiO2 + nSolH2O(liquid) = albite sani dine quartz Na(1.67x )KxAl1.67Si10.33O24nSolH2O (3-2) mordenite In this formulation x and represent the num ber of moles of K+ in mordenite sample, and nSol is the water content of sample at 25 C and the RH of laboratory (Table 3-5). Recalling that HSol,50 and Hf,50 for reactants (left side of a reaction) and products (right side of a reaction) were related to one another as: HSol,50 (reactants) HSol,50 (products) = Hf,50 (products) Hf,50 (reactants) (3-3) In Eq. 3-3, HSol,50 of mordenite (this study) and the other species (previous studies) have been determined (Table 3-6), Hf,50 of mordenite is t hus the only unknown quantity. Because the temperat ure of the calorimetric measurements (50 C) differs from the reference temperatur e for most compilations of thermodynamic data (25 C), enthalpies of formation must be adjusted to account for the difference in heat content and water content between these temperatur es. As a consequence, three steps were take to calculate the actual enthalpy of hydrated mordenite ( Hf,hydrated) of mordenite. First, enthalpy of formati on of mordenite at 25 C with water content of nSol ( Hf,25) was estimated by the following equation: 50 f,50f,25 P 25HHCd T (3-4) 63

PAGE 64

where CP is the heat capacity of mordenite, the integral represents the change in heat content of the mineral from 25 to 50 C. Heat capacity of zeolites can be assessed by assuming that the heat capacity of reaction is negligible for a constructive reaction between structurally related zeolites and t he oxides (cf. equation 2; Helgeson et al., 1978). For hydrated zeolites, t he reliability of estimated CP generated using this method (as well as other published estimation algor ithms; Berman and Brown 1985; Robinson and Haas 1983) is dependent on the similarity of CP, H2O(zeolite) between the zeolites participating in the algorithm. Using two zeolites with significantly different CP, H2O(zeolite) as structural analogs for each other w ould lead to large uncertainties in CP estimates. This source of error is generally minimized when this method is used for the structurally similar zeolites. The CP of mordenite solid solutions at 25 C (Table 3-6) were thus estimated from the CP of a natural mordenite (Johns on et al. 1992) using the above method. In addition, th e difference between the CP of mordenite at 25 to 50 C is very small, and the impact of the change in CP on the results of Hf is negligible. As a consequence, in assuming that the CP of mordenite is indepen dent of temperature, Hf,25 of mordenite can be calculated by Eq. 3-4. Second, because mordenite exhibited different water content when RH changes, enthalpy of formation of dehydrated mordenite ( Hf,dehydrated) was derived from Hf,25 and nSol based on the results of hydration in mordenite ( HHyd,W1 and HHyd,W2 in Table 3-4). Third, Hf,hydrated was calculated for fully hydrated mo rdenite from the results of Hf,dehydrated. The data of Hf,hydrated are shown as a function of XK+ for mordenite in Figure 3-10. Note that Hf,hydrated generally becomes less exothermic with increasing XK+, which is consistent with previous observations that enthalpies of formation of zeolites become 64

PAGE 65

less exothermic when the average ionic poten tial of their extr aframework cations decreases (Yang and Navrotsky 2000). Enthalpy of formation for dehydrated mordenite in this study was also assessed us ing the linear relationship between Hf and aluminum content derived by Navrotsky and Tian (2001). The results ar e in agreement with those determined in this study (Table 3-6). An ex cellent linear fit of points in Figure 3-10 indicates that the cation mixture in mordeni te is nearly ideal. Because excess enthalpy (Hex) is defined as the difference betw een the measured enthalpy of a given composition and that of an isochemical mechanical mi xture of relevant endmember compositions, a line of ideal mixi ng was computed from the observed Hf,hydrated of mordenite endmembers (Figure 3-10). Ex cess enthalpy was thus determined by substracting heat of mechanic mixing from Hf,hydrated of mordenite solid solutions along the line of ideal mixing. The maximum Hex is 0.73 kJ/mol at XK+ of 0.5, which is within the uncertainty of Hf,hydrated. Mordenite containing Na+ and K+ is thus essentially an ideal solid solution. Discussion Comparison with Structure Refinement Results The thermogravimetric and calorimetric observations on the behavior of water molecules in mordenite were compared with previous results of structure refinement for better understanding the energetically distinct se ts of water molecules in mordenite. According to the notation of Mo rtier (1982), there are three ma jor cation sites A, D, and E in the extraframework of mordenite, and water molecules can be accommodated into sites B, C, D, E, G, H, and J. Martucci et al (2003) st udied the step by step thermal dehydration process of mordenite from Pas han (Poona, India) using the time-resolved synchrotron powder diffraction. They found that sites C, D, G, and H dehydrated 65

PAGE 66

completely below 300 C, whereas water was re moved from site B, J, and E above this temperature. In comparison with the observations in this stud y, water sites C, D, G, H correspond to W1, and water sites B, J, E are similar to W2. The resu lts of the structure refinement also illustrate that the water sites belonging to W2 are generally more closely associated to the extraframework cations, le ading to the relatively higher energetic properties than associated with W1. Mortier et al. (1978) observed that site A became empty in K-mordenite, and K+ possibly co-occupied site B with water molecules. The cooccupancy of K+ and water molecules in site B shoul d influence the energetics of W2 to some extent, which is in agreement with the calorimetric results of W2 in mordenite solid solutions. Hydration Thermodynamics of W1 The isotherms on Figure 3-8 also a llow calculation of standard-state thermodynamic properties for hydration in W1 (e.g., Carey and Bish 1996; Fridriksson et al. 2003; Fialips et al. 2004). Because the water content of a fully occupied W2 is 3 H2O pfu, the water content of W1 in Na and K-mordenite must be 3.99 and 3.32 H2O pfu, respectively. The hydration of W1 in Na-mo rdenite accounting for one mole of water can thus be expressed by the following reaction: Na0.42Al0.42Si2.59O6.02.75H2O + H2Ovaport = Na0.42Al0.42Si2.59O6.02.75H2O (3-4) W1 vacant W1 occupied The standard states for the endmem bers of W1 solid solution (i.e., Na0.42Al0.42Si2.59O6.02.75H2O and Na0.42Al0.42Si2.59O6.02.75H2O) are unit activities at all temperatures and pressures. T he standard state for water vapor is unit activity of the ideal gas at 1 bar and any temperature. The equilibrium constant (K) for reaction 3-4 is expressed as 66

PAGE 67

2W1,occ W1,vacHOa K af (3-5) where aW1,occ and aW1,vac are the activities of Na-mordenite with occupied and vacant W1, respectively; and H2O refers to the fugacity of water vapor. Because of the assumption of ideal gas for water vapor under current experimental conditions, H2O can be taken to be equal to PH2O. The fractional occupancy of W1 (XW1) in Na-mordenite can be calculated by 2HO W13 3.99n X (3-6) For the ideal Langmuir sorption in W1 (e.g ., Carey and Bish 1996; Fialips et al. 2004), the relationship between the ideal equilibrium constant (KLangmuir) and water content is expressed by 2W1 Langmuir W1HO(1)X K XP (3-7) where the unit of PH2O is bar. When the mordenite-H2O system reaches equilibrium, the chemical potential of reaction 3-4 ( Hyd,W1) is zero, and the thermodynamics of hydration in W1 can be described by t he following equation (Carey and Bish 1996) Hyd,,W1 Hyd,W1o o,Mor Hyd,W1 oo W1 P, R 2 oo111 () lnV H CdTdTdPRK TTTTTT 0 (3-8) where Hyd,W1oH is the standard-state partial mola r enthalpy of hydration in W1; To is the reference-state tem perature (25 C); R is the gas constant (8.314 J/mol K); is the standard-state heat capacity of re action 3-4, which is assum ed to be constant and equal to 2.5R (e.g., Johnson et al. 1992; Carey 1993; Carey and Bish 1996); o P, RC o,M or W1V is the 67

PAGE 68

standard-state partial molar volume of H2O in W1, which is gener ally small and has negligible contribution to the energetics of hydration over t he small range in P along the liquid vapor saturation curve for water at the temperatures of thes e experiments (e.g., Bish 1984; Carey and Bish 1996). Due to the non-ideality in the mordenite-H2O system, the activity of H2O in W1 is assumed to be represented by a polynomial expansion of XW1 (Carey and Bish 1996), Eq. 3-5 thus becomes Langmuir W1exp()i iKKWX (3-9) where Wi is temperature independent polynom ial coefficient. Based on the above assumptions, the combinatio n of Eq. 3-6 and 3-7 yields: Hyd,W1 Hyd,W1o Hyd,W1 o o Langmuir W1 ooo11 ()2.5[ln()(1)]lni iT TW HRR K TTTTTT T 0 X (3-10) Previous studies of zeolite-H2O equilibrium suggested that a second-order polynomial expansion (i=2 in Eq. 3-9) was enough to describe the activity of H2O in W1 (e.g., Carey and Bish 1996). Eq. 3-8 can thus be rearranged to become 2 o Langmuir W1W1 oln2.5[ln()(1)] T TB C KRAX TTTTT D X (3-11) where A B C and D are temperature and XW1 independent parameters, which can be determined by non-linear regression of lefthand side of Eq. 3-11 as a function of XW1 for the isotherms on Figure 3-8. The fitted parameters were used to reproduce the curves for the absorption equi librium in W1 (Figure 3-11), and the predicted values are generally consistent with t he experimental results. The predictions for the water contents of mordenite endmembers as a function of PH2O was calculated from the fit parameters (Figure 3-12). It can be seen that the predictions are in excellent ag reement with the experimental data. As shown above, the 68

PAGE 69

water content of mordenite at saturated PH2O is linearly related to XK+. Assuming that the thermodynamic properties of W1 water vary in a similar linear fashion with XK+, the water contents of mordenite solid solutions was calculated as a function of PH2O from the properties of W1 in the two endmembers (Figure 3-12). It can be seen that the calculated values are in excellent consiste ncy with the experimental data, suggesting that the linear relationsh ip between water content and XK+ applies to any hydration state of mordenite at certain temperature and PH2O. The thermodynamic properties for hydratio n in W1 of mordenite were estimated from the fit parameters in Eq. 3-11 using the following relations (Carey and Bish 1996). Hyd,W1o o(RTAB) (3-12) Hyd,W1oHRB (3-13) Hyd,W1oSR A C D (3-14) 1WR (3-15) 2WR (3-16) The integral molar enthalpy of W1 ( ) at 25 C were estimated by the following equation when XW1 equals 1. Hyd,W1H Hyd,W1o2 12 Hyd,W1 W1 W1 o()()2.5( 23 WW HHXXRT ) T (3-17) Intergral enthalpy of formation of W1 in K-mordenite was determined by the same method, and the results are listed in Table 3-4. It can be seen that of Naand K-mordenite are consistent with the results calculated from calorimetric data within the uncertainties. In addition, the thermodynamic properties of hydrat ion in W1 determined Hyd,W1H 69

PAGE 70

by mordenite-H2O equilibrium also indicate that t he energetics of W1 has little change in Naand K-mordenite, which is in coincidence with the interpretation in calorimetric analyses. Cation Composition and Water Molecule Behavior The thermodynamic analysis of dehydrati on and hydration for the ion-exchanged mordentie indicates that the water molecule behavior is strongly related to the cation composition. The nH2O of K-mordenite is lower than that of Na-mordentie, which is coincident with previous observations on t he water contents of t he other ion exchanged zeolites, such as clinoptilolite (Carey and Bish 1996; Yang et al 2001), chabazite (Shim 1999), and merlinoite (Donahoe et al 1990). The variation of water content in the ion exchanged forms of zeolites probably results from the different sizes of cations. When relatively large cations such as K+ substitutes for the relatively small ones like Na+ in zeolites, it occupies more space of the extraframework, which can influence the water sorption capacity of zeolites. It is a similar case regarding the energetics of water in zeolites. The results of for Naand K-mordenite in this study are in agreement with the argument that the hydration of zeolites becomes less exothermic with increasing radius of the cation (e.g., Barre r and Cram 1971; Yang et al. 2000; Bish and Carey 2001). More importantly, our results demonstrate that both nH2O and of mordenite change linearly with increasing XK+ (Figure 3-5 and 3-9). Based on these linear relationships, the behavior of water molecules (including nH2O and ) in mordenite solid solutions can be assessed from those of the endmembers, and vice versa. HydH HydH HydH 70

PAGE 71

The water molecules in W1 and W2 al so behave differently when the cation composition of mordenite varies. The water content of W2 is independent of XK+, whereas W1 loses water with increasing XK+. In the case of energetics, HHyd,W1 appears to be constant, but HHyd,W2 becomes less exothermic with increasing XK+. The different behaviors of W1 and W2 may be attr ibuted to their struct ural features. The space of extraframework is limited in mordenite, the exchange of K+ for Na+ results in partial dehydration. Because W1 is less energetic than W2, it is relatively easier for W1 to lose the water during ion exchange reac tion, and there is no influence on the water content of W2. However, due to the clos e correlation between W2 and cations, the ion exchange may cause some rear rangement of molecules in W2 (such as co-occupancy in a site), which correspondingly changes the energetics of W2. Applications The thermodynamic analysis for the behavior of the energetically distinct sets of water molecules in mordenite can be used to assess its stabilit y in the natural environment. Mordenite occurs in a wide range of temperature, fo r instance, 150-230 C for hydrothermal alteration in the Wa irakei geothermal area (Sheppard and Hay 2001), compared with temperatures of 40-90 C for burial diagenesis in the Neogene Green Tuff region of Japan (Iijima 1988). Because of the high energetics of W2 in mordenite (dehydration occurs at ~ 300 C 1 bar), the variation of geological conditions has little impact on the occupancy of W2. The predictions for the water content of W1 in Naand K-mordenite at 1 bar and 1 m bar were calculated as a f unction of temperature based on the thermodynamic parameters of W1 (Figur e 3-13). It can be seen that at certain PH2O 71

PAGE 72

the cation composition has a significant effe ct on the water content of mordenite at relatively high temperatures. One of the important i ndustrial applications of zeolites is solar or industrial waste energy storage, the principle of which is the dehydration and rehydration of zeolites. The synthetic and natural zeolites used for thermal storage are pr imarily those having large water contents and energetically distinct sets of water mole cules like mordenite. A key that determines which zeolite is good for this application is the energy storage density of zeolite (corresponding to in this study). However, most solar or industrial waste heat collector wor ks at 50 200 C, which limits the dehydration/rehydration to cert ain amount of water in zeolites. Taking mordenite as an example, only HHyd,W1 should be considered in the evaluation of it s energy storage density. Note that the HHyd,W1 is independent of the cati on composition of mordenite, but the W1 of K-mordenite becomes fully dehydrated at much lo wer temperature than that of Na-mordenite (Figur e 3-13). In addition, the re latively dry condition also decreases the temperature for the dehydration of mordenite. At PH2O of 1 mbar (~ 3% RH at 25 C), the highest temperature requi red to make K-mordenite completely dehydrated on W1 is only ~ 165 C. HydH 72

PAGE 73

73 Table 3-1. Chemical compositions of anhydrous mordenite samples Sample Na-MOR NaK-MOR1 NaK-MOR2 NaK-MOR3 NaK-MOR4 NaK-MOR5 K-MOR Component concentrationa (wt.%) Na2O 6.00 4.29 4.14 2.62 1.36 0.55 0.02 MgO 0.06 0.05 0.05 0.04 0.04 0.04 0.02 Al2O3 10.18 10.41 10.47 10.44 10.38 10.39 10.6 SiO2 74.93 76.14 75.99 76.01 75.88 75.66 77.43 P2O5 0.02 0.02 0.03 0.03 0.03 0.02 0.03 K2O 0.04 1.82 3.05 5.27 7.14 8.37 9.21 CaO 0.04 0.05 0.05 0.03 0.03 0.03 0.03 TiO2 0.39 0.38 0.38 0.38 0.39 0.38 0.38 Fe2O3 0.16 0.19 0.2 0.18 0.18 0.18 0.15 Total 91.82 93.35 94.36 95.00 95.43 95.62 97.87 Cations per 24 framework Ob Na 1.580 1.275 1.074 0.680 0.354 0.143 0.005 Mg 0.012 0.010 0.010 0.008 0.008 0.008 0.004 Al 1.630 1.640 1.651 1.647 1.641 1.647 1.644 Si 10.179 10.177 10.165 10.175 10.176 10.176 10.191 P 0.001 0.001 0.002 0.002 0.002 0.001 0.002 K 0.007 0.310 0.520 0.900 1.222 1.436 1.547 Ca 0.006 0.007 0.007 0.004 0.004 0.004 0.004 Ti 0.178 0.171 0.171 0.171 0.176 0.172 0.168 Fe 0.016 0.019 0.020 0.018 0.018 0.018 0.015 FC 1.65 1.66 1.67 1.67 1.66 1.67 1.66 Si/Al 6.24.10 6.21.09 6.16.09 6.18.09 6.20.09 6.18.09 6.20.09 XK+c 0.00 0.20.02 0.33.02 0.57.03 0.76.03 0.89.03 1.00 XK+d 0.00 0.19.01 0.32.02 0.57.03 0.80.04 0.92.05 1.00 aDirectly measured from XRF. bNormalized to LOI-free 100 wt.%. cCalculated from XRF data. dCalculated by mass balan ce of the IC data.

PAGE 74

Table 3-2. Chemical compositions of hydrous mordenite Unit-cell dimensions () Sample XK+ a Water contentb (wt.%) Formula a b c Vm (cm3/mol) Na-MOR 0 14.270.14 Na1.67Al1.67Si10.33O24.99H2O 18.09 20.44 7.52 1674.22 NaK-MOR1 0.200.02 13.920.14 Na1.34K0.33Al1.67Si10.33O24.86H2O 18.09 20.43 7.50 1669.55 NaK-MOR2 0.330.02 13.690.14 Na1.12K0.55Al1.67Si10.33O24.77H2O 18.05 20.43 7.50 1665.23 NaK-MOR3 0.570.03 13.330.13 Na0.72K0.95Al1.67Si10.33O24.61H2O 18.10 20.46 7.50 1673.03 NaK-MOR4 0.780.03 13.040.13 Na0.37K1.30Al1.67Si10.33O24.47H2O 18.11 20.48 7.52 1679.35 NaK-MOR5 0.910.04 12.900.13 Na0.15K1.52Al1.67Si10.33O24.38H2O 18.12 20.49 7.51 1679.46 K-MOR 1 12.610.13 K1.67Al1.67Al1.67Si10.33O24.32H2O 18.10 20.50 7.52 1680.69 aAverage of the results fr om XRF and mass balance. bMaximum water content extr apolated from mordenite-H2O equilibrium. 74

PAGE 75

Table 3-3. Integral enthalpy of hydration of mordenite at 25 C Dehdyratred mordenite Sample mass (mg) Durationa (min) Water absorbedb (mg) HydH (kJ/mol-H2O) Na-MOR 18.97 172 2.46 -73.560.74 NaK-MOR1 27.56 225 3.50 -71.930.72 NaK-MOR2 20.85 170 2.68 -70.480.70 NaK-MOR3 23.91 185 3.03 -68.200.68 NaK-MOR4 22.37 190 2.78 -66.170.66 NaK-MOR5 20.10 195 2.49 -65.250.65 K-MOR 20.40 172 2.43 -64.160.65 CaNa-MORc -73.8 1.8 Na-MORd -70.3 2.8 aDuration of immersion portion of experiment used in data regression. bMass of water absorbed during the rehydration. cCa0.289Na0.361Al0.940Si5.060O12.468H2O, Johnson et al. (1992). dH0.1Na0.9AlSi5O12, Barrer and Cram (1971). 75

PAGE 76

Table 3-4. Thermodynamic properties of hydration for energetically di stinct water sites in mordenite Thermodynamic parameters of W12 HHyd,W1 a (kJ/mol) HHyd,W2 a (kJ/mol) Hyd,,W1o (kJ/mol) Hyd,W1oH (kJ/mol) Hyd,W1oS (J/molK) W1 (J/mol) W2 (J/mol) Hyd,W1H b (kJ/mol) Na-MOR -58.11.16 83.53.67 -47.10.65 86.04.01 -130.61.15 80822. 07 -53222.95 -63.37.35 NaK-MOR1 -60.65.21 79.52.59 NaK-MOR2 -57.03.14 79.09.60 NaK-MOR3 -57.92.16 74.56.49 NaK-MOR4 -58.20.16 71.98.44 NaK-MOR5 -58.48.17 70.92.42 K-MOR -60.13.20 -70. 13.40 -27.11.57 -67. 35.63 -134.98.15 37060.79 -27938.96 -58.13.09 aCalculated from calorimetric data. bCalculated from mordenite-H2O equilibrium data. 76

PAGE 77

Table 3-5. Solution calorime tric data for mordenite (50 C) SAMPLE RH (%) Water content (pfu) Sample mass (g) Ta (K) Cp Beforeb (J/K) Cp Afterc (J/K) HSol Befored (kJ/mol) HSol Aftere (kJ/mol) Na-MOR 64 6.72 0.07403 0.039564 3881.7 3880.5 -1827.76 -1827.18 Na-MORf 59 6.67 0.07323 0.039412 3876.8 3875.8 -1836.40 -1835.93 Na-MOR 57 6.65 0.07079 0.038028 3880.8 3882.3 -1834.14 -1834.84 NaK-MOR1 59 6.51 0.07336 0.039332 3883.0 3880.9 -1831.25 -1830.29 NaK-MOR1f 57 6.49 0.07477 0.040100 3877.1 3877.4 -1834.82 -1834.94 NaK-MOR1 60 6.52 0.09032 0.048152 3879.2 3879.5 -1826.02 -1826.17 NaK-MOR2 58 6.40 0.07013 0.037449 3881.9 3881.8 -1833.34 -1833.30 NaK-MOR2f 56 6.37 0.07108 0.038141 3877.3 3877.6 -1838.95 -1839.12 NaK-MOR2f 57 6.38 0.08192 0.043813 3877.3 3877.3 -1833.24 -1833.24 NaK-MOR3 60 6.26 0.07512 0.039841 3878.0 3879.6 -1826.83 -1827.59 NaK-MOR3 56 6.21 0.07217 0.038454 3880.9 3882.0 -1834.80 -1835.32 NaK-MOR3 61 6.27 0.07091 0.037434 3879.8 3879.4 -1819.58 -1819.40 NaK-MOR4f 58 6.11 0.07241 0.038225 3877.9 3878.2 -1824.48 -1824.64 NaK-MOR4 55 6.08 0.07075 0.037422 3879.5 3880.4 -1827.78 -1828.09 NaK-MOR4 57 6.10 0.07529 0.039720 3880.2 3880.3 -1824.05 -1824.11 NaK-MOR5f 57 6.04 0.07135 0.037648 3878.1 3877.9 -1828.29 -1828.22 NaK-MOR5 57 6.04 0.07215 0.037763 3880.6 3880.5 -1814.72 -1814.69 NaK-MOR5 56.5 6.04 0.07266 0.038242 3880.1 3880.1 -1824.57 -1824.57 K-MOR 66 6.01 0.07814 0.040955 3887.3 3885.1 -1823.71 -1822.70 K-MOR 59 5.93 0.07364 0.038663 3880.7 3885.1 -1820.89 -1822.90 K-MORf 60 5.94 0.10024 0.052865 3876.0 3876.6 -1827.16 -1827.45 HSol (kJ/mol) Albite -627.34.44g Sanidine -612.94.44h Quartz -137.36.35i Water -0.19.01j aTemperature increase during dissolution. bCalorimeter heat capacity before dissolution. cCalorimeter heat capacity after dissolution. dEnthalpy of solution at 323.15 K based on heat capacity before dissolution. eEnthalpy of solution at 323.15 K based on heat capacity after dissolution. fDissolution performed in acid of preceding calorimetric experiment. gHovis (1988). hHovis (2004). iHovis (1982). jHovis (unpublished results). 77

PAGE 78

78 Table 3-6. Thermodynamic properties us ed in thermochemical cycles and calculated enthalpies of formation for mordenite Sample Hf,50 a (kJ/mol) CP b (J/mol K) Hf,hydrated c (kJ/mol) Hf,dehydrated (kJ/mol) -13261.860.82 964.97.93 -13367.111.52 -11192.321.52 -13238.970.79 962.24.92 -13359.171.50 -11184.381.50 Na-MOR -13234.960.87 961.14.92 -13361.141.55 -11186.351.55 -13204.290.99 955.90.91 -13333.341.62 -11209.641.62 -13194.490.72 954.81.91 -13329.531.47 -11205.831.47 NaK-MOR1 -13211.800.72 956.45.91 -13337.871.47 -11214.171.47 -13174.070.71 947.32.89 -13308.921.47 -11213.571.47 -13159.830.72 945.68.89 -13303.661.47 -11208.301.47 NaK-MOR2 -13168.470.71 946.23.89 -13309.301.47 -11213.951.47 -13146.930.89 946.91.89 -13275.761.56 -11242.071.56 -13124.880.80 944.18.89 -13268.661.51 -11234.971.51 NaK-MOR3 -13157.490.73 947.46.89 -13283.331.47 -11249.631.47 -13112.670.72 941.14.88 -13244.361.47 -11260.471.47 -13100.770.75 939.50.88 -13241.441.48 -11257.541.48 NaK-MOR4 -13110.310.72 940.59.88 -13244.991.47 -11261.101.47 -13092.670.72 938.92.88 -13218.301.47 -11264.621.47 -13106.220.71 938.92.88 -13231.851.47 -11278.171.47 NaK-MOR5 -13096.360.71 938.92.88 -13221.991.47 -11268.311.47 -13091.651.01 938.37.88 -13208.251.63 -11274.981.63 -13070.231.59 934.01.87 -13210.762.04 -11277.492.04 K-MOR -13067.660.74 934.55.87 -13205.191.48 -11271.921.48 Na-MORd -11211.283.40 K-MORd -11300.826.60 Suporting data for thermochemical cycles Albitee -3929.8.6 Sanidinef -3960.5.6 Quartze -909.6.0 Watere -283.9.1 aWith same water content as in Table 3-5. bCalculated from CP of natural mordenite (see text). 3cFully hydrated mordenite. dCalculated from Hf (oxides) of mordenite using the formula in Navrotsky and Tian (2001). eNeuhoff et al. (2004). fHovis et al. (2004).

PAGE 79

Figure 3-1. Unit cell parameters and molar volume of mordenite solid solutions as a function of XK+. The solid line represents volume of ideal mixing in mordenite solid solutions. 79

PAGE 80

Figure 3-2. The dehydration behavior of Na-MOR (1), Na K-MOR3 (2), and K-MOR (3) between 25 and 650 C at dry condition Dashed curves are the mass change during the dehydration. U ndulating solid curves represent dTGA. Smooth solid curves show the heat flow as a function of temperature. 80

PAGE 81

Figure 3-3. The dehydration behavior of partially rehydrated Na(1) and K-mordenite (2). See figure caption to Figure 1 for explanation of curves. Only one peak in DSC has been observed for the partia lly rehydrated mordenite. The small peak at ~ 50 C in DSC actually does not represent an energetically distinct water site, which has no corresponding change shown by the dTGA curve. 81

PAGE 82

Figure 3-4. Isothermal hydration experiment s on Na(1) and K(2) mordenite at 25 C and ~ 40% RH. See figure caption to Fi gure 3-1 for explanation of curves. 82

PAGE 83

Figure 3-5. The of mordenite solid solutions at 25 C as a function of XK+. The solid curve is a linear fitting of the points (R2 = 0.9989). HydH 83

PAGE 84

Figure 3-6. Partial molar enthalpy of hydrat ion as a function of water content for Naand K-mordenite. The points at lowest and highest water contents are so scatter that they are not involved in the diagram. 84

PAGE 85

A B Figure 3-7. Cumulative heat evolved during absorption of water into A) Naand B) Kmordenite as a function of mass absorbed calculated from TGA and DSC data shown in Figure 3-3. The solid curv es are the linear regressions for the lowand high-degree hydration in mordenite, respectively. 85

PAGE 86

A B Figure 3-8. Water content as a function of PH2O and temperatures for A) Naand B) Kmordenite. The open circles are the resu lts of isothermal desorption, and the solid triangles represent thos e of isothermal absorption. 86

PAGE 87

Figure 3-9. Maximum water contents of mordenite at saturated PH2O (in terms of number of H2O per 24 framework O) as a function of XK+. The solid curve is a linear fitting of the points (R2 = 0.9870). 87

PAGE 88

Figure 3-10. Enthalpy of formation of fully hydrated mordenite as a function of XK+. The solid line represents ideal mixing of Na+ and K+ computed from the points. 88

PAGE 89

A B Figure 3-11. Isothermal result s of absorption in W1 for A) Naand B) K-mordenite in terms of ln KLangmuir as a function of XW1. The open circles are the experimental results and the solid curves depict the predicted values calculated from fitted parameters. 89

PAGE 90

Figure 3-12. Prediction for the water c ontent of mordenite as a function of PH2O. The points represent the exper imental data for Na-MOR, Na-MOR3, and K-MOR. See caption to Figure 3-8 for explanati on of the symbols. The solid curves show the prediction for Na-MOR and KMOR calculated from mordenite-water equilibrium data. The dashed curve is the prediction for NaK-MOR3, which was calculated from those for Na-MOR and K-MOR in a ssuming that the water content of mordenite is linear to XK+ at certain temperature and PH2O. 90

PAGE 91

91 Figure 3-13. Fractional occupancy of W1 in mordenite as a function of temperature at PH2O of 1 bar and 1 mbar.

PAGE 92

CHAPTER 4 THERMODYNAMICS OF NA-K ION EXCHANGE IN MORDENITE Introduction Zeolites are aluminosilicates whose stru cture is characterized by a tectosilicate framework of Siand Al-cent ered tetrahedra arranged to fo rm 2-10 channels that contain water molecules and extraframework cations (Gottardi and Galli 1985). The open structure of the zeolite framework and the requirement of charge balance provide many zeolites with high ion-exchange capaciti es that are widely used in the industry of hard water softening and wastewat er treatment (e.g., Mur phy et al. 1978; Kallo 2001). Furthermore, because of the common occurrence of zeolites as products of interaction between aqueous solutions and low-grade me tamorphic and sedimentary rocks, ion exchange by zeolites plays an important role during the fl uid-rock interaction in the natural systems (e.g., James and Silver 1988; Sturchio et al. 1989; Neuhoff et al. 2000) and in passive engineering applications like Yucca Mountain (e.g., Smyth 1982; Bish 1984; Broxton et al. 1987; Bish 2003). Evalua tion of the consequences of ion exchange in these environments requires physical-chemic al models of the partitioning of ions between zeolites and solutions. Despite considerable experimental invest igation of the ion ex change properties of zeolites, rigorous models applicable to the in situ behavior of these minerals are lacking. Most studies have focused on the cation sele ctivity and the equilibri um of ion exchange at room temperature (e.g ., Ames 1964a, 1964b; Fletcher et al. 1984; Shibue 1998; Pabalan 1994). However, zeolites occur in nature over a r ange of temperatures, particularly those formed by hydrothermal alteration which appear to form at temperatures as high as ~ 573.15 K (e.g., Reykjanes, Iceland, Tomasson and 92

PAGE 93

Kristmannsdottir 1972; Ohaki-Broadlands, New Zealand, Henley and Ellis 1983). Despite this, there have been few studies per taining to the temperature (or pressure) dependence of these reactions (e.g., Shibue 1998; Fridriksson et al. 2004). In part, this can be attributed to the previous suggesti ons that ion excha nge equilibria between zeolites and aqueous fluids are relatively temperatureand pr essure-independent owing to relatively small enthalpies and volumes of reaction, respective ly (e.g., Helfferich 1962). Nonetheless, geologic observations suggest that exchangeabl e ion compositions in zeolites often do vary with depth/temperature in geothermal, diagenetic, and metamorphic systems. For example, Ogi hara and Iijima (1989) investigated the chemistry of zeolite zones in Japanese ar c-related basins and found several trends of clinoptilolite composition with depth: in some areas Na-K-clinoptilolite was gradually converted to Na-clinoptilolite with increasi ng depth whereas in others Na-clinoptilolite was converted to Ca-clinoptilolite with depth. Similarly, petrologic studies at Yucca Mountain by Broxton et al. ( 1987) demonstrated that clinoptilolites were generally K-rich at shallow levels but became more Naor Ca-rich with depth. Br oxton et al. (1987) hypothesized that the exchangeable cation contents were influenced by the differences in the composition of groundwaters and/or the minerals cation exchange properties. However, Ogihara (2000) found that the clinoptilo lite in calcite-cemented tuff at 1100 m depth in the offshore MITI-S omaoki borehole was not Ca-rich although it had certainly been in contact with Ca-rich pore water during calcite cementation. This suggests that some other factor like temperature should have significant impact on the exchangeable ion contents. 93

PAGE 94

This study investigated the thermody namics of ion exchange in mordenite, a common zeolite with high ion-exchange capac ity and favoring higher temperature than clinoptilolite and heulandite in hydrothermal systems. Experiments involving binary ion exchange between Na+ and K+ in mordenite were conducted at three different temperatures. Resulting isotherms and equilibrium constants of reaction ( K ) were compared to assess the temperature dependenc e of Na-K ion exchange in mordenite. A thermodynamic model was subsequently dev eloped to predict the temperature dependence of Na-K ion exchange in mordenite to facilitate predict ion of the stability and composition of mordenite in geological systems. Experimental Methods Samples and Characterization The zeolite sample used in this study was synthetic Na-mordenite powder (CBV 10A) obtained from Zeolyst International. Sample purity and phase identity were confirmed by X-ray powder diffraction (XRD) data collected at 25 C on a Rigaku Ultima IV X-ray diffractometer with Ni-filtered Cu K radiation generat ed at 40 kv and 44 mA. Bulk composition of dehydrated mor denite sample was determined to be Na1.67Al1.67Si10.33O24 by x-ray fluorescence (XRF) at the Department of Geociences, University of Wisconsin-Milwaukee (e.g., McHenry 2009). The sample was found to be very sensitive to relative humidity (RH), and the water content varied at different RH. Therefore, the wate r content measured by loss on i gnition was not suitable for the sample in this study. The actual water cont ent was measured by thermogravimetry in a humid environment (see Chapter 3 for details ). The total mass loss was determined to be 14.27.14 wt%, and this va lue was adopted as the wate r content of Na-mordenite. 94

PAGE 95

Consequently, the chemical formula of Na-m ordenite sample used in this study was Na1.67Al1.67Si10.33O24.99H2O. Homoionic Exchange Homoionic K-mordenite was produced by immersing ~ 50 g of Na-mordenite in 500 mL of 1 N KCl solution. The high-densit y polyethylene (HDPE) bottle containing Namordenite powder and KCl solution was kept in an oven at 353.15 K. The bottle was frequently shaken to make the powder and so lution well mixed. The KCl solution was replaced every week, and the progress of homoionic exchange was monitored by inspecting the Na+ concentrations in replaced chloride solutions using a sodium ion selective electrode (ISE) attached to a Denv er Instrument Model 250 electrochemical meter. Large interferences with Na+ concentration caused by the high concentrations of K+ in replaced solutions were detected in the ISE measurements. However, the concentration of Na+ was found to decrease in each r eplaced solution. After twelve weeks, the accomplishment of the hom oionic exchange was c onfirmed by nearly constant values of Na+ concentrations in three more replaced chloride solutions. Finally, the ion-exchanged mordenite pow der was separated from the solution by a centrifuge, then washed with deionized water and dried nat urally at room te mperature. The bulk composition of homoionic K-mordenite was analyzed by XRF, and the result indicated that the Na+ in the original sample was fully substituted by K+. The water content of Kmordenite was 12.61.13 wt% when meas ured by the same technique as Namordenite, the chemical formul a of K-mordenite was thus K1.67Al1.67Si10.33O24.32H2O. Binary Ion Exchange The binary ion exchange between Na+ and K+ in mordenite was conducted in the chloride solutions of 0.1 N at 298.15, 323.15, and 348.15 K. The chloride solutions were 95

PAGE 96

prepared by mass from reagent-grade NaCl and KCl, respectively. Each ion exchange system involved mixing weighted amount of homoionic mor denite powder with chloride solutions containing two competing cations in a known equivalent concentration ratio, but at constant normality. The masses of Na-mordenite and equivalent concentration rations of cations were designed based on the previous ion exchange study for clinoptilolite (Pabalan 1994) to yield relati vely evenly distribut ed points along the ionexchange isotherm. The 20 mL glass vials containing well mixed solutions and mordenite powder were kept in a climate-cont rolled lab with const ant room temperature at 298.15 K, and frequently shaken for at least one week. In addition, ion exchange isotherms were experimentally reversed by using K-mordenite instead of Na-mordenite in the above procedures. The consistent resu lts of reversed experiments starting with Kmordenite indicated that the binary ion exchange experiments had reached equilibrium. After establishing equilibrium, the aqueous solution was separated from the solid and sealed off for future analysis. The bi nary ion exchange expe riments at 323.15 and 348.15 K were prepared in the same way as de scribed above, but the vials were placed in a vertical airflow oven that could maintain constant temperature wi thin K. The solid samples in the vials were allowed to sett le for two weeks so that they could be separated from the solutions at the experimental temperatures. Concentrations of Na+ and K+ in the equilibrated solutions were analyzed by a Dionex ion chromatography (IC) system in the hydrochemistry lab of the University of Florida. The samples were prepared by diluting the experimental solutions (1:100) with deionized water. The standards were made from the combined si x cation standard-I solution obtained from Dionex. Test results of the standards showed great linearity for 96

PAGE 97

both cations (R2 equals 1.0000 for K+ and 0.9975 for Na+). The instrument drift was checked by measuring the concentrations of a standard and blank solution after every five samples. Repeated experiments showed that t he results of Na+ were usually accompanied with relatively la rger drift than those of K+, which was probably caused by the residual Na+ in the instrument after l ong-term use for seawater. Results Tables 4-1 through 4-3 show the experimental data of binary ion exchange between Na+ and K+ in mordenite at 298.15, 323.15 and 348.15 K, respectively. The molar concentrations of Na+ and K+ (mNa+ and mK+) in the aqueous solutions were calculated from IC analyses of the diluted experimental solutions. Due to the large dilution factor, errors of the cation c oncentrations particularly those with low concentrations were enlarged correspondingly. Cation concentrations of the initial 0.1 N chloride solutions were also measured by IC in order to reduce the errors, and the values were used in the following calculations of this study. The ratio of Na+ and K+ in mordenite solid solution after ion exchan ge was reported as mole fraction of K+ (XK+), which can be calculated by the change of K+ concentrati on during the ion exchange reaction (e.g., Fridriksson et al. 2004), as expressed by: XK+ = (mKClVKCl mK+VTotal) / nNa+ (4-1) where mKCl and VKCl are the concentration and volume of KCl solution, respectively (Table 4-1); VTotal is the total volume of chloride solution for binary ion exchange; and nNa+ is the number of moles of Na+ in the weighted amount of Na-mordenite. Isotherms of ion exchange at different temperatures ar e presented in Figures 4-1 through 4-3 by plotting the equivalent fraction of K+ in the solid phase ( K) as a function 97

PAGE 98

of the equivalent fraction of K+ in the coexisting aqueous phase ( EK). Both EK and K can be calculated by the following equations: EK = mK+ / ( mK+ + mNa+) (4-2) K = XK+ / ( XK+ + XNa+) (4-3) where XNa+ is the mole fraction of Na+ in mordenite solid soluti on, which is equal to (1XK+). Ion exchange experiments included se veral steps: analyzing the chemical composition of mordenite samples; preparing the initial chloride solutions; measuring the mass of mordenite samples and aqueous solutions; diluting the ion-exchanged aqueous solutions; and determining the concent rations of cations by IC. Each step contributes to uncertainty in the results, and all of them were included in the error calculations. Equivalent fractions in the solid phase were calculated by mass balance from the resultant aqueous concen trations. Uncertainties in K and EK take into account both the uncertainties in the process of the ion exchange experiments and uncertainties resulting from IC analyses and subsequent calculations. The uncertainty of EK is generally lower than that of K, because the concentrations of cations in the aqueous solution were from direct measurements whereas the major s ource of error for K involved the accumulation of errors resulting from the calculation of cation concentrations in solid phase. The isotherm was constructed by interpretation from K+ due to its relatively small error in IC anal yses. In addition, results of the reversed experiments (using K-mordenite as starting sample) also show an excellent agreement with those of Na-mordenite (Figure 4-1, 42, 4-3), indicating that ion exchange in mordenite has reached equilibrium and the reaction is reversible. 98

PAGE 99

It can be seen that the point s at the extrema of isot herms generally have large uncertainties. The reason is that small amounts of mordenite samples were used in these ion exchange experiments to obtain the values of EK close to 0 and 1. Under these conditions, even a small error in cation concentration can lead to large errors in K. The isotherms at elevated temperatures s how relatively larger errors compared to that at room temperature, which may partly result from evaporatio n during the collection of experimental solutions. Discussion Thermodynamics of Na-K Exchange The ion exchange isotherm between Na+ and K+ in mordenite was constructed based on the concentration change of cations in both aqueous and solid solutions. The ion exchange reaction involving Na+ and K+ can be expressed as K+ + NaAlSi6.19O14.38.19H2O = Na+ + KAlSi6.19O14.38.78H2O + 0.41H2O(liquid) (4-4) Standard states for the homoionic mordenite and liquid water are unit activity of the pure phase at all temperatures and pressures. For the aqueous ions, the standard state corresponds to unit activity of a hypothetical one molar solution of the ion referenced to infinite dilution at all temperatures and pressures. The thermodynamic analyses of ion exchange are based on the previous approach for equilibrium of binary ion excha nge reaction (e.g., Pabalan 1994; Pabalan and Bertetti 1999, 2001; Fridriksson et al. 2004) The equilibrium constant for reaction 4-4 can be expressed as K = (mNa+ Na+)(XK+ K+)( aH2O)0.41 / [( mK+ K+)(XNa+ Na+)] (4-5) 99

PAGE 100

where is the activity coefficient of cation in the aqueous solution; refers to the rational activity coefficient for mordenite component and account for the nonideality in the zeolite pahse; and aH2O is the activity of water, which is taken to be unity in the relatively dilute experiment al solutions of this study. Activity coefficients of Na+ and K+ in the aqueous solution was calculated using the EQ3NR aqueous speciation software (Wolery 1991) from mK+ and mNa+ determined by IC analysis. All the measurable parameters in Eq. 4-5 were summariz ed through the Vanselow coefficient (Kv) defined by Kv = (mNa+ Na+) XK+ / [( mK+ K+) XNa+] (4-6) Margules equations describing the excess Gibbs energy ( Gex) across the solid solution and associated activity coeffici ents have been successfully applied to describe mixing of zeolite endmembers in binary ion exchange experiments. (e.g., Pabalan 1994; Shibue 1998; Fridriksson et al. 2004). In this approach, the ac tivity coefficients of the endmember on a per mole of char ge absis can be expressed as lnK+ = XNa+2 [WK+ + 2XK+( WNa+ WK+)] / ( RT ) and lnNa+ = XK+2 [WNa+ + 2 XNa+( WK+ WNa+)] / ( RT ) (4-7) where WK+ and WNa+ are binary Margules interaction parameters, R is the gas constant (8.314 J/mol K), and T is temperature in Kelvin. Substi tution of Eq. 4-6 and 4-7 into Eq. 4-5 yields lnKv = ln K + WNa+(2 XK+ 3 XK+ 2 + 2 XNa+3 2 XNa+2) / ( RT ) + WK+(-2 XK+ 3 + 2XK+ 2 2 XNa+3 + XNa+2) / ( RT ) (4-8) Nonlinear regression of Eq. 48 allows determination of ln K WK+, and WNa+ (Table 4-4) for the ion exchange at individual temperatures. The isotherm and result of ln K indicate 100

PAGE 101

that mordenite has great preference for K+ over Na+. This interpretation is consistent with the results of previous studies of ion exchange in mordenite (e.g., Ames 1964a, 1964b; Wolf et al. 1978). The solid curve in Figure 4-1 repr esents the isotherm computed from the above ther modynamic parameters for i on exchange in mordenite at 298.15 K. It can be seen that the predicted isotherm is in excellent agreement with the experimental results. Similar analyses were performed for the ion exchange in mordenite at 323.15 and 348.15 K. Excellent agreement s within the uncertaint ies are shown between the predicted isotherms and the ex perimental values at elevat ed temperatures (Figure 4-2 and 4-3). The uncertainties of the experim ental results and calc ulated thermodynamic parameters generally becom e larger with increasing temper ature. It is partly caused by the evaporation of chloride solutions at el evated temperatures. Although the mordenite samples and chloride solutions were contai ned in the sealed HDPE bottles during ion exchange, evaporation was inevit able when taking the solution for IC analysis. To avoid the impact of temperature c hange on the experiment, the solid sample was separated from solution by settling for a week in t he oven. And then a small amount of solution was taken out by pipet for dilution. Du ring the procedure of dilution by mass, evaporation could somewhat reduce the mass of original so lution, which increased the cation concentrations in diluted sample for IC analysis. Equilibrium constants determi ned for ion exchange between Na+ and K+ in mordenite are compared to the literature dat a (Table 4-4). The discrepancy is primarily caused by the different chemical compositions of the materials used in these studies. For instance, the same ion exchange experiments performed on zeolon (Ames 1964a) 101

PAGE 102

and AW-300 (Ames 1964b) presented different ln K although they are both synthetic mordenites. Note that Wolf et al. (1978) determined ln K for ion exchange between Na+ and K+ in a synthetic mordenite at 278 and 293 K, and the latter is lower than the former. This observation is coincident with our results that the value of ln K becomes lower when temperature increases from 323.15 to 348.15 K. The other thermodynamic properties of Na-K ion exchange in mordenite are also listed in Table 4-4. Gibbs free energy of reaction ( GR) can be calculated from ln K and enthalpy of reaction ( HR) was interpreted from Hf of all the phases in reaction 4-4. Entropy of reaction ( SR) can thus be calculated from the determined GR and HR by: SR = ( HR GR) / T (4-9) Because the thermal entropy of ion-exchan ged zeolites can be estimated by assuming equilibrium between the zeo lites and their component oxides (e.g., Robie and Hemingway 1991), previously determined ther mal entropy of natural mordenite by Johnson et al. (1992) was used to assess the entropies for Naand K-mordenite in this study. As a consequence, SR can be estimated from the entropies of all phoases in reaction 4-4, and the result (-26.95.23 J/mol K) is in agr eement with our value within the uncertainties (Table 4-4). Role of Water Molecules in Ion Exchange Ion exchange has a significant impact on the water content of mordenite, and water behavior also plays an important role in the ion exchange reaction. The loss of water molecules from the cryst al structure of mordenite dur ing the substitution of K+ for Na+ is caused by relatively bigger cation size of K+ than that of Na+. This change of water content in different cation forms of z eolites was also observ ed in clinoptilolite 102

PAGE 103

(Carey and Bish 1996; Yang et al 2001), chabazite (Shim 1999), and merlinoite (Donahoe et al 1990). It appears that cations and water molecules occupy certain part of space in the extraframework of mordenite. Ion exchan ge changes the space occupancy of cations and then water molecules wi ll be adjusted to yield or fill the extra space. The de/hydration occurs simultaneou sly with the ion ex change, and the amount of water lost/gained in this process should be proportional to the change of space for cations. The role of water during the ion excha nge in mordenite was analyzed in terms of the enthalpy of accompanying partial dehydr ation. The ion exch ange reaction can be divided into two parts regarding HR: ion exchange involving no water (reaction 4-10, HR,cation) and conversion between the water molecules in Naand K-mordenite (reaction 4-11, Hr,H2O). K+ + NaAlSi6.19O14.38 = Na+ + KAlSi6.19O14.38 (4-10) 4.19H2O(Na-MOR) = 3.78H2O(K-MOR) + 0.41H2O(liquid) (4-11) Enthalpy of reaction 4-10 was evaluated by Hf of dehydrated homoionic mordentie and cations, and HR,H2O was calculated from enthalpy of water molecules in Naand Kmordenite, which can be assesse d by the difference between Hf of hydrated and dehydrated mordenite. It can be seen that contributions of these two parts to the ion exchange reaction are very close but one is exothermic (reaction 410) and the other is endothermic (reaction 4-11), leadi ng to a relatively small HR of ion exchange (Table 44). The energetic of water behavior appears to be able to offset that of cation substitution in terms of the thermodynamics of ion exchange reaction. This offset effect 103

PAGE 104

was also observed in the other thermodynamic properties such as entropy and volume (see Chapter 2). Thermodynamic Analysis of Temperature Dependence of Ion Exchange Temperature dependence of the ion exchange equilibri um can be demonstrated by the variation of HR with temperature. The relationship between ln K and HR can be expressed by the vant Hoff equation: R 2ln H dK dTRT (4-12) where HR as a function of temperature can be derived from the heat capacity of reaction ( CP,R) via the relationship: R,R,298 P,R 298 T THHCd T (4-13) Enthalpy and heat capacity of reaction at 298.15 K were estimated by summation of Hf and CP of all the species in reaction 4-4, respectively. However, both Naand Kmordenite become partially dehydrated at elevated temperatures, and their water contents change with temperature. The water loss also changes Hf of mordenite, which contributes to the change of HR. As a consequence, this part should be considered in modeling t he temperature dependence of ion exchange. The temperature for the occurrence of mordenite is generally no greater than 523.15 K, and the partial dehydration of mordenite in this range of tempatures only occurs in W1. Therefore, the water content of mordenite as a function of temperature (Figur e 4-4) can be derived from the mordenite-H2O equilibrium data in Chapter 3. In combination with the thermodynamic properties of water in mordenite, the contribution of partial dehydration to HR can be determined at different temperatures In contrary to the contribution of 104

PAGE 105

CP,R which makes HR become less exothermic with in creasing temperature, the loss of water in both endmembers in creases the exothermisity of reaction. The results of these two parts show that when temper ature changes from 298.15 to 523.15 K, the magnitudes of their contributions to HR are quite similar. Therefore, HR has only a small change during this temperature range. Using the experimental value of ln K at 298.15 K as the initial point, the ln K of ion exchange in mor denite as a function of temperature is demonstrated from 298.15 to 523.15 K in Figure 4-5. The predicted temperature dependence of ln K is consistent with the experimental results within uncertainties (Figure 4-5), and it shows that the preference of mordenite for K+ over Na+ decreases with increasing temperatur e. At relatively low temperatures, the variation of lnK as a function of temperature depends primarily on the magnitude of HR (Eq. 4-12). When temperat ure increases, the impact of HR on the temperature dependence of lnK becomes small. Ther efore, the change of ln K at high temperatures is relatively smaller compared to that at low te mperatures. In order to find out the impact of partial dehydration of mordenite on te mperature dependence of ion exchange, ln K as a function of temperatur e without considering the hydrat ion state change of mordenite is also shown in Figure 4-5. It can be seen t hat at relatively low temperatures (i.e., <288.15 K) the two curves are similar to one another, because the cont ribution of partial dehydration in mordenite to HR is very small. However, at relatively high temperatures, the hydration state of mordenite changes dramtically with temperature, making this part of contribution overwhelm that of CP,R. As a consequence, high temperatures reduce the preference of mordenite for K+ over Na+. 105

PAGE 106

Excess Gibbs energy for mordenite solid solutions of mixing Na+ and K+ can be calculated from Margules parameters WK+ and WNa+ (Table 4-5) by the: Gex = (WK+XNa+ + WNa+XK+) XK+XNa+ (4-13) The results of Gex as a function of XK+ for mordenite solid solu tions at 298.15, 323.15, and 348.15 K are shown in Figure 4-6. Note that the maximum Gex of mordenite solid solution is -0.16 kJ/mol, which is within the uncertainty of GR. This small magnitude of Gex is consistent with that of Hex (Chapter 3), which confirms that mordenite of mixing Na+ and K+ is close to ideal solid solution. Excess entropy ( Sex) of mordenite solid solution is related to Gex and Hex by: Gex = Hex TSex (4-14) Excess entropy of mordenite so lid solution is positive (Table 4-5), which can not result from the ordering of cations in mordenite. A potential explanation is that mixture of Na+ and K+ in mordenite changes the occupancy of water molecules in extraframework, and the disordering of water molecules leads to a positive Sex. Geological Implications The temperature dependence of ion exchan ge in mordenite predicted by our thermodynamic model can help explain the variation of cation compositions in the natural mordenite. Passaglia (1975) examined the crystal chemistry of 35 mordenite samples taken from the lit erature and concluded that mordenite in igneous rocks showed small variations in chemistry, Si occupied 80-85% of t he tetrahedral sites and the extraframework cations are mainly Na+ and Ca2+ with minor K+. However, the diagenetic mordenite from Ponza island (e.g ., Passaglia et al. 1995) and Sardinia (e.g., Ghiara et al. 1999), Italy generally has high K+ content (K2O > 1.0 wt%). Mordenite is commonly formed by hydrothermal alteration of volcaniclastic minerals within veins, 106

PAGE 107

fissures, or amygdales of some igneous ro cks (Gottardi & Galli 198 5). The hydrothermal fluids interacting with the rock should have various cation compositions at different locations. Even when the fluid has very low concentrations of K+ similar to that of seawater (Na+ : K+ 46: 1), the content of K2O can be greater than 1.0 wt% in mordenite based on our experimental results of ion exchange at 298.15 K. A potential reason for the generally low K content of mordenite in the igneous rocks is the high temperature generated by t he hydrothermal circulati on. The temperature of hydrothermal fluid for the formation of mord enite is generally high er than 373.15 K (e.g., Kitsopoulos 1997; Pe-Piper 2000). The temperature dependenc e of ion exchange in mordenite predicted in this study suggests that the preference of mordenite for K+ drops a lot when the temperature reaches 373.15 K. On the other hand, the diagenetic mordenite is formed in hydrological open systems via hydrolysis of vitric material. The interacting fluid is probably percolating met eoric water or groundwat er (e.g., Passaglia et al. 1995; Ghiara et al. 1999). Besides the generally high concentration of K+ in these fluids, the low temperature can also encourage the formation of K-rich mordenite. In addition, Muhe and Stoffers (1995) studied zeolit es (i.e., mesolite, cowlesite, mordenite, and ferrierite) that precipitated exclusivel y in porous zones in the mid-oceanic ridge basalts. They estimated the formation temperat ure of the zeolites in porous zones to lie in the range of 423 573 K, and the dominance of Ca and Na over K in these zeolites distinguished them from the K-rich minerals assemblage t hat forms in the course of palagonitization (i.e., low temperature alteration) This observation is coincident with the results predicted by our thermodynamic model. 107

PAGE 108

Table 4-1. Experimental data for binary ion exchange between Na+ and K+ in mordenite with 0.1 N chloride solutions at 298.15 K Mordenite (mg) 0.1 N NaClc (mL) 0.1 N KCld (mL) mNa+ (mmol/mL) mK+ (mmol/mL) XK+ e 1a 250.1 9.5 0.5 97.728 0.780 0.090 2 251.3 9 1 98.008 1.606 0.178 3 153.5 9 1 97.015 2.690 0.252 4 101.3 9 1 95.866 3.855 0.320 5 152.7 8 2 90.161 7.040 0.448 6 189.9 7 3 89.201 10.033 0.555 7 168.5 7 3 86.937 11.747 0.570 8 172.2 6 4 79.876 18.109 0.666 9 131.6 6 4 77.758 22.021 0.709 10 124.2 5 5 69.107 30.482 0.810 11 150.4 4 6 60.608 35.695 0.834 12 167.5 3 7 57.053 42.372 0.850 13 166.7 2 8 47.811 50.680 0.902 14 167.1 1 9 39.946 59.902 0.918 15 167.2 0 10 29.936 68.871 0.944 16 130.4 0 10 25.586 74.727 0.965 17 101.1 0 10 16.617 79.690 0.976 18 49.4 0 10 11.738 88.489 1.023 19 61.4 0 20 7.040 91.566 1.101 20b 222.9 1 9 17.669 88.366 0.986 21 155.1 2 8 26.380 78.826 0.972 22 155.4 3 7 35.863 69.721 0.948 23 145.1 4 6 43.051 59.927 0.946 24 165.7 5 5 52.344 51.706 0.900 25 139.4 6 4 61.230 42.579 0.856 26 161.3 7 3 68.438 34.894 0.803 27 149.8 8 2 76.450 25.778 0.764 28 157.8 9 1 83.012 18.930 0.671 29 158.7 10 0 90.130 12.370 0.560 30 77.7 10 0 94.666 6.854 0.501 31 145.6 20 0 127.902 6.503 0.495 32 82.3 20 0 125.923 4.301 0.409 33 38.8 20 0 123.999 2.598 0.243 a1-19 were the experiments on Na-mordenite. b20-33 were the experiments on Kmordenite. cThe normality of NaCl solution m easured by IC was 99.978 mmol/mL. dThe normality of KCl solution measured by IC was 97.747 mmol/mL. eMole fraction of K+ in mordenite calculated from mass balance of IC data. 108

PAGE 109

Table 4-2. Experimental data for binary ion exchange between Na+ and K+ in mordenite with 0.1 N chloride solutions at 323.15 K Mordenite (mg) 0.1 N NaClc (mL) 0.1 N KCld (mL) mNa+ (mmol/mL) mK+ (mmol/mL) XK+ e 1a 251.0 9.5 0.5 123.075 1.346 0.076 2 249.2 9 1 124.976 2.050 0.167 3 155.0 9 1 113.598 2.800 0.242 4 104.7 9 1 120.202 3.900 0.301 5 149.9 8 2 110.368 7.387 0.435 6 192.5 7 3 107.664 11.285 0.503 7 171.9 7 3 119.696 12.234 0.533 8 179.2 6 4 114.692 17.766 0.637 9 133.6 6 4 105.264 21.604 0.698 10 126.5 5 5 98.079 30.624 0.764 11 152.9 4 6 90.427 35.795 0.793 12 166.2 3 7 84.707 42.946 0.812 13 167.6 2 8 78.285 50.350 0.879 14 167.6 1 9 71.275 58.471 0.930 15 165.8 0 10 70.165 68.180 0.938 16 131.4 0 10 68.308 74.209 0.933 17 103.4 0 10 60.654 78.877 0.939 18 53.5 0 10 47.933 87.466 0.937 19 64.5 0 20 45.787 90.295 1.074 20 31.1 0 20 44.879 93.825 0.985 21b 162.4 1 9 42.712 88.819 0.936 22 149.0 2 8 58.582 77.421 0.996 23 134.2 3 7 62.024 69.538 0.920 24 129.2 4 6 62.657 59.682 0.926 25 162.0 5 5 87.718 51.185 0.900 26 157.9 6 4 90.118 42.978 0.845 27 155.7 7 3 97.066 34.834 0.788 28 145.7 8 2 101.972 26.198 0.733 29 165.3 9 1 109.317 19.473 0.665 30 152.5 10 0 118.340 12.625 0.532 31 106.8 10 0 125.631 9.585 0.493 32 144.5 20 0 120.987 7.021 0.451 33 82.2 20 0 124.004 4.502 0.381 34 39.4 20 0 117.122 2.621 0.248 a1-20 were the experiments on Na-mordenite. b21-34 were the experiments on Kmordenite. cThe normality of NaCl solution m easured by IC was 111.406 mmol/mL. dThe normality of KCl solution me asured by IC was 96.632 mmol/mL. eMole fraction of K+ in mordenite calculated from mass balance of IC data. 109

PAGE 110

Table 4-3. Experimental data for binary ion exchange between Na+ and K+ in mordenite with 0.1 N chloride solutions at 348.15 K Mordenite (mg) 0.1 N NaClc (mL) 0.1 N KCld (mL) mNa+ (mmol/mL) mK+ (mmol/mL) XK+ e 1a 250.2 9.5 0.5 131.778 1.884 0.066 2 250.6 9 1 123.668 2.486 0.161 3 151.6 9 1 126.848 3.789 0.218 4 99.8 9 1 132.001 4.721 0.281 5 149.1 8 2 119.042 8.614 0.406 6 191.6 7 3 122.331 12.560 0.484 7 171.7 7 3 117.576 13.714 0.504 8 180.2 6 4 112.832 19.671 0.598 9 132.7 6 4 112.669 24.361 0.619 10 126.1 5 5 104.071 32.250 0.736 11 153.4 4 6 91.080 37.507 0.769 12 164.4 3 7 88.907 44.961 0.797 13 167.2 2 8 83.769 53.415 0.829 14 166.4 1 9 74.807 63.975 0.810 15 167.4 0 10 72.362 70.621 0.909 16 130.1 0 10 60.705 77.780 0.869 17 102.9 0 10 55.186 80.715 0.943 18 53.0 0 10 52.811 89.665 0.908 19 66.0 0 20 43.931 92.786 0.942 20 31.6 0 20 47.086 95.849 0.906 21b 159.1 1 9 53.390 90.388 0.937 22 151.1 2 8 55.381 82.368 0.866 23 134.1 3 7 69.788 72.925 0.832 24 129.0 4 6 76.733 62.774 0.838 25 161.1 5 5 70.710 53.742 0.842 26 158.2 6 4 92.220 45.677 0.775 27 154.1 7 3 105.336 36.966 0.728 28 147.2 8 2 110.843 27.449 0.702 29 165.1 9 1 116.385 21.379 0.605 30 152.6 10 0 112.593 14.266 0.472 31 107.6 10 0 115.552 10.865 0.429 32 143.9 20 0 126.228 8.027 0.370 33 82.4 20 0 126.778 5.322 0.270 34 38.7 20 0 131.319 3.109 0.092 a1-20 were the experiments on Na-mordenite. b21-34 were the experiments on Kmordenite. cThe normality of NaCl solution m easured by IC was 111.406 mmol/mL. dThe normality of KCl solution me asured by IC was 96.632 mmol/mL. eMole fraction of K+ in mordenite calculated from mass balance of IC data. 110

PAGE 111

Table 4-4. Thermodynamic properties of Na-K ion exchange in mordenite T (K) lnK GR (kJ/molcharge) SR (kJ/molcharge) HR (kJ/molcharge) HR,cation (kJ/molcharge) HR,H2O (kJ/molcharge) 298.15a 2.38.20 -5.90 .99 -15.10 .69 -10.40.49 -40.36.99 29.96.04 323.15a 2.49.39 -6.17 .97 -10.16.60 -9.20.49 -39.58.99 30.38.04 348.15a 2.20.39 -5.45 .97 -8.37.60 -7.95.49 -37.98.99 30.03.04 298.15b 1.94 298.15c 3.00 278.15d 2.45 293.15d 2.22 aThis study. bZeolon, Ames (1964). cAW-300, Ames (1964). dSynthetic mordenite, Wolf et al. (1978) 111

PAGE 112

Table 4-5. Margules paramet ers and excess properties of mo rdenite solid solution with 0.5 mole fraction of K+ T (K) WNa+ (kJ/molcharge) WK+ (kJ/molcharge) Gex (kJ/molcharge) Hex (kJ/molcharge) Sex (kJ/molcharge) 298.15 -0.47.08 -0.79 .13 -0.16.03 0.73 0.01 2.99.07 323.15 -0.37.13 -0.42 .18 -0.10.04 348.15 -0.26.15 -0.41 .24 -0.08.05 112

PAGE 113

Figure 4-1. Isotherm for binary ion exchange of K+/Na+ in mordenite with 0.1 N chloride solutions at 298.15K. The open circles show experimental results for ion exchange with Na-mordenite, and the solid ones represent those with Kmordenite. The curve depicts an equilibri um ion exchange is otherm computed based on regressed thermodyna mic parameters (see text). 113

PAGE 114

Figure 4-2. Isotherm for binary ion exchange of K+/Na+ in mordenite with 0.1 N chloride soluctions at 323.15 K. S ee figure caption to Figure 4-1 for explanation of symbols and curve. 114

PAGE 115

Figure 4-3. Isotherm for binary ion exchange of K+/Na+ in mordenite with 0.1 N chloride soluctions at 348.15 K. S ee figure caption to Figure 4-1 for explanation of symbols and curve. 115

PAGE 116

Figure 4-4. Hydration states of Naand K-mordenite as a func tion of temperature during ion exchange. 116

PAGE 117

Figure 4-5. The variation of equilibrium c onstant for reaction 4-4 as a function of temperature. The open circles are the experimental results at 298.15, 323.15, and 348.15 K. The solid curve depicts the temperatur e dependence of ion exchange in mordenite predicted by the thermodynamic model, and the dashed curve shows the prediction wit hout considering hydration state change of mordnite. 117

PAGE 118

118 Figure 4-6. Excess Gibbs free energyl of mordenite solid so lution as a function of XK+ at different temperatures.

PAGE 119

CHAPTER 5 EXPERIMENTAL INVESTIGATION OF CA-BEARING SOLID SOLUTIONS IN MORDENITE Introduction Zeolites often exhibit extensive solid solutions between Si and Al in their frameworks, between occupied and vacant wa ter sites within their channels, and among charge balancing extraframework cations such as Na+, K+, Ca2+, Sr2+, and Ba2+ (e.g., Passaglia and Sheppard 2001; Neuhoff and Ruhl 2006). The extent of extraframework cation substitution varies widely between zeo lite species, with some species exhibiting no substitution, some in which the substituti on is coupled to framework substitutions as in plagioclase feldspars, and others in wh ich extensive substitution occurs independent of framework composition (Neuhoff and Ruhl 2006). Although zeolites in the latter category often exhibit complete solid solution between homoionic forms for virtually all exchangeable cations (e.g., chabazite and cli noptilolite), in others only incomplete exchange is possible for some cations (e.g., Barrer et al 1973; Townsend and Loizidou 1984). Among the rock-forming zeolites, two of the most important examples of this phenomenon include the par tial exchange of Na+ and Ca2+ for K+ in phillipsite (Shibue 1998) and the partial exchange of Ca2+ for Na+, K+ in mordenite (Passaglia, 1975). Mordenite (nominally [Na2,K2,Ca]4Al8Si40O9628H2O) is a common secondary mineral in hydrothermal and sedimentary environments. It has been found in vugs of volcanic and intrusive rocks (e.g., Sukhes wala et al. 1974; Passaglia 1975; Wise and Tschernich 1978; Nativel 1986; Vezzalini et al. 1994), as well as a diagenetic product of silicic tuff, pitchstone and volcanic mari ne sediments (Passaglia and Sheppard 2001). Mordenite shows a high Si/Al ratio with nat ural samples ranging from 4.2 to 5.9, and synthetic samples as high as 10.1 (Gotta rdi and Galli 1985). Mordenite also has high 119

PAGE 120

water content (up to 15 wt.%) and high t hermal stability (Passaglia 1975). Mordenite exhibits extensive solid solution and r eady ion exchange between extraframework cations, most notably Na+, K+, and Ca2+. Alkali metal cations are more prevalent than Ca2+ and no mordenites with Ca2+ comprising greater than 70% of the extraframework charge are known (Figure 5-1; Passaglia 1975; Kitsopoulos and Dunham 1998; PePiper 2000; Bartlett 2002). Barrer and Kli nowski (1974) investigated the cation selectivity of a synthetic mordenite (Z eolon), and found that a ll the ion exchange reactions were reversible except those involving Ca2+ and Sr2+. It thus appears that the compositional limits observed in nature ar e related to the ion exchange behavior of mordenite. The goal of this study is to invest igate the ion exchange behavior of Ca2+ in mordenite and its relationship to natural mordenite compositions. The ion exchange experiments focused on the cation pairs of Ca2+/ Na+ and Ca2+/ K+ and determination of the maximum Ca2+ content through homoionic i on exchange. The behavior of water molecules in ion exchanged mordenite was studied through the dehy dration of these samples as a function of temperature. In comb ination with the result s of previous crystal structure refinements of mordenite, some important features of the crystal chemistry of Ca2+ in mordenite are elucidated that help ex plain in compositiona l variation in both natural and synthetic syst ems. Binary ion exchange equilibria involving Ca2+ exchange in mordenite are then used together with the binary Na+-K+ equilibria reported in Chapter 4 to develop a therm odynamic model for ternary exchange equilibria that facilitate assessment of exchangeable ion cont ents in mordenite as a function of fluid composition in natural systems. 120

PAGE 121

Materials and Methodology The mordenite samples used in this study were produced from synthetic Namordenite powder obtained from Zeolyst International (CBV 10A). The raw Namordenite powder was used in its obtained form, and K-mordenite was generated by immersing Na-mordenite in 1 N KCl solution for eight weeks with KCl solution replaced every week. Sample purity and phase identity were confirmed by X-ray powder diffraction (XRD) data collected at 25 C on a Rigaku Ultima IV X-ray diffractometer with Ni-filtered Cu K radiation generated at 40 kV and 44 mA. Bulk composition of dehydrated mordenite sample (T able 5-1) was determined to be Na1.67Al1.67Si10.33O24 by x-ray fluorescence (XRF) at the Department of Geociences, University of WisconsinMilwaukee (e.g., McHenry 2009). All mordenite samples were found to be very sensitive to the relative humidity (RH), and the water content varied at different RH. Therefore, the water contents measured by loss on ignition were not suitable for the samples in this study. Two types of experiments were c onducted to determine the maximum water content of mordenite. Firs t, complete dehydration of mordenite was conducted on a Netzsch STA 449C Jupiter simultaneous thermal analysis system by heating 20-30 mg of sample to 650 C (e.g., Neuhoff and Wang 2007b; Wang and Neuhoff 2008a, 2008b). The sample was stablized at ~ 40% RH (close to the limit of the Jupiter system) at the beginning of the experiment, so that the mass change was t he water content of sample at 40% RH. Second, the mass change of the same sample was measured as a function of RH between 5 and 95% RH by a TA Instrum ents Q5000 sorption analyzer so that the maximum sorption capacity of the sample could be extrapolated. The total water contents for Naand K-mordenite we re determined to be 14.27.14 wt% and 12.61.13 wt%, respective ly. As a consequence, Na1.67Al1.67Si10.33O24.99H2O and 121

PAGE 122

K1.67Al1.67Si10.33O24.32H2O were adopted as the chemical formulae of the mordenite samples used in this study. Homoionic experiments to produce Ca-mor denite were conducted by immersing ~ 10 g of Na-mordenite in 500 ml of 1 N CaCl2 solution. Attempts were also made to produce such material from K-mo rdenite, but owing to the high slectivity of mordenite for K+ (see below) these experiments were less successful and are not reported here. The high-density polyethylene (HDPE) bottle containing mordenite powder and CaCl2 solution was kept in an oven at 80 C. The bottle was frequently shaken to keep the powder and solution well mixed. The CaCl2 solution was replaced every three days, and the progress of the hom oionic exchange was monitored by inspecting Na+ concentrations in the replaced chloride solutions using ion select ive electrodes (ISE) attached to a Denver Instrument Model 250 electrochemical meter. Large interferences with the concentration of Na+ caused by the high concentrations of Ca2+ in replaced solutions were detected in the ISE measur ements. However, t he concentration of Na+ was found to decrease in each replaced solution. After four weeks of this process, nearly constant concentrations of Na+ were observed in the replaced chloride solutions, indicating no further ion exchange between the sample and CaCl2 solutions. The Caexchanged mordenite powder wa s separated from the soluti ons by centrifuge, then washed with deionized water and dried passively at room temperature. Ion exchange equilibria and ca tion selectivity between Ca2+ and either Na+ or K+ were determined through binary ion exchan ge experiments. In each experiment, a precisely weighed amount of homoionic Naor K-mordenite powder was placed into chloride solutions containing two competing cations in a known equivalent concentration 122

PAGE 123

ratio, but at constant normality (0.1 N). The 20 ml glass vials containing well mixed solutions and mordenite powder were kept in a climate-controlled lab with constant room temperature at 25 C, and frequently shaken for at least one week. Some of the experiments were conducted in duplicate and allowed to react fo r two weeks with results essentially identical to those r eacted for one week. Afte r reaction, the aqueous solution was separated from the solid an d sealed off for future analysis. The concentrations of cations in the experimental solutions were analyzed by a Dionex ion chromatography (IC) system in the hydrochemistry lab of the University of Florida. The instrument drift was checked by measuri ng the concentrations of a standard and blank solution after every five samples. Dehydration behavior of the produced mordenite samples was evaluated with the aid of the Netzsch STA 449C Jupiter device previously described. For each experiment, 20-30 mg of mordenite was placed into a Pt-Rh crucible with unsea led, perforated lids. Samples were heated under ultrapure, dry N2 gas at 15 K/min from 25 to 650 C. During the heating process, the mass change and heat flow of the sample were simultaneously recorded by thermogravimetric analysis (T GA) and differential scanning calorimetry (DSC) signals, respectively. Ca Exchange of Mordenite The chemical composition of Ca-mordenite generated through attempts at homoionic exchange from Na-mordenite is shown in Table 5-1. Despite rigorous attempts at exchanging Ca2+ into the mordenite structure, the maximum mole fraction of Ca2+ (XCa2+) generated was ~ 0.44. This value is close to the maximum XCa2+ (~ 0.5) observed in the natural mordenite samp les (Figure 5-1). The generation of Camordenite used Na-mordenite as the starti ng material, and there was supposed to be a 123

PAGE 124

large amount of Na+ in Ca-mordenite if the sample could not be fully ion exchanged. However, instead of Na+, a large fraction of K+ has been observed in Ca-mordenite (Table 5-1). The cause of this is appear s to be the presence of an unavoidable KCl impurity in the CaCl2 used to make the experimental fluids. Analysis of the CaCl2 solution showed that the ratio of cation equivalent fraction in the solution was Ca2+ : K+ : Na+ = 96.1 : 1.7 : 2.2. Due to the greater selectiv ity of mordenite for K+ over Na+, K+ was selectively incorporated into mordenite as the amount of Na+ in the system decreased during subsequent replacement of the experimental fluids. Figures 5-2 and 5-3 show the bi nary ion exchange isotherms for Ca2+/Na+ and Ca2+/K+ in mordenite, respectively The results are presented in terms of the equivalent fraction of Ca in the solid phase ( Ca) as a function of the equivalent fraction of Ca in the coexisting aqueous phase ( ECa) calculated by: ECa = 2 mCa2+ / (2 mCa2+ + mA+) (5-1) Ca = 2 XCa2+ / (2 XCa2+ + XA+) (5-2) where mCa2+ and mA+ are the molar concentrations of Ca2+ and Na+/K+ respectively, in the aqueous solutions; XA+ is the mole fraction of Na+ or K+ in the mordenite solid solution. The ion exchange experiments included several steps: analyzing the chemical composition of the initial Naand K-mordenite samples; preparing the initial chloride solutions; measuring the mass of mordeni te samples and aqueous solutions; diluting the ion-exchanged aqueous solutions; and determi ning the concentrations of cations by IC. Each step could cause somewhat error for the results. Equivale nt fractions in the solid phase were calculated by mass balance from the resultant aqueous concentrations. Uncertainties in Ca and ECa take into account both the uncertainties in 124

PAGE 125

the process of the ion exc hange experiments and uncertainties resulting from IC analyses and subsequent calculat ions. The uncertainty of ECa is generally lower than that of Ca, because the concentrations of cations in the aqueous solution were from direct measurements whereas t he major source of error for Ca involved the accumulation of errors resulting from the ca lculation of cation concentrations in solid phase. This is especially true for points at the extrema of isotherms, where small amounts of mordenite samples were used to obtain the values of ECa close to 0 and 1. Under these conditions, even a small error in the cation concentration can lead to large errors in Ca. As a consequence, the isotherm was c onstructed by interpretation from the data of both Ca2+ and Na+/K+. Repeated experiments s howed that analyses of Na+ were usually accompanied with relatively larger drift than those of K+ and Ca2+, which was probably caused by the residual Na+ in the instrument after long-term use for seawater. The isotherm points calculated fr om different cations show good agreement with one another within the uncertainties. Due to the incomplete exchange of Ca2+ into mordenite realized in both the homoionic and binary exchange ex periemtns, thermodynamic analysis of the isotherms shown in Figures 5-2 and 5-3 was performed by normalizing Ca to the observed maximum value of Ca (Ca,max = 0.67, corresponding to XCa of 0.5) in natural samples (e.g., Barrer et al. 1973; Pabalan and Be rtetti 2001) by the following equation Ca = Ca (5-3) where Ca is normalized equivalent fraction of Ca in the solid phase; and refers to the normalization coefficient, equaling 1/ Ca,max. The ion exchange reaction involving two competing cations Ca2+ and Na+ in mordenite can then be expressed as 125

PAGE 126

0.5Ca2+ + Na2Al2Si12.38O28.76.38H2O + ( n -8.38)H2O(liquid) = Na+ + NaCa0.5Al2Si12.38O28.76 n H2O (5-4) where n refers to the maximum number of wate r molecules in Ca-mordenite at 25 C, which could not be determined due to our inabi lity to generate the composition of the maximum Ca mordenite endm ember. Note that the Ca exhibits unit value for a sample whose composition is equival ent to the Ca-bearing mordeni te endmember in reaction 54. Standard states fo r the homoionic mordenite and liquid water are unit activity of the pure phase at all temperatures and pressure s. For the aqueous i ons, the standard state corresponds to unit activity of a hypothetical one molar solution of the ion referenced to infinite dilution at all temperatures and pressures. The normalized isotherm for reaction 5-4 is also shown in Figure 5-2, and the equilibrium constant of reaction ( K ) can be expressed as K = (mNa+ Na+)(XCa2+ Ca2+) / [( mCa2+ Ca2+)0.5( XNa+Na+)(aH2O)(n-8.38)] (5-5) where X is the normalized mole fraction of cations calculated from ; XCa2+ and XNa+ are the mole fractions of the homoionic Ca2+ and Na+ forms in reaction 5-4, respectively; mCa2+ and mNa+ are the molalities of the free aquo calcium and sodium ions, respectively; is the activity coefficient of cation in the aqueous solution; refers to the rational activity coefficient for mordenite component and account for the nonideality in zeolite phase; and aH2O is the activity of water, which is ta ken to be unity in the relatively dilute experimental solutions of this study. The aqueous phase activity coefficients were calculated from the measured experimental solution compositions using the computer code EQ3NR (Wolery, 1991). Associated with a ll the measurable param eters in Eq. 5-5, a Vanselow coefficient (Kv) is defined as 126

PAGE 127

Kv = (mNa+ Na+) XCa2+ / [( mCa2+ Ca2+)0.5XNa+] (5-6) The consequences of non-ideal mixing between the mineral components in reaction 5-4 leading to the non-integer values of in Eq. 5-5 represented through an asymmetric Margules formulation that has been successf ully applied to the binary ion exchange in other zeolites (e.g., Pabalan 1994; Shibue 19 98; Fridriksson et al. 2004). In this model, the activity coefficients may be represented by lnCa2+ = [(2 WNa+ WCa2+)XNa+ + 2( WCa2+ WNa+)XNa+] / ( RT ) (5-7) and lnNa+ = [(2 WCa2+ WNa+)XCa2+ + 2( WNa+ WCa2+)XCa2+] / ( RT ) (5-8) where WCa2+ and WNa+ are binary Margules in teraction parameters, R is the gas constant (8.314 J/mol K), and T is temperature in Kelvin. Substitu tion of Eq. 5-6, 5-7, and 5-8 into Eq. 5-5 yields lnKv = ln K WCa2+(2XCa2+ + 2XNa+ 2 XCa2+ XNa+) / ( RT ) + WNa+(2XCa2+ + 2XNa+ XCa2+ 2 XNa+) / ( RT ) (5-9) Nonlinear regression of equation Eq. 5-9 allows determination of ln K WCa2+, and WNa+ (Table 5-4) for reaction 5-4 at 25 C The above thermodynamic model can be applied to the ion exchange between K-mordenite and CaCl2 solution by replacing Na+ in reaction 5-4 and the corresponding data a nd stoichiometries with those for K+, and the results are list ed in Table 5-4. The thermodynamic parameters determined by Margules equation were used to predict the isotherm curves fo r the ion exchange reactions (Figure 5-2 and 5-3). It can be seen that the predictions are in good agr eement with both normalized and original experimental data. The results from this st udy were compared with the literature data. 127

PAGE 128

Tarasevich et al. (2006) det ermined the isotherms for Ca-Na-mordenite ion exchange system, and the results were similar to that in this study. Their results not only showed incomplete exchange of Ca2+ for Na+ (with Ca,max of ~ 0.7), but also indicated that mordenite was more selective for Ca2+ than Na+ at low ECa and the selectivity reversed at high ECa. Ames (1964b) also derived K for Ca2+/Na+ exchange in a synthetic mordenite (Zeolon), but no corresponding isot herm was shown in his study. There are relatively large discrepancies between the values of K and the Margules parameters determined in this study and thos e found in the literature for Ca2+/Na+ exchange. The primary reason for this discrepancy is likely the nature of mordenite samples used in these studies. For instance, Tarasevich et al. (2006) performed same ion exchange experiments for two diffe rent mordenite samples, and the values of K suggested different preference for Ca2+. In the case of Ca2+/K+ exchange in mordenite, no literature data were available for comparison. However, our results indicate that mordenite has more preference for K+ over Ca2+, which is coincident with the general selectivity sequence postulated by Suzuki et al (1978) for a synthetic mordenite. Water Molecules in Ion-exchanged Mordenite The dehydration of ion exchanged mordenite, CaNa-mordenite and CaKmordenite, as a function of temperature is compared with that of homoinic Naand Kmordenite in Figures 5-4 and 5-5, respecti vely. The water content and behavior of water molecules are reflected by the TGA curve and its first derivative (dTGA), respectively. It can be seen that Naand K-mordenite can be fully dehydrated at 650 C, whereas the dehydration of Ca-mordenite is incomplete at this temperature (as manifested by fact that mass was still decreasing at this temper ature). Two energetically distinct types of water are observed in mordenite denoted W1 (wat ers lost at relatively low temperatures) 128

PAGE 129

and W2 (waters lost at relatively high temperatures; see Chapter 3). The ranges of temperature for the peaks of W1 and W2 shown by DSC and dTGA are exactly the same for the three ion-exch anged mordenite samples. There is also some obvious diffe rence among the DSC and dTGA for the dehydration of ion-exchanged mordenite, indicating the cation composition can affect the abundance and energeti cs of water molecules in mo rdenite. Substitution of K+ for Na+ has significant impact on the abudance of W1 but not their energetics, and the energetics of W2 water but not their abunda nce (see Chapter 3). The dTGA of Camordenite is quite similar to that of it s corresponding monovalent endmember. It thus appears that the ion ex change involving Ca2+ has little effect on the energetics of W1 and W2. However, the total water content in Ca-mordenite is higher than both Naand K-mordenite, which is attributed to the rela tively smaller size and high valence of Ca2+ that collectively permit more space for wate r molecules within the mordenite channels. In addition, the dehydration of Ca-mordenite is incomplete at 650 C, suggesting that the exchange of Ca2+ for Na+ or K+ may increase the abundance of both W1 and W2. Martucci et al (2003) studied the step by step thermal dehy dration process of mordenite from Pashan (Poona, India) using the time -resolved synchrotron powder diffraction. They also found that the water site B which was co-occupied by Ca2+ could not be completely dehydrated above 630 C, which is consistent with our observations. Controls on Natural Mordenite Compositions The exchange systems of Ca2+/Na+ and Ca2+/K+ in mordenite behave differently with ECa. The isotherm for Ca2+/Na+ shows that mord enite prefers Ca2+ to Na+ at low ECa, but the selectivity becomes reversed at higher ECa, leading to the opposite signs of Margules parameters for this reaction. The isotherm for Ca2+/K+ indicates that mordenite 129

PAGE 130

has a preference for K+ over Ca2+ over the whole range of XCa2+, which is consistent with the negative value of lnK (Table 5-4). Nonetheless, both ion exchange systems are incomplete, and the isotherms de monstrate that the maximum XCa2+ in mordenite is ~ 0.5, which is consistent with the compositio ns of natural samples shown in Figure 5-1. Passaglia (1975) investigated the crystal chemistry of 21 natural mordenite samples from different locations. The XCa2+ of these samples has been calculated to range from ~ 0.25 to ~ 0.46. The studies on mordenite from Polyegos Island, Greece (Kitsopoulos and Dunham 1998), Sardinia, Italy (Ghiara et al. 1999), Nova Scotia, Canada (Pe-Piper 2000), and Marysvale, Utah (Barlett 2002) also showed no mordenite with XCa2+ exceeding 0.5. The incomplete exchange of Ca2+ into mordenite likely reflects the crystallochemical consequences of the high Si content of mordenite. At relatively high Si contents, the framework charge due to the presence of Al (which is ultimately accommodated by the presence of the exchangeab le extraframework cations) is fairly dilute. Thus, the statistical chances (assuming complete Si-Al disorder) of two Al atoms occurring in close proximity within the fr amework (in order to accommodate the local charge balance constraints imposed by a divalent extraframework cation like Ca2+) are low. This is compounded by Al-avoidance in al uminosilicate structures (cf. Lowenstein, 1954) which precludes the occurrence of Al-O-A l linkages in the framework. Cation sites D ( in the twelve-membered rings) and A (in the eight-membered rings) are located near the four-membered rings that are normally somewhat enriched in Al, thus potentially providing sufficient local framework c harge to acceommodate (nonexclusively) a divalent ion such as Ca2+ (e.g., Alberti 1997; Simonc ic and Armbruster 2004). In 130

PAGE 131

contrast, the E cation site is located within the large channel s in the structure which are somewhat depleted in Al relative to the bulk. We propose that site E will generally only accommodate monovalent cations in order to meet the constrai nts of local charge balance in the large channels. The ion exchange equilibria for binary systems including K+/Na+, Ca2+/Na+ and Ca2+/K+ in mordenite can be used to assess the chemical composition of natural fluids interacting with mordenite. Elprince et al. (1980) employed the subregular model to determine the composition of an aqueous solution including three cations as a function of exchanger phase composition. Because the interaction parameters in mordenite solid solutions are relatively small, the subregul ar model can be applied in this study to determine the activity coefficient of ionexchanged mordenite in a ternary system (Na-KCa). Note that the thermodynamic parameters for binary system of Ca2+/Na+ and Ca2+/K+ were interpreted from normalized data, the actual XCa2+ in mordenite needs to be normalized to the observed maximum XCa2+ (0.5, corresponding to Ca,max = 0.67) in the subregular model. The mole fractions of Na+ and K+ are thus adjusted corresponding to the normalized XCa2+, but with the same Na/K ratios. In the following model, subscripts 1, 2, 3 stand for the solid phase of Na, K, and Ca, respectively. Taking reaction 5-4 as an example, the ex cess Gibbs free energy for the Na-Ca binary system ( ) can be calculated by the Margules parameters interpreted from ion exchange isotherm: 13 exG13 ex133113(GWXWXXX ) (5-10) where X is the mole fraction of cations in the binary system, X3 is the normalized mole fraction of Ca2+, and X1+X3=1. When subregular model is applied to a binary ion 131

PAGE 132

exchange system, can be expressed by the follo wing equation (e.g., Hardy 1953; Grover 1977): 13 exG13 ex 131313113[ln(lnln)]GRTXXfXff (5-11) where is the activity coefficient of solid phase 1 when its mole fraction approaches zero in a binary mixure containing 1 and 3. The values of 13f13lnf and can be determined by a linear r egression of the plot 31lnf13 ex 1 2 / (GRTX)f X vs. X1. The activity coefficients for the other two binary system s can be determined by the same method. When the subregular model is ap plied to a ternary system, the activity coefficients of the three solid phases can be expressed by the following equations (e.g., Currie and Curtis 1976; Elprince et al. 1980): (1)2(1)2(1) (1)3(1)3(1)2(1)2 12223332323222233332232333232lnfcxcxcxxdxdxdxxdxx (2)2(2)2(2) (2)3(2)3(2)2(1)2 21113331313111133331131333131lnfcxcxcxxdxdxdxxdxx and (3)2(3)2(3) (3)3(3)3(3)2(3)2 31112221212111122221121222121lnfcxcxcxxdxdxdxxdxx (5-12) where x is the mole fraction of cati ons in the ternary stystem, and x1+ x2+x3=1; c and d are constant coefficients, and (1) 22 21122lnlncf, (1) 23 31211213321 (3ln3lnlnlnln 2 cfffff )ff, (1) 222 12212(lnln) df and (1) 223 1221133123323(lnln)(lnlnlnln) dfffff (5-13) 132

PAGE 133

With the activity coefficients and mole fractions of solid phases in the tenary system, the activities of aqueous phases (in terms of aCa2+/ aNa+ 2 and aCa2+/ aK+ 2) can be calculated by Eq. 5-5. Results of the above analyses were used to generate a phase diagram (Figure 5-6) illustrating the exchan geable ion content of mordenite as a function of the composition of coexiting solutions. The curves are isopl eths of Ca content in mordenite, which increases to the upper right of the diagram. The diagonal lines repres ent isopleths of XNa+/( XNa++ XK+) in mordenite, increasing to the lower right. The corresponding fluid compositions for the format ion of natural mordenite samples (Figure 5-1) were calculated and plotted in Figure 5-6. The chemic al compositions of some natural fluids associated with the formation of some mo rdenites such as seawater, hydrothermal fluids in Skagafjordur, Iceland (S. Arnor sson, personal communication), and present Yucca Mountain water (Kerrisk 1987) are shown for comparion. The veracity of the phase relations shown in Figure 5-6 is illustrated by the excellent agreement between t he compositions of nominally coexisting modenite and solution compositions. The solution composit ions reported from Skagafjordur, Iceland, are meteoric solutions that have exper ienced varying degrees of basalt-water interaction over the temper ature range (40-120 C) in whic h mordenite is common in Icelandic lavas (Neuhoff, unpublished data). T he mordentie analysis from Teigarhorn Iceland reported by Passaglia (1975) lies within this field. Similarly, the mordenite composition reported by Broxton et al. (1987) from Yu cca Mountain, Nevada, USA is consistent with the composition of present-day solutions report ed by Kerrisk (1987). 133

PAGE 134

It can be seen that the wide range of c hemical compositions of natural fluids results in quite different cation compositions in mordenite (e.g., Passaglia 1975). Bartlett (2002)s samples were collected from the same area (Marysvale, Utah), and the same fluid resulted in similar compositions in t hese samples. However, some of the samples have higher XCa2+ than the others, and the Ca content of the corresponding fluids are relatively high. A possible scenario is t hat the fluids were mixed with Ca-rich groundwater. The samples from Kitsopoulos and Dunham (1998) were divided into two groups based on their chemical compositions: typical NaCa-mordenite and K-rich mordenite. Kitsopoulos and D unham (1998) argued that t he typical NaCa-mordenite was formed by hydrothermal alteration and the hydrothermal fluids were recharged by seawater. This argument is consistent with the results shown in Figure 5-6, the fluid compositions for this group of samples are relatively close to that of seawater, but with lower Na+ and higher Ca2+ content. A potential reason is t hat the seawater was diluted by Ca-rich groundwater. In the ca se of K-rich mordenite, the K+ contents of some of the corresponding fluids are higher than that of seawater (Figure 5-6), but there was no evidence found for any other source of K in that area. As a consequence, they suggested that the reason for K-rich mordenite is that it had K-rich clinoptilolite crystals as precursors. The ion exchange results of our study indicate that mordenite has a higher preference for K+ over Ca2+ and Na+. Therefore, even the K-rich mordenite had further contact with hydrothermal fluids in rich of Na+ and Ca2+, it could keep the high K+ conent. The mordenite samples formed by hy drothermal alteration in Nova Scotia, Canada (Pe-Piper 2000) show sim ilar Na/K ratio but wide range of XCa2+. Fluid compositions for some of the samples are quite similar to that of seawater, indicating 134

PAGE 135

the equilibrium of these samples with seaw ater, potentially due to reequilibration after formation. The others wi th relatively higher XCa2+ reflect the fact that the fluids involved in the occurrence of these samples were s eawater modified by fl uid-rock interaction. A critical point for the geological application of our thermodynam ic model is that this model was developed based on ion exch ange in a synthetic instead of a natural sample. Most natural mordenites have properties characteristic of small-port mordenites, whereas the synthetic sample in this study is large-port mordenite. A major difference between the largeand small-port mordenite is the size of ex traframework channel. Molecules with a diameter > 4.5 can be introduced into large-port mordenite, and small-port mordenite can only accept molecu les with a diameter < 4.2 (Simoncic and Armbruster 2004). However, the diameters of K+, Na+, and Ca2+ are all smaller than 4.2 Consequently, there should be little difference between the gener al selectivity of natural and synthetic mordenite in term s of these three cations. Another major difference between natural and synthetic mordeni te is the Si-Al ordering, which has a significant effect on the extraframework char ge as well as the distribution of cations in mordenite (Simoncic and Armbruster 2004) For instance, the compressed eightmembered rings of natural mordenites are nor mally enriched in Al, providing sufficient extraframework charge for Ca2+. However, there are less Al in the compressed eightmembered rings of synthetic mordenites, and correspondingly less extraframework charge for local balance. As a consequence, the Ca occupancy in natural and synthetic mordenite might be somewhat different. Neve rtheless, the phase relations predicted by our model still show excellent agreement with the geological observations. In addition, 135

PAGE 136

the application of thermodynamic model serves as a heuristic tool for interpreting the natural conditions from experimental results. 136

PAGE 137

Table 5-1. Compositions of dehydrated ion-exchanged mor denite samples analyzed by XRF Sample Na-mordenite K-mordenite Ca-mordenite Component concentrationa (wt.%) Na2O 6.00 0.02 0.67 MgO 0.06 0.02 0.04 Al2O3 10.18 10.6 10.28 SiO2 74.93 77.43 75.21 P2O5 0.02 0.03 0.02 K2O 0.04 9.21 2.72 CaO 0.04 0.03 3.4 TiO2 0.39 0.38 0.36 Fe2O3 0.16 0.15 0.15 Total 91.82 97.87 92.85 Cations per 24 framework Ob Na 1.580 0.005 0.176 Mg 0.012 0.004 0.008 Al 1.630 1.644 1.640 Si 10.179 10.191 10.180 P 0.001 0.002 0.001 K 0.007 1.547 0.470 Ca 0.006 0.004 0.493 Ti 0.178 0.168 0.164 Fe 0.016 0.015 0.015 aDirectly measured from XRF. bNormalized to LOI-free 100 wt.%. 137

PAGE 138

Table 5-2. Experimental data fo r binary ion exchange involving Ca2+ and Na+ in mordenite with 0.1 N chloride solutions at 25 C Na-MOR (mg) 0.1 N NaCla (g) 0.1 N CaCl2 b (g) mNa+ (mmol/kg) mCa2+ (mmol/kg) Na Ca 1 253.4 9.5 0.5 96. 618 0.256 0.972 0.084 2 251.6 9 1 95.818 0.583 0.887 0.167 3 199.2 9 1 97.013 0.789 0.825 0.199 4 149.6 9 1 95.110 1.094 0.837 0.243 5 240.6 8 2 93.991 1.986 0.712 0.311 6 147.4 8 2 90.537 3.652 0.657 0.384 7 173.8 7 3 82.796 6.461 0.659 0.427 8 192.4 6 4 79.838 11.322 0.511 0.360 9 202.8 5 5 70.255 14.148 0.542 0.428 10 201.1 4 6 60.863 17.882 0.540 0.468 11 199.0 3 7 52.836 21.869 0.499 0.497 12 201.5 2 8 44.506 26.309 0.478 0.489 13 200.7 1 9 34.782 30.353 0.486 0.511 14 250.8 0 10 30.400 31.961 0.480 0.532 15 201.6 0 10 25.697 34.117 0.481 0.545 16 168.4 0 10 22.062 36.201 0.497 0.517 17 120.1 0 10 17.633 38.593 0.496 0.507 18 180.9 0 20 14.897 39.785 0.496 0.529 19 122.6 0 20 11.497 41.768 0.553 0.433 aThe normality of NaCl solution measured by IC was 99.978 mmol/kg. bThe normality of CaCl2 solution measured by IC was 88.319 mmol/kg. 138

PAGE 139

Table 5-3. Experimental data fo r binary ion exchange involving Ca2+ and K+ in mordenite with 0.1 N chloride solutions at 25 C K-MOR (mg) 0.1 N KCla (g) 0.1 N CaCl2 b (g) mK+ (mmol/kg) mCa2+ (mmol/kg) K Ca 1 161.9 9 1 88.085 4.333 0.996 0.017 2 155.9 8 2 81.437 8.351 0.889 0.059 3 149.0 7 3 71.448 12.552 0.899 0.090 4 159.1 6 4 62.851 16.623 0.870 0.121 5 146.3 5 5 54.027 20.783 0.829 0.164 6 155.8 4 6 45.974 23.715 0.783 0.274 7 154.7 3 7 37.547 28.793 0.739 0.239 8 154.8 2 8 29.192 32.216 0.694 0.324 9 156.1 1 9 21.150 36.244 0.640 0.361 10 155.7 0 10 13.251 40.881 0.577 0.358 11 84.2 0 10 8.629 42.301 0.529 0.471 12 64.9 0 20 4.575 44.302 0.484 0.526 aThe normality of KCl solution me asured by IC was 97.578 mmol/kg. bThe normality of CaCl2 solution measured by IC was 91.623 mmol/kg. 139

PAGE 140

Table 5-4. Thermody namic properties of ion exchange between CaCl2 solution and Na-, K-mordenite T (C) ln K WCa2+ (kJ/molcharge) WK+ (kJ/molcharge) WNa+ (kJ/molcharge) K+/Na+a 25 2.38.20 -0.79.13 -0.47.08 Ca2+/Na+a 25 -1.05.32 -1.53.65 -5.86.79 Ca2+/K+a 25 -3.25.26 0.20.10 -6.17.85 Ca2+/Na+b 25 -0.37 Ca2+/Na+c 20 0.04 Ca2+/Na+d 20 -1.55 aThis study. bZeolon, Ames (1964b). cTranscarpathian Na-form, Ta rasevich et al. (2006). dCrimean Na-form, Tarasevich et al. (2006). 140

PAGE 141

Figure 5-1. Extraframework cation content s reported for natural mordenites from a variety of different parageneses along with the composition of the maximum Ca-bearing mordenite generated in this study. 141

PAGE 142

Figure 5-2. Binary i on exchange isotherm for Ca2+/Na+ in Na-mordenite with 0.1 N chloride solutions at 25 C. The circles are the original experimental data, and the triangles depict the normalized data (see text). The open symbols show the experimental results calculat ed from the concentrations of Ca2+, and the solid ones represent those calculated from Na+. The dashed and solid curves represent the predicted isotherms for normalized and original experimental data, respectively. 142

PAGE 143

Figure 5-3. Binary i on exchange isotherm for Ca2+/K+ in K-mordenite with 0.1 N chloride solutions at 25 C. The circles are t he original experim ental data, and the triangles depict the nor malized data (see text). The open symbols show the experimental results calculated from the concentrations of Ca2+, and the solid ones represent those calculated from K+. The dashed and solid curves represent the predicted isotherms for normalized and original experimental data, respectively. 143

PAGE 144

Figure 5-4. The dehydration behavior of Na-Mordenite (1) and CaNa-Mordenite (2) between 25 and 650 C. Dashed curves represent the mass change during the dehydration. Waved curves are the first derivative of mass change. Solid curves show the heat flow as a function of temperature. 144

PAGE 145

Figure 5-5. The dehydration behavior of K-Mordenite (1) and CaK-Mordenite (2) between 25 and 650 C. See figure captio n to Figure 3 for explanation of curves. 145

PAGE 146

146 Figure 5-6. Phase diagram illustrating the ex traframework cation contents of mordenite as a function of solution compositi on at 298.15 K, 1 bar. The curves are isopleths of Ca content in mordenite, represented by XCa2+. The diagonal lines represent XNa+/( XNa++ XK+) in mordenite, increasing to the lower right. Shown for comparison are predicted fluid compos itions in equilibrium with mordenite samples reported by Passaglia ( 1975), Barlett (2002), Kisopoulos and Dunham (1998), Pe-piper (2000), and Brox ton et al. (1987) along with the compositions of seawater and fluids associated with mordenite formation at Skagafjordur, Iceland (S. Arnorsson, personal communication and Yucca Mountain, Nevada, USA (Kerrisk 1987).

PAGE 147

CHAPTER 6 CONCLUSIONS A thermodynamic model for ion exchange as well as the water behavior in zeolites has been developed in this study, which prov ides important heuristic insights into the energetics and stability of zeolites. Thermo gravimetric and calorimetric analyses on the de/hydration of ion exchanged mordenite reveal that ion exchange has a significant impact on the behavior of water molecules in mordenite. Results for mordenite solid solutions containing Na+ and K+ indicate that the maximum water content of mordenite decreases and the enthalpy of hydration becomes less exothermic with increasing XK+. More importantly, both maximum water content and enthalpy of hydr ation of mordenite are demonstrated to change linearly with XK+. Two energetically dist inct sets of water molecules have been observed in mordenite : W1 (relatively less energetic) and W2 (relatively high energetic). This observation is consistent with the results of structure refinement of mordenite at different temperatures. T hermodynamic analysis on the hydration of W1 and W2 reveals that these two sets of water molecules behave differently when cation composition of mordenite changes. The water content of W2 is independent of cation com position, whereas W1 loses water with when XK+ increases; the enthalpy of hydration of W1 appears to be constant, but that of W2 becomes less exothermic with increasing XK+. The determined therm odynamic properties of mordenite solid solution including volume and enthalpy of formation also vary with cation composition. There is very little ex cess enthalpy and excess Gibbs free energy in mordenite solid solutions, sugges ting that the mixture of Na+ and K+ in mordenite is nearly ideal. 147

PAGE 148

148 The systematic study on the binary ion exchange invoving Na+, K+, and Ca2+ mordenite illuminates the correlation between t he chemical compositions of zeolites and natural fluids. Isotherms of these binary ion exchange system s show that mordenite has a general cation selectivity of K+> Na+ >Ca2+. Homoionic Naand K-mordenite can be generated by monoionic exchan ge, whereas the maximum Ca2+ content (mole fraction) generated in the laboratory is ~ 0.44, which is consistent with the crystal chemistry results of numerous natural mordenite samp les. The subregular model was employed to calculate the thermodynamic prameters of a ternary ion exchange system based on those of binary systems determined in th is study. The phase diagram generated based on the thermodynamic data demons trates that the wide range of chemical compositions of natural fluids result in quite different cation compositions in mordenite. The fluid compositions of some mordenite samples interpreted by our m odel are in agreement with the local observations. The results of Na-K ion exchange in mo rdenite at three different temperatures show that mordenites preference for K+ over Na+ decreases with increasing temperature. A thermodynamic model has been built to predict the temperature dependence of ion exchange in mordenite. The pr ediction is generally consistent with the experimental results within the unc ertainties. A potential reason for the discrepancies is that the Margules formulat ions applied to the experimental data does not take into account the hydration state ch ange of mordenite at elevated temperatures. As a consequence, previous thermodynamic models for ion exchange in zeolites need to be modified to include the behavior of wa ter molecules during ion exchange reaction. This unsolved problem will be the focus of our future work.

PAGE 149

LIST OF REFERENCES Alberti, A., Davoli, P ., and Vezzalini, G. (1986) The cr ystal structure refinement of a natural mordenite. Zeitschrift Fu r Kristallographi e, 175, 249-256. Alberti, A. (1997) Location of the Bronsted sites in mordeni te. Zeolites, 19, 411-415. Alt, J.C., Honnorez, J., Laverne, C., and Emmermann, R. ( 1986) Hydrothermal alteration of a 1 km section thr ough the upper oceanic crust, deep-sea drilling project hole 504B: Mineralogy, chemistry, and evolution of seawater-basalt interactions. Journal of Geophysical Research, 91, 10309-10335. Ames, L.L. (1960) The cation si eve properties of clinoptilolit e. American Mineralogist, 45, 689-700. (1961) cation sieve properties of the open zeolites chabazite, mordenite, erionite and clinoptilolite. Americ an Mineralogist, 46, 1120-1131. (1964a) Some zeolite equilibr ia with alkali metal ca tions. American Mineralogist, 49, 127-145. (1964b) Some zeolite equilibria with alkaline earth metal cations. American Mineralogist, 49, 1099-1110. Armbruster, T. and Gunter, M.E. (2001) Crystal structures of natural zeolites. In D.L. Bish and D.W. Ming, Eds., Natural Zeolit es: Occurrences, Properties, Applications, 45, p. 1-67. Reviews in Mineralogy and G eochemistry, Mineralog ical Society of America, Chantilly, Virginia. Baldar, N.A. and Whittig, L. D. (1968) Occurrence and synthes is of soil zeolites. Soil Science Society of America Proceedings, 32, 235-238. Bartlett, E. (2002) Crystallochemical variat ions and genesis of mordenite, Marysvale, Utah. MS thesis, Brigham Y oung University, Provo, UT. Barrer, R.M. (1978) Zeolites and clay minerals as sorbents and molecular sieves, 497 p. Academic Press, London. Barrer, R.M. and Cram, P.J. (1971) Heats of immersi on of outgassed ion-exchanged zeolites. Advances in Chemistry Series, 102, 105-131. Barrer, R.M. and Klinowsk.J. (1974) Ion-exchange selectivity and electrolyte concentration. Journal of the Chemical Society Faraday Transactions I, 70, 20802091. 149

PAGE 150

Barrer, R.M., Klinowsk.J., and Sherry, H.S. (1973) Zeolite exchangers: Thermodynamic treatment when not all ions are exchangeable. Journal of the Chemical Society Faraday Transactions I, 69, 1669-1676. Benning, L.G., Wilkin, R.T., and Barnes, H.L. (2000) Solubility and stability of zeolites in aqueous solution: II. Calcic clinoptilolite and mordenite. Americ an Mineralogist, 85, 495-508. Berman, R.G. and Brown, T.H. (1985) Heat capacity of mineral s in the system Na2OK2O-CaO-MgO-FeO-Fe2O3-Al2O3-SiO2-TiO2-H2O-CO2 representation, estimation, and high-temperature ex trapolation. Contributions to Mineralogy and Petrology, 89, 168-183. Bish, D.L. (1984) Effects of exchangeable cation composition on t he thermal-expansion contraction of clinoptilolite. Cl ays and Clay Minerals, 32, 444-452. Bish, D.L. and Carey, J.W. (2001) Thermal behav ior of natural zeolites. In D.L. Bish and D.W. Ming, Eds., Natural Zeolites: Occu rrences, Properties, Applications, 45, p. 403-452. Reviews in Mineralogy and Geochem istry, Mineralogical Society of America, Chantilly, Virginia. Bish, D.L., Vaniman, D.T., Ch ipera, S.J., and Carey, J.W. (2003) The distribution of zeolites and their effects on the perform ance of a nuclear waste repository at Yucca Mountain, Nevada, USA. American Mineralog ist, 88, 1889-1902. Boettinger, J.L. and Southard, R.J. (1995) Phyllosilicate distribution and origin in aridisols on a granitic pediment, Western Mojave Desert. Soil Science Society of America Journal, 59, 1189-1198. Boles, J.R. and Coombs, D.S. (1977) Zeolite facies alteration of sandstones in Southland Syncline, New Zealand. Amer ican Journal of Science, 277, 982-1012. Broxton, D.E., Bish, D.L., and Warren, R. G. (1987) Distributi on and chemistry of diagenetic minerals at Yucca Mountain, Nye County, Nevada. Clays and Clay Minerals, 35, 89-110. Broxton, D.E., Warren, D.C., Hagan, R. C., and Luedemann, G. (1986) Chemistry of diagenetically-altered tuffs at a potential nuclear waste r epository, Yucca Mountain, Nevada, 160 p. Los Alamos National Laboratory report LA-10802-MS. Los Alamos, New Mexico. Buatier, M.D., Monnin, C., Fruh-Green, G.L. and Karpoff, A.M. (20 01) Fluid-sediment interactions related to hydrothermal circul ation in the Eastern Flank of the Juan de Fuca Ridge. Chemical Geology, 175, 343-360. 150

PAGE 151

Byers, F.M., Jr., Cart, W. J., Orkild, P. P., Quinlivan, W.D., and Sargent, K.A. (1976) Volcanic suites and related cauldrons of Timber Mountain-Oasis Valley caldera complex, southern Nevada, 70 p. U.S. Geological Survey Professional Paper 919. Cammenga, H.K., Eysel, W., Gmelin, E., Hemminger, W., Hohne, G.W.H., and Sarge, S.M. (1993) The temperatur e calibration of scanning ca lorimeters. 2. Calibration substances. Thermochimica Acta, 219, 333-342. Carey, J.W. (1993) The heat capacity of hydrous cordierite above 295 K. Physics and Chemistry of Minerals, 19, 578-583. Carey, J.W. and Bish, D.L. (1996) Equilibrium in the clinoptilolite-H2O system. American Mineralogist, 81, 952-962. Currie, K.L. and Curtis, L.W. (1976) Applic ation of multicomponent solution theory to jadeitic pyroxenes. Journal of Geology, 84, 179-194. Donahoe, R.J., Hemingway, B.S., and Liou, J.G. (1990) Thermochemical data for merlinoite: 1. Low-temperature heat c apacities, entropies and enthalpies of formation at 298.15 K of 6 synthetic samples having various Si/Al and Na/(Na+K) ratios. American Miner alogist, 75, 188-200. Dyer, A. and Zubair, M. (1998) Ion exchange in chabazite. Microporous and Mesoporous Materials, 22, 135-150. Elprince, A.M., Vanselow, A.P., and Sposit o, G. (1980) Heterovalent, ternary cationexchange equilibria: NH4 +-Ba2+-La3+ exchange on montmorillonite. Soil Science Society of America Journal, 44, 964-969. Evans, J.P. and Chester, F.M. (1995) Fluid-rock interaction in faults of the San Andreas system: Inferences from San Gabriel faul t rock geochemistry and microstructures. Journal of Geophysical Research, 100, 13007-13020. Fialips, C.I., Carey, J.W., and Bish, D. L. (2005) Hydrationdehydration behavior and thermodynamics of chabazite. Geochimica et Cosmochimica Acta, 69, 2293-2308. Fletcher, P., Franklin, K.R., and Townsend, R.P. (1984) Thermodynamics of binary and ternary ion exchange in zeolites: T he exchange of sodium, ammonium and potassium ions in mordenite. Philosophica l Transactions of the Royal Society of London Series A, Mathematical a nd Physical Sciences, 312, 141-178. Fridriksson, T., Carey, J.W., Bish, D.L., N euhoff, P.S., and Bird, D.K. (2003) Hydrogenbonded water in laumontite II: Experiment al determination of site-specific thermodynamic properties of hydrati on of the W1 and W5 sites. American Mineralogist, 88, 1060-1072. 151

PAGE 152

Fridriksson, T., Neuhoff, P.S., Viani, B. E., and Bird, D.K. (2004) Experimental determination of thermodynamic properties of ion-exchange in heulandite: Binary ion-exchange experiments at 55 and 85 C involving Ca2+, Sr2+, Na+, and K+. American Journal of Science, 304, 287-332. Galindo Jr.C., Ming, D.W., Morgan, A. and Pickering, K. (2000) Use of Ca-saturated clinoptilolite for ammonium from NASAs advanced life support wastewater system. In C. Colella and F.A. Mumpton, Eds., Natural Zeolites fo r the Third Millennium, p. 363-371, De Frede Editore, Naples, Italy. Galli, E., and Ghittoni, G.L. (1972) The cr ystal chemistry of phillipsites. American Mineralogist, 57, 1125-1145. Ghiara, M.R., Petti, C., Franco, E., Lonis, R., Luxoro, S., and Gnazzo, L. (1999) Occurrence of clinoptilolite and mordenite in Tertiary calc-alkaline pyroclastites from Sardinia (Italy). Cla ys and Clay Minerals, 47, 319-328. Gmelin, E. and Sarge, S.M. (2000) Temperature, heat and heat flow rate calibration of differential scanning calorimeters Thermochimica Acta, 347, 9-13. Gopal, R., Hollebone, B.R., Langf ord, C.H., and Shigeishi, R. A. (1982) The rates of solar-energy storage and retrieval in a zeolite-water system. Solar Energy, 28, 421-424. Gottardi, G. (1989) The genesis of zeolites. European Journal of Mineralogy, 1, 479-487. Gottardi, G. and Galli, E. (1985) Natural Zeolites, 409 p. Springer-Verlag, Berlin. Greenspan, L. (1977) Humidity fixed poin ts of binary saturated aqueous solutions. Journal of Research of the Nati onal Bureau of Standards-A. Physics and Chemistry, 81A, 89-96. Grigorieva, L.V., Salata, O.V., Kolesn ikov, V.G., and Malakhova, L.A. (1988) Effectiveness of the sorptive and coagulatio nal removal of enteric bacteria and viruses from water. Khim Teknol Vody, 10, 458-461. Grover, J. and Malek, E. (1977) Ranks of whitehead groups Journal of Engineering Sciences, 3, 15-26. Hardy, H.K. (1953) A sub-regular solution m odel and its application to some binary alloy systems. Acta Metallurgica, 1, 202-209. Hay, R.L. (1986) Geologic occurrence of z eolites and some associated minerals. Pure and Applied Chemis try, 58, 1339-1342. 152

PAGE 153

Hay, R.L. and Moiola, R.J. (1963) Authi genic silicate minerals in Searles Lake, California. Sedimentology, 2, 312-332. Hay, R.L., and Sheppard, R.A. (2001) Occurrence of zeolites in sedimentary rocks. In D.L. Bish and D.W. Ming, Eds., Natu ral Zeolites: Occurrences, Properties, Applications, 45, p. 217-234. Reviews in Mineralogy and Geochemistry, Mineralogical Society of Am erica, Chantilly, Virginia. Helfferich, F. (1962) Ion exchange, 624 p. McGraw-Hill, New York, New York. Helfferich, F.G. and Hwang, Y.L. (1991) Ion exchange kinetics. In K. Dorfner, Ed, Ion Exchangers, p. 1277-1309. Walt er de Gruyter, Berlin. Helgeson, H.C., Delany, J.M., Nesbitt, H.W., and Bird, D.K. (1978) Summary and critique of the thermodynam ic properties of rock-forming minerals. American Journal of Science, 278, 1-229. Henley, R.W., and Ellis, A. J. (1983) Geothermal systems ancient and modern: A geochemical review. Earth Science Reviews, 19, 1-50. Hohne, G.W.H., Cammenga, H.K., Eysel, W., Gmelin, E., and Hemminger, W. (1990) The temperature calibration of scanning ca lorimeters. Thermochimica Acta, 160, 112. Hovis, G.L. (1982) Resolution of a systematic interlaboratory discrepancy in recent calorimetric data, and the heats of solu tion of quartz, low albite, adularia, and gibbsite. American Mi neralogist, 67, 950-955. (1988) Enthalpies and volumes related to K-Na mixing and Al-Si order-disorder in alkali feldspars. Journal of Petrology, 29, 731-763. Hovis, G.L., Harlov, D., and Gottschalk, M. (2004) Solution calorimetric determination of the enthalpies of formation of NH4-bearing minerals buddingtonite and tobelite. American Mineralogist, 89, 85-93. Hovis, G.L. and Roux, J. (1993) Thermodyna mic mixing properties of nepheline-kalsilite crystalline solutions. American Journal of Science, 293, 1108-1127. Hovis, G.L., Roux, J., and Rich et, P. (1998) A new era in hydrofluoric acid solution calorimetry: Reduction of required samp le size below ten milligrams. American Mineralogist, 83, 931-934. Hovis, G.L., Roux, J., and R odrigues, E. (2002) Thermodynamic and structural behavior of analcime-leucite anal ogue systems. American Mi neralogist, 87, 523-532. 153

PAGE 154

Iijima, A. (1988) Diagenetic transformations of minerals as exemplified by zeolites and silica minerals-A Japanese view. In G.V. Chilingarian and K.H. Wolf, Eds., Diagenesis II, 43, p. 147-211. Elsevier. Amsterdam. James, E.W., and Silver, L.T. (1988) Implications of zeolites and their zonation in the Cajon Pass Deep Drillhole. Geophysi cal Research Letters, 15, 973-976. Johnson, G.K., Flotow, H.E ., Ohare, P.A.G., and Wise, W.S. (1983) Thermodynamic studies of zeolites: Natrolite, mesolite and scolecite. American Mineralogist, 68, 1134-1145. Johnson, G.K., Tasker, I.R., Flotow, H.E ., Ohare, P.A.G., and Wise, W.S. (1992) Thermodynamic studies of mordenite dehydrated mordenite, and gibbsite. American Mineralogist, 77, 85-93. Johnson, J.W., Oelkers, E.H., and Helges on, H.C. (1992) SUPC RT92: A software package for calculating the standard molal thermodynamic properti es of minerals, gases, aqueous species, and reactions from 1 to 5000 bars and 0 to 1000 C. Computers & Geosciences, 18, 899-947. Kallo, D. (2001) Applications of natural zeolites in wate r and wastewater treatment. In D.L. Bish and D.W. Ming, Eds., Natu ral Zeolites: Occurrences, Properties, Applications, 45, p. 519-550. Reviews in Mineralogy and Geochemistry, Mineralogical Society of Am erica, Chantilly, Virginia. Keith, T.E.C., Thompson, J. M., and Mays, R.E. (1983) Selective concentration of cesium in analcime during hydrothermal alteration, Yellowstone National Park, Wyoming. Geochimica et Cosmochimica Acta, 47, 795-804. Kerrisk, J.F. (1987) Gr oundwater chemistry at Yucca Mo untain, Nevada, and vicinity. Los Alamos National Laborator y Report LA-9912-MS, 41 p. Kitsopoulos, K.P. (1997) The genesis of a mordenite deposit by hydrothermal alteration of pyroclastics on Polyegos Island, Greece. Clays and Clay Minerals, 45, 632-648. Kitsopoulos, K.P. and Dunham, A. C. (1998) Compositional variations of mordenite from Polyegos Island, Greece: Na-Ca and K -rich mordenite. European Journal of Mineralogy, 10, 569-577. Long, J.C.S. and Ewing, R.C. (2004) Yucca Mountain: Ea rth science issues at a geologic repository for high-level nuclear waste. Annual Review of Earth and Planetary Sciences, 32, 363-401. Loewenstein, W. (1954) The distribution of aluminum in the tetrahedra of silicates and aluminates. American Mi neralogist, 39, 92-96. 154

PAGE 155

Maier, C.G. and Kelley, K. K. (1932) An equation for t he representation of hightemperature heat content data. Journal of the Americ an Chemical Society, 54, 3243-3246. Martucci, A., Sacerdoti, M., Cruciani, G., and Dalconi, C. (2003) In situ time resolved synchrotron powder diffraction study of mordenite. European Journal of Mineralogy, 15, 485-493. McCulloh T.H., Cashman S.H., and Stewar t R.J. (1973) Diagnetic baselines for interpretive reconstructions of maxi mum burial depths and paleotemperatures in clastic sedimentary rocks. In A Sympos ium in Geochemistry: Low Temperature Metamorphism of Kerogen and Clay Minerals, p. 18-46. Pacific Section of SEPM, Los Angeles, California. McCulloh, T.H., and Stewart, R.J. (1980) S ubsurface interface between zeolitized and overlying less-altered rocks, Souther n San Joaquin Valley, California: Configuration and implications for petrol eum entrapment. Americ an Association of Petroleum Geologist Bulletin, 64, 446. McHenry, L.J. (2009) El ement mobility during zeolitic and argillic alteration of volcanic ash in a closed-basin lacustrine environmen t: Case study Olduvai Gorge, Tanzania. Chemical Geology, 265, 540-552. Meier, W.M., Meier, R., and Gramlich, V. (1978) Mordenite: Interpretation of a superposed structure. Zeitschrift fr Kristallographie, 147, 329. Ming, D.W. and Dixon, J.B. ( 1987) Quantitative determination of clinoptilolite in soils by a cation-exchange capacity method. Clays and Clay Miner als, 35, 463-468. Ming, D.W., and Mumpton, F.A. (1989) Zeolites in soils. In J.B. Dixon and S.B. Weed, Eds., Minerals in Soil Environments, 1, p. 873-911. Soil Science Society of America, Madison, Wisconsin. Mortier, W. J. (1982) Compilat ion of extra framework sites in zeolites, 67 p. Butterworths, London. Mortier, W.J., Pluth, J.J., and Smith, J.V. (1978) Positions of cations and molecules in zeolites with the mordenite-type fram ework. IV. Dehydr ated and rehydrated Kexchanged ptilolite. In L.B. Sand and F. Mumpton Eds., Natural Zeolites, Occurrence, Properites, Use, p. 53-75. Pergamon Press, Oxford. Muhe, R. and Stoffers, P. (1995) Rock structures and mineral assemblage associated with hydroexpansion in midoceanic ri dge basalts. Mineralogy and Petrology, 54, 71-83. 155

PAGE 156

Mumpton, F.A. (1977) Utilization of natural zeolites. In F.A. Mumpton Ed., Mineralogy and Geology of Natural Zeolites, p. 177-204. Reviews in Mineralogy 4, Mineralogical Society of Am erica, Chantilly, Virginia. Murphy, C.B., Hrycyk, O., and Gleason, W.T. (1978) Natural zeolite: Novel uses and regeneration in wastewater treatment. In L.B. Sand an d F.A. Mumpton, Eds., Natural Zeolites: Occurrence, Properties Use, p. 5 471-478. Pergamon Press, Oxford. Nativel, P. (1986) Occurrence of mordenite at Cilaos, R eunion Island, Indian Ocean. Bulletin De Miner alogie, 109, 337-347. Navrotsky, A. and Tian, Z.R. (2001) Systematics in the enthalpies of formation of anhydrous aluminosilicate zeolites, glasses, and dense phases Chemistry, a European Journal, 7, 769-774. Neuhoff, P.S. (2000) Thermodynamic pr operties and parageneses of rock-forming zeolites. PhD thesis, Stanford University, Stanford, CA. Neuhoff, P.S. and Bird, D.K. ( 2001) Partial dehydration of laumontite: thermodynamic constraints and petrogenetic implications Mineralogical M agazine, 65, 59-70. Neuhoff, P.S., Fridriksson, T., Arnorsson, S., and Bird, D.K. (1999) Porosity evolution and mineral paragenesis during low-grade metamorphism of basaltic lavas at Teigarhorn, eastern Iceland. Americ an Journal of Science, 299, 467-501. Neuhoff, P.S., Fridriksson, T., and Bird, D.K. (2000) Zeolite parageneses in the north Atlantic igneous province: Implications for geotectonics and groundwater quality of basaltic crust. International Geology Review, 42, 15-44. Neuhoff, P.S., Hovis, G.L., Ba lassone, G., and Stebbins, J.F. (2004) Thermodynamic properties of analcime solid solutions. Am erican Journal of Science, 304, 21-66. Neuhoff, P.S., and Ruhl, L.S. (2006) Mechanisms and geochemical significance of Si-Al substitution in zeolite solid solu tions. Chemical Geology, 225, 373-387. Neuhoff, P.S. and Wang, J. (2007a) Heat capacity of hydration in zeolites. American Mineralogist, 92, 1358-1367. (2007b) Isothermal measur ement of heats of hydrat ion in zeolites by simultaneous thermogravimetry and diffe rential scanning calorimetry. Clays and Clay Minerals, 55, 239-252. Neuhoff, P.S., Watt, W.S., Bird, D.K., and P edersen, A.K. (1997) Timing and structural relations of regional zeolite zones in bas alts of the East Greenland continental margin. Geology, 25, 803-806. 156

PAGE 157

Newell, P.A. and Rees, L.V.C. (1983) Ion-exchange and cation site locations in zeolite L. Zeolites, 3, 22-27. Nitkiewicz, A.M., Kerrick, D.M. and Hemingway, B.S. (1983) T he effect of particle size on the enthalpy of solution of quartz: Implications for phase equilibria and solution calorimetry. Geological Society of Amer ica Abstracts with Programs, 15, 653. Ogihara, S. (1996) Diagenetic trans formation of clinoptilolite to analcime in silicic tuffs of Hokkaido, Japan. Mineraliu m Deposita, 31, 548-553. (2000) Composition of cl inoptilolite formed from vo lcanic glass during burial diagenesis. Clays and Clay Minerals, 48, 106-110. Ogihara, S. and Ijima, A. (1989) Clinoptilolite to heuland ite transformation in burial diagenesis. In P.A. Jacobs, Ed., Zeolites : Facts, Figures, Future: Proceedings of the 8th International Zeolite Conference, 49, p. 491. Amsterdam, the Netherlands. Pabalan, R.T. (1994) Ther modynamics of ion exchange between clinoptilolite and aqueousbsolutions of Na+/K+ and Na+/ Ca2+. Geochimica et Cosmochimica Acta, 58, 4573-4590. Pabalan, R.T. and Bertetti E.P. (1999) Experimental and modeling study of ion exchange between aqueous solutions and the z eolite mineral clin optilolite. Journal of Solution Chemistry, 28, 367-393. Pabalan, R.T., and Bertetti, F.P. (2001) Cation-exchange properties of natural zeolites. In D.L. Bish and D.W. Ming, Eds., Na tural Zeolites: Occurrences, Properties, Applications, 45, p. 453-518. Reviews in Mineralogy and Geochemistry, Mineralogical Society of Am erica, Chantilly, Virginia. Passaglia, E. (1975) Crystal chemistry of mordenites. Contributions to Mineralogy and Petrology, 50, 65-77. Passaglia, E., and Sheppard, R.A. (2001) The crystal chemistry of zeolites. In D.L. Bish and D.W. Ming, Eds., Natural Zeolites: Occurrences, Properties, Applications, 45, p. 69-116. Reviews in Mineralogy and Geoc hemistry, Mineralog ical Society of America, Chantilly, Virginia. Pe-Piper, G. (2000) Mode of occurrence, chemical variat ion and genesis of mordenite and associated zeolites from the Morden area, Nova Scotia, Canada. Canadian Mineralogist, 38, 1215-1232. Rao, A., and Rees, L.V.C. (1966) Kinetics of ion exchange in mordenite. Transactions of the Faraday Society, 62, 2505-2512. 157

PAGE 158

Robie, R.A., and Hemingway, B.S. (1991) Heat capacities of kaolin ite from 7 to 380 K and of DMSO-intercalated kaolinite from 20 to 310 K. The entropy of kaolinite Al2Si2O5(OH)4. Clays and Clay Minerals, 39, 362-368. Robie, R.A., Hemingway, B.S., and Wils on, W.H. (1978) Low -temperature heat capacities and entropies of feldspar glasses and of anorthite. American Mineralogist, 63, 109-123. Robinson, G.R. and Haas, J.L. (1983) Heat c apacity, relative enthalpy, and calorimetric entropy of silicate mineral s: An empirical method of prediction. American Mineralogist, 68, 541-553. Ross, M., Flohr, M.J.K., and Ross, D.R. ( 1992) Crystalline solution series and orderdisorder within the natrolite mineral group. American Miner alogist, 77, 685-703. Sabbah, R., An, X.W., Chickos, J.S., Leitao, M.L.P., Roux, M.V., and Torres, L.A. (1999) Reference materials for calorimetr y and differential thermal analysis. Thermochimica Acta, 331, 93-204. Sarge, S.M., Gmelin, E., Hohne, G.W.H., Cammenga, H.K ., Hemminger, W., and Eysel, W. (1994) The caloric calibration of scanni ng calorimeters. Thermochimica Acta, 247, 129-168. Scarmozzino, R., Aiello, R., and Santucci, A. (1980) Chabazitic tuff for thermal storage. Solar Energy, 24, 415-416. Schlenker, J.L., Pluth, J.J., and Smith, J.V. (1979) Positions of cations and molecules in zeolites with the mordenite-type fram ework. 8. Dehydrat ed sodium-exchanged mordenite. Materials Resear ch Bulletin, 14, 751-758. Selvidge, M., and Miaoulis, I.N. (1990) Evaluation of reversible hydration reactions for use in thermal energy storage. Solar Energy, 44, 173-178. Sheppard, R.A. and Hay, R.L. (2001) Formation of zeolit es in open hydrologic systems. In D.L. Bish and D.W. Ming, Eds., Na tural Zeolites: Occurrences, Properties, Applications, 45, p. 261-275. Reviews in Mineralogy and Geochemistry, Mineralogical Society of Am erica, Chantilly, Virginia. Shibue, Y. (1998) Cation-exch ange properties of phillipsite (a zeolite mineral): The differences between Si-rich and Si-poor phillipsites. Separation Science and Technology, 33, 333-355. (1999) Calculations of flui d-ternary solid solution equili bria: An application of the Wilson equation to fl uid-(Fe, Mn, Mg)TiO3 equilibria at 600 C and 1 kbar. American Mineralogist, 84, 1375-1383. 158

PAGE 159

Shigeishi, R.A., Langford, C. H., and Hollebone, B.R. (1979) Solar energy storage using chemical potential changes associated with drying of zeolites. Solar Energy, 23, 489-495. Shim, S.H., Navrotsky, A., Gaffney, T.R., and MacDougall, J.E. (1999) Chabazite: Energetics of hydration, ent halpy of formation, and effe ct of cations on stability. American Mineralogist, 84, 1870-1882. Simoncic, P., Armbruster, T., and Pattison, P. (2004) Cationic thionin blue in the channels of zeolite mordenite: A single-cr ystal X-ray study. Journal of Physical Chemistry B, 108, 17352-17360. Smyth, J.R. (1982) Zeo lite stability constraints on radioa ctive waste isolation in zeolitebearing volcanic rocks. Journal of Geology, 90, 195-201. Spooner, E.T.C., Chapman, H.J., and Smewing, J.D. (1977) Strontium isotopic contamination and oxidati on during ocean-floor hydrot hermal metamorphism of ophiolitic rocks of t he Troodos Massif, Cyprus. Geochimi ca et Cosmochimica Acta, 41, 873-890. Sposito, G. (1981) Cation exchange in soils: An historical and theoretical perspective. In M. Stelly, Ed., Chemistry in the Soil Environment, 40, p. 13-30. Soil Science Society of America, Madison, Wisconsin. Stolen, S., Glockner, R., and Gronvold, F. (1996) Heat capacity of the reference material synthetic sapphire (alpha-Al2O3) at temperatures from 298.15 to 1000 K by adiabatic calorimetry. Increased a ccuracy and precision through improved instrumentation and computer control. Journal of Chemical Thermodynamics, 28, 1263-1281. Sturchio, N.C., Bohlke, J.K., and Binz, C. M. (1989) Radium-thorium disequilibrium and zeolite-water ion exchange in a Ye llowstone hydrother mal environment. Geochimica et Cosmochimica Acta, 53, 1025-1034. Sukheswa.Rn, Avasia, R.K., and Gangopad.M. (1974) Zeolites and associated secondary minerals in the Deccan Traps of Western India. Miner alogical Magazine, 39, 658-671. Suzuki, N., Saitoh, K., and Hamada, S. (197 8) Ion-exchange properties of a synthetic mordenite on alkali and alkaline-ea rth metal ions. Radiochemical and Radioanalytical Letters, 32, 121-126. Tarasevich, Yu.I., Krysenko, D.A., and Polyakov V.E. (2006) Equilibr ia and heats of ion exchange in the system of mordenite: Alka li and alkaline earth ca tions. Theoretical and Experimental Chem istry, 42, 320-326. 159

PAGE 160

Tomasson, J. and Kristmannsdot tir, H. (1972) High temper ature alteration minerals and thermal brines, Reykjanes, Iceland. Contributions to Mineralogy and Petrology, 36,123-134. Townsend, R.P. and Loizidou, M. (1984) Ion-exchange properties of natural clinoptilolite, ferrierite and mordenite .1. Sodium a mmonium equilibria. Zeolites, 4, 191-195. Tsitsishvili, G.V., Andronikashvili, T.G., Ki rov, G.N., and Filizova, L.D. (1992) Natural zeolites, 295 p. Ellis Horwood, Chichester. Vaniman, D.T., Chipera, S.J., Bish, D. L., Carey, J.W., and Levy, S.S. (2001) Quantification of unsaturated-zone alteration and cation exchange in zeolitized tuffs at Yucca Mountain, Nevada, USA. Geochimica et Cosmochimica Acta, 65, 3409-3433. van Reeuwijk, L.P. (1974) The thermal dehydration of natural zeolites, 88 p. Dissertation, Wageni ngen, Netherlands. Vaughan, D.E.W. (1978) Properties of natural zeolites. In L.B. Sand and F.A. Mumpton Eds., Natural Zeolites: Occurance, Proper ties, Use, p. 353-372. Pergamon Press, Oxford. Vezzalini, G., Quartieri, S., Rossi, A., Alber ti, A., and Ricci, C.A. (1994) Occurrence of zeolites from northern Vict oria Land (Antarctica); Proc eedings of the 4th meeting on Earth sciences in Antarctica. Terra Antartica, 1, 96-99. Waldbaum, D.R. and Robie, R.A. (1970) Internal sample container for hydrofluoric acid solution calorimetry. Journal of Geology, 78, 736-741. Wang, C.Y. (1984) On the constitution of the San Andreas fault zone in central California. Journal of Geophysical Research, 89, 5858-5866. Wang, C.Y., Feng, R., Yao, Z.S., and Shi, X.J. (1986) Gravity anomaly and density structure of the San Andreas fault z one. Pure and Applied Geophysics, 124, 127140. Wang, J. and Neuhoff, P.S. (2008a) Anomalous temperature dependence of the heat of hydration in natrolite. The O pen Mineralogy Journal, 2, 43-48 (2008b) Thermodynamics of dehydration in analcime: Absorption calorimetry and equilibrium observations. Amer ican Mineralogist, 93, 1449-1459. White, A.F., Claasen, H.C., and Benson, L.V. (1980) The effect of glass dissolution on the water chemistry in a tuffaceous aquifer, Ranier Mesa, Nevada. U.S. Geological Survey Water Supply Paper, 1535-1565. 160

PAGE 161

161 Wilkin, R.T. and Barnes, H.L. (1998) Solub ility and stability of zeolites in aqueous solution: I. Analcime, Na-, and K-clinopt ilolite. American Miner alogist, 83, 746-761. Wise, W.S. and Tschernich, R.W. (1978) Daciardite-bearing zeolite assemblages in the pacific northwest. In L.B. Sand and F.A. Mumpton, Eds., Natural Zeolites: Occurrence, Properties, Use, p. 105-111. Pergamon Press, Oxford. Wolery, T.J. (1991) EQ3nr. A computer program for geochemical aqueous speciation solubility calculations: Theoretical Manual, UserS Guide and Related Documentation (Version 7.0). Lawr ence Livermore National Laboratory. Wolf, F., Georgi, K., and Pilc howski, K. (1978) Ion-exc hange of mono-valent cations on synthetic mordenite.1. Equilibrium and thermodynamics of exchange K+-Na+, H+Na+ And H+-K+. Zeitschrift Fur Physikalische Chemie (Leipzig), 259, 717-726. Yang, P., Stolz, J., Armbruster, T., and Gunter, M.E. (1997) Na, K, Rb, and Cs exchange in heulandite single crystals: Di ffusion kinetics. American Mineralogist, 82, 517-525. Yang, S.Y. and Navrotsky, A. (2000) Energetics of form ation and hydration of ionexchanged zeolite Y. Microporous and Mesoporous Materials, 37, 175-186. Yang, S.Y., Navrotsky, A., and Wilkin, R. (2001) Thermodynamics of ion-exchanged and natural clinoptilolite. Amer ican Mineralogist, 86, 438-447. Zamzow, M.J., Eichbaum, B.R., Sandgren, K. R., and Shanks, D.E. (1990) Removal of heavy metals and other cations from wa ste water using zeolites. Separation Science and Technology, 25, 1555-1569.

PAGE 162

BIOGRAPHICAL SKETCH Jie Wang was born in January 1981 in a small town of Ezhou, China. He lived with his parents and attended elementary and high schools in the town. He entered the University of Science & Technology of China in September 1999 with excellent grades of College Entrance Exams. After five year s study in Geoscience, he graduated with a bachelor degree of science. He was admitted to the PHD program in the Department of Geological Sciences, University of Florida, in August 2004 with a research assistantship and the Grinter Fellowship. He worked with Dr. Philip Neuhoff and focused his research on mineralogy and physical geochemistry. He spent two years on the project Thermodynamics of dehydrati on and hydration of zeolites and obtained his master degree of science in Geological Sciences in August 2006. After that he stayed in University of Florida to pursue his P HD degree, and worked on the ion exchange and water molecule behavior in zeolites. Upon completion of PHD study, he will find a full time position as a geologist to unitilize his research experience and skills in industry. 162