r'.‘l I I\?\|I"'|Il|l'" ‘ 489's “Es. OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. METAL ION COMPLEXATION STUDIES OF A SULFUR-OXYGEN MIXED DONOR MACROCYCLIC LIGAND BY SOLVENT EXTRACTION AND NUCLEAR MAGNETIC RESONANCE BY Leslie Ann Greenbauer A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1978 ABSTRACT METAL ION COMPLEXATION STUDIES OF A SULFUR-OXYGEN MIXED DONOR MACROCYCLIC LIGAND BY SOLVENT EXTRACTION AND NUCLEAR MAGNETIC RESONANCE BY Leslie Ann Greenbauer Measurements of metal ion complexing strengths with three macrocyclic ligands are reported. Two tech- niques, solvent extraction and nuclear magnetic resonance, were employed to assess the cation-ligand binding strengths. The extraction equilibrium constants (log Ke) for potassium with dibenzo-18C6 were determined to be 4.79 i 0.22 and 6.38 i 0.06 with 18C6. Constants were also obtained for cesium and thallium (I), where log Ke for the (Cso18C6)+ complex was determined to be 5.56 i 0.41 and for complexes of thallium (I), with 18C6 and 1,10-dithia- 18C6, were found to be 5.81 i 0.05 and 3.80 i 0.01, respectively. Cesium-133 and thallium-205 NMR chemical shift measurements provided a second method for determining complex formation constants. Cesium-1,10-dithia-18C6 complex formation constants (log Kf) were determined in Leslie Ann Greenbauer four solvents: nitromethane (0.65 i 1.09), dimethylforma- mide (0.56 i 0.29), acetonitrile (0.90 i 0.27), and pyri- dine (0.84 H- 0.15). The thallium ion binding with 1,10- dithia-18C6, in pyridine, produced a stronger complex 1.39 i 0.19. where log Kf A Raman-IR study of the complex formed between the solvent, acetonitrile, and the 18C6 ligand, was per- formed to identify bands arising due to this interaction. To my family Mom Dad Julie Janet Dan Cindy and Barney and my family to be Gary ii ACKNOWLEDGMENTS Although a thesis is an account of an individual's research accomplishments, the final product is actually a summation of the efforts of several people. I would like to thank Dr. Alexander Popov for serving as my research preceptor and for his recommendations toward completing my masters project. I extend my thanks to Dr. Harvey Nikkel, from Grand Valley State Colleges, for introducing me to research and encouraging me to continue my studies in graduate school. A very special thanks to my coworkers Fred Smetana, whose professional help and friendship made the time spent a little more worthwhile, and John Hoogerheide, for his help in writing the computer program used in the extrac- tion studies. I would like to thank Davette Whitaker for making 405 more fun to do work in. My deep thanks to Gary, who provided me with his invaluable gifts of love, support, and understanding which he unselfishly provided when he was trying to com- plete his own work. Forever my gratitude and love to my parents and family for instilling in me the awareness of iii a sound education and the desire to obtain it and to the youngest member, Barney, who quietly listened and helped in his own special way. To all of you, thank you. iv TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . vii LIST OF FIGURES . . . . . . . . . . . . ix I. INTRODUCTION . . . . . . . . . . . 1 II. HISTORICAL REVIEW . . . . . . . . . 8 Influences upon Macrocyclic Polyether- Alkali Metal Ion Complexing Stability. . . 8 Polythiaethers and Sulfur-Oxygen Mixed Donor Ligands . . . . . . 15 Metal Ion—Crown Complexation Studies by the Extraction Technique. . . . . . 21 Raman Studies of Solvation and Ion Association . . . . . . . . . . 30 Nuclear Magnetic Resonance. . . . . . 32 Cesium-133 NMR . . . . . . . . . 34 Thallium-ZOS NMR 0 o o o o a o o 37 III. EXPERIMENTAL MATERIALS, INSTRUMENTS, AND PROCEDURES 0 O O O O O O O I O O 44 Materials 0 O O O O O O O O O O 44 Reagents . . . . . . . . . . . 44 Ligands . . . . . . . . . . . 45 Solvents . . . . . . . . . . . 51 Instruments and Procedures. . . . . . 53 Extraction Studies. . . . . . . . 53 Infrared Absorption and Laser Raman . . 55 NMR Specrometer. . . . . . . . . 57 IV. SOLVENT EXTRACTION TECHNIQUE FOR THE DETER- MINATION OF MACROCYCLIC LIGAND-METAL EQUILIBRIUM CONSTANTS . . . . . . . 61 Introduction . . . . . . . . . . 61 Distribution Equilibria. . . . . . . 62 V Page Characterization of Picric Acid Ultra— violet Absorption . . . . 65 Possible Sources of Error in Extraction Procedures . . . . . . 67 Ligand Partition Coefficient Studies . . 71 Dibenzo-18C6: Spectrophotometric Determination. . . . . . . 71 18C6 and 1,10-Dithia-18C6: Gravimetric Determination. . . . . . . . . 71 Extraction Systems . . . . . . . . 72 Extraction Studies Utilizing the DB18C6 Ligand . . . . 73 Extraction Studies Utilizing the 18C6 Ligand . . . . 77 Extraction Studies Utilizing the 1,10- Dithia-18C6 Ligand . . . . . . . 92 Extraction Equilibrium Constant Determi— nation. . . . . . . . . . . . 92 V. INFRARED AND RAMAN STUDIES OF 18C6'ACETO- NITRILE COMPLEX . . . . . . . . . 110 Introduction . . . . . . . 110 Infrared and Raman Spectral Data. . . . 111 VI. DETERMINATION OF CESIUM- AND THALLIUM- DT18C6 COMPLEX STABILITY CONSTANTS IN NONAQUEOUS SOLVENTS BY NMR. . . . . 115 Introduction . . . . 115 Concentration Formation Constants from NMR Chemical Shift Measurements . . . 116 Cesium-133 NMR Study. . . . . . . . 119 Thallium-205 NMR Study . . . . . . . 126 VII. COMMENTARY. . . . . . . . . . . . 132 APPENDICES A. EXTRACTION EQUILIBRIUM CONSTANT DETERMI- NATION FROM DISTRIBUTION EQUILIBRIA STUDIES . . . . . . . . . . . . 134 B. COMPUTER PROGRAM EXTRACT . . . . . . . 137 LIST OF REFERENCES . . . . . . . . . . . 139 vi 10. 11. 12. LIST OF TABLES Page NMR Properties of the 133Cs and 205T1 Nuclei . . . . . . . . . . . . 35 Physical Properties of Thallium (I) and the Alkali Metal Cations . . . . . . 38 The Chemical Shifts, Both Observed and Literature Values, Obtained for DT18C6/ Deuterated Chloroform Proton NMR . . . 50 General Instrumental Parameters Used for the Collection of Raman Spectra. . . . 56 Some Solvent Properties and Bulk Suscepti- bility Corrections Used in This Work . . 59 Stability Constants (K5) in Water for the Meta1-Ligand Complexes. . . . . . 72 Equilibrium Extraction Study of the System CHZCI2 + D818C6 / KOH + PicH + H20 . . 74 Equilibrium Extraction Study of the System CH2C12 + 18C6 / KOH + PicH + H20 . . . 79 Equilibrium Extraction Study of the System CH2C12 + 18C6 / CsOH + PicH + H20 . . . 82 Equilibrium Extraction Study of the System CH2C12 + 18C6 / TIC H O + LiOH + PicH 2 3 2 + H O O O O O O O O O O O 84 2 Equilibrium Extraction Study of the System CH2C12 + DT18C6 / KC H O + PicH + LiOH 2 3 2 + H20 0 O O O O O O O 93 Equilibrium Extraction Study of the System CH2C12 + DT18C6 / CsC2H3O2 + PicH + LiOH + H20 . . . . . . . . . . 96 vii Table Page 13. Equilibrium Extraction Study of the System CH2C12 + DT18C6 / T1C H o + PicH + LiOH 3 2 + H O O O I I O O O O O O O O O 97 2 l4. Extraction Equilibrium and Complex Dissoci- ation Constants. . . . . . . . . . 105 15. Equilibrium Constants for the Extraction of Picrates . . . . . . . . . . . . 108 16. Infrared and Raman Band Frequencies of 18C6 and ACN Samples (Regions: IR 600-4000 cm”1 and Raman 840-3010 cm’l) . . . . . 112 17. Cesium-133 Chemical Shifts of (Cs-DT18C6)+ Complexes in Various Solvents (ambient probe temperature 32-33°C) . . . . . . 121 18. Complex Formation Constants for Cesium Com- plexes with DT18C6 in Four Nonaqueous Solvents . . . . . . . . . . . . 125 19. Thallium-205 Chemical Shifts for (Tl-DT18C6)+ Complexes in Two Nonaqueous Solvents. . . 127 viii LIST OF FIGURES Figure 1. Representative synthetic macrocyclic ligands . . . . . . . . . . . . . 2. Proton NMR of DT18C6 in deuterated chloroform with tetramethylsilane internal reference (SSB = spinning side band). . . . . . 3. Beers law plot for picric acid in water, A = 356 nm . . . . . . . . . . . . 4. Extraction equilibrium determination of the concentration of picrate ion extracted into methylene chloride as a function of time 0 O O I O O O O O O O O O O 5. Extraction of potassium picrate into methylene chloride by D818C6 at various cation and ligand concentrations . . . . . . . . 6. Extraction of potassium picrate into methylene chloride by 18C6 at various cation and ligand concentrations . . . . . . . . 7. Extraction of cesium picrate into methylene chloride by 18C6 at various cation and ligand concentrations . . . . . . . . 8. Extraction of thallium picrate into methylene chloride by 18C6 at various cation and ligand concentrations . . . . . . . . 9. Comparative potassium picrate extraction efficiency into methylene chloride and toluene . . . . . . . . . . . . . 10. Comparative extraction of thallium picrate using two different counterions . . . . . ix Page 49 66 70 76 86 87 88 90 91 Figure Page 11. Extraction of potassium picrate into methylene chloride by DT18C6 at various cation and ligand concentrations . . . . . . . . 100 12. Extraction of cesium picrate into methylene chloride by DT18C6 at various cation and ligand concentrations . . . . . . . . 101 13. Extraction of thallium picrate into methylene chloride by DT18C6 at cation concentrations of 0.024M and 0.079M with various ligand concentrations . . . . . . . . . . 102 14. High efficiency extractions of thallium picrate into methylene chloride by DT18C6 at various cation and ligand concentra- tions . . . . . . . . . . . . . 103 15. Cesium-133 chemical shifts versus ligand to metal mole ratio for (CS°DT18C6)+ complex formation . . . . . . . . . . . . 122 16. Thallium-205 chemical shifts versus ligand to metal mole ratio for (T1°DT18C6)+ complex formation . . . . . . . . . . . . 129 CHAPTER I INTRODUCTION Since the middle 19605 the synthetic macrocyclic ligands and their complexes have received much attention. Extensive investigation of these polydentate ligands have supplied valuable information concerning these' molecules and their interaction with a variety of metal cations. Several subdivisions of the general class of syn- thetic macrocyclic ligands may be recognized. The divisions may be based upon the number of rings that make up the molecule, the size of the ring, the type of donor atoms which make up the internal surface of the ring (which act as the binding sites for complexation), or by the presence of substituent groups. Another impor- tant variation is the possibility of mixed donor atoms. The broadest classification and commonly used in literature includes: (a) Macrocyclic polyethers. This group includes the monocyclic polyethers, as well as their nitrogen and sulfur derivatives, first reported in 1967 by Pedersen (1). (b) Macrobicyclic ligands. These molecules, referred to as cryptands and first synthesized by Lehn and co-workers (2, 3), consist of a tridimen- sional intramolecular cavity. The central cavity is formed by the three bridges meeting at common nitrogen atoms on each end of the molecule. Figure 1 contains examples of some of the many types of macrocyclic ligands which have been synthesized and studied. Also shown in Figure l are the IUPAC names given to these ligands as well as the more abbreviated "crown nomenclature." Although the IUPAC provides a very descriptive and unique name for each ligand, it is too complicated for convenience. It was for this reason that trivial names were proposed by Pedersen (1). The assign- ment of the trivial names to the single cycle ligand may be represented by the general expression x-crown-y, where x is the total number of atoms in the macrocyclic ring, crown is the class name, and y is the number of oxygen atoms (or other donor atoms) in the ring (Figure 1-A). Variation in the type of donor atoms within the ring and their positions (Figure 1-B, l-C, and 1-D) or any substituents attached to the ring (Figure l-E) are noted initially. Figure l-F is an example of a bicyclic ligand. The bicyclic ligands are called "cryptands." Since the bicyclic ligands may have a variable number of ether oxygens in each strand, TRIVIAL NAME: 18-Crown-6 IUPAC: 1,4,7,10,13,16-hexaoxacyclooctadecane Figure 1-A 1,10-dithia-18-Crown-6 4,7,13,l6-tetraoxa-l,10-dithiacyclooctadecane Figure l-B aza-lS-Crown-S 4,7,10,13-tetraoxa-1-azacyclopentadecane Figure 1-C Fig. l.--Representative synthetic macrocyclic ligands NAN NVN 1,4,8,ll-tetraaza-l4-Crown-4 1,4,8,11-tetraazacyclotetradecane Figure 1-D dibenzo-lB-Crown-G 2,3,11,12—dibenzo-1,4,7,10,13,16-hexaoxacyclooctadeca-2, ll-diene Figure 1-E W Cryptand-222 (C222) 4,7,13,16,21,24-hexaoxa-1,10-diazabicyclic (8,8,8) hexacosane Figure 1-F Fig. l.--Continued it is convenient to use the number and the distribution of ether oxygens for coding the cryptand. Thus, cryptand F of Figure l is designated by C222. The attraction of a great number of researchers to the study of the crown and crown-related compounds was a result of Pedersen's findings concerning the bind- ing ability and selectivity the ligands possess for alkali and alkaline earth metals. The ligand-metal com- plexes formed were stable both in solutions and in the crystalline form. Before Pedersen's work, investigators found alkali salt solutions to be relatively inert toward complexation. Thus began the tremendous and lengthy studies of the complexation properties of the macrocyclic ligands, not only with alkali and alkaline earth cations but also with several transition metal ions. Although a critical survey of the literature would yield a multitude of articles concerning crown and cryptand complexation, a continuation of these studies is certainly not futile or repetitious. Stability con- stants of metal complexes are important. These constants are not only important from the point of view of chemical interest but are valuable in many areas of science. For example, in theoretical chemistry they help establish relationships between the stability and certain properties of the ligand or metal ion; in analytical chemistry they allow evaluation of new methods of measurement by a thorough understanding of the extent of interference from system components and in physical chemistry the constants are of interest in studying solution thermodynamics and reaction kinetics involving metal complexes, etc. These values are also of great importance in the various fields of biology when considering the effects of metal ions. Therefore, complete and careful studies of complexation by a variety of techniques are necessary. The crowns and cryptands do promise applications in different areas of chemistry and biology. The ability of these ligands to selectively bind cations makes pos- sible their use in areas where this selectivity property is important, such as sensing elements in cation selective electrodes (4), in enhancing the solubility of salts in low polarity solvents (5), and in separation processes where they may separate not only different metals (6, 7), but also different isotopes of the same metal (8, 9). They are also excellent models for the naturally occurring antibiotic macrocycles. The cyclic polyethers have not only been shown to be structurally similar to some of the antibiotics but also capable of increasing the transport of alkali metal ions across reconstituted biological membranes (10, 11). A variety of methods are available for use in studying the formation of the complexes between macro— cyclic ligands and metal salts. These include such techniques as calorimetric titration (12, 13), poten- tiometry with a cation selective electrode (4, 14), electrical conductance (15, 16), optical spectroscopy (17, 18), proton and alkali metal nuclear magnetic resonance (NMR) (19, 20), infrared spectrosc0py (21), and distribution equilibria (22). The work presented here deals with complexation studies of several monocyclic crown compounds utilizing two of the above techniques. Complexation of dibenzo—18C6 (DBlSC6) with potassium and, 18C6 and 1,10-dithia-18C6 (DT18C6) with potassium, cesium, and thallium were studied by an extraction technique. The two phase systems were analyzed by the distribution equilibria between the organic and aqueous phases. The intent of this particular study was to compare and contrast the calculated equilibrium extraction constants, due to the cation-crown complexes, of the different metals and ligands employed. Nuclear magnetic resonance studies were also per- formed for the mixed donor macrocyclic compound DT18C6 in various nonaqueous solvents. Specifically, Cs-133 and T1-205 NMR information was collected to deduce the complexing abilities of the sulfur containing crown for the alkali metal, cesium as compared to the toxic, heavy metal thallium. CHAPTER II HISTORICAL REVIEW The ever-growing number of existing synthetic macrocyclic molecules has provided for an exciting area of coordination chemistry research. As mentioned they have already provided many interesting and chemically important applications in the field of metal complexation in solution. Influences upon Macrocyclic Polyether-Alkali Metal Ion Complexing Stability Probably the most captivating characteristic of these compounds is their ability to bind cations selec— tively. This selectivity is a function of not only the various characteristics of the ligand but also of the cation and the solution composition, that is the solvent. Several excellent reviews concerning these parameters, for both the monocyclic and bicyclic ligands, have appeared in literature and, therefore, only a brief description of the effects will be presented and only in reference to the monocyclic crown compounds (23-26). The metal-ligand complexes are a result of ion- dipole interaction between the cation and the electro- negative donor atoms which are symmetrically arranged in the ligand ring. The factors affecting the formation and thermodynamic stabilities of the complexes include: (a) The relative sizes of the ion and ligand cavity; (b) The positioning of the donor atoms in the ring; (c) The type and charge of cation involved in com- plexation; (d) The type of donor atom which is present in the ligand; (e) The number of binding sites in the macrocyclic ring; (f) Substitution on the crown ring; (9) The solvent and extent of solvation of the ion and of the donor atoms. The cyclic polyethers (all donor atoms are oxygens) have been found to exist in predominantly one to one (1:1) meta1:ligand complexes. However, depending on the ratio of the diameter of the cavity and metal ion diameters, higher ratio complexes are also formed (27). In the case of the 1:1 complex, Pedersen suggested that the metal was centrally located within the ligand cavity with the oxygen atoms coplanar. This idea was substantiated by 10 x-ray crystal structure analysis performed by Bright and Truter (28). Bush and Truter (29) found that for the potassium complex of DB30C10 and other large ring macro— cyclic ligands, the structure of the complex was one in which all the oxygen atoms are wrapped around the potas- sium ion and are equidistant from it; however, they are not in the same plane. In reference to the polyethers of 18C6, it appears that the maximum stability, for alkali and alkaline earth cations, occurs in situations where the cation and cavity diameters match. This has also been found for transition metal ions Ag+, T1+, Hg2+, and Pb2+ by Izatt and coworkers (12). In the case of smaller and larger ringed crown compounds, not as great a selectivity is observed. In these cases, complexation reactions are complicated by formation of additional species, of stochiometries other than 1:1. Larger crown complexes may exist in which the metal ion is quite small in comparison to the ligand cavity. Hence, cavity size in relation to complex strength loses importance. The macrocyclic ligands possess a reasonable amount of flexibility and, therefore, conformational changes can often occur during complexation. This idea of nonrigidity appears chemically logical if one con- siders the basic donor groups, which are directed into the cavity, should possess a repulsive force against 11 each other. In terms of complex formation, Pedersen (30) found that the strongest oxygen donor atom was in a link of -OCH CH20-, followed by -O(CH2)3O- and became ineffec- 2 tive beyond -O(CH2)4O-. With respect to the alkali and alkaline earth metals, the metal ion appears to the ligand as a posi- tively charged sphere. The binding of the macrocyclic ligand, being electrostatic in nature, requires only that the ligand furnishes an electronically basic environment and that it can successfully compete with the solvation of the cation (15). The smaller ions are more strongly solvated than the larger ions and; there- fore, more energy must be gained through ligand-metal complexation to overcome the solvation energy. However, with a small charge density, the localized charge increases the affinity for the ligand. The larger cations possess a comparatively lower charge density, which reduces their affinity for water, and also for the ligand. Therefore, cations of intermediate size yield an Optimum change in enthalpy and generally greatest ligand selectivity lies with these ions. In terms of cation charge, the divalent ions generally have larger stability constants than the univalent ions of the same size. Also for large ions the divalent ions are more selectively complexed than the univalent ions. The opposite is true for small ions (24). 12 Substitution of either nitrogen or sulfur for an ether oxygen in the macrocyclic ring decreases the com- plexing ability of the ligand for the alkali and alkaline earth metal ions. Frensdorff (4) studied the effect of substitution and reported that for the 18C6 and DB18C6 ligands the potassium ion complexation is weakened by nitrogen and sulfur substitution and that the stability constants decrease in the same order as the electro- negativity of the donor atoms, O>NR>NH>S. This is not a surprising trend, since as the negative charge on the heteroatom decreases, the electrostatic attraction between it and the cation lessens. At the same time, however, there is an increase in the affinity toward transition metal ions of proper sizes. It would seem logical that the larger the number of binding sites, the stronger would be the complex. In the case of the polyether ligands, Pedersen (1) found that those containing five to ten oxygen atoms form the most stable complexes with the alkali and alkaline earth ions, with several transition metal ions, and with several quaternary ammonium organic cations. Cram and coworkers (31) reported that l8C5 complexed t-butylammonium to a much lesser degree than 18C6. Progressive replacement of oxygen atoms by sulfur atoms in the ring increases the stability of silver (I) and mercury (II) complexes (4). 13 Izatt et a1. (12) studied the complexation selec- tivity of 18C6 and substituted 18C6 ligands. For the two isomers of dicyclohexyl-18C6 (DCH18C6), with cis- syn-cis and cis-anti-cis conformation, quite different stability constants were obtained, with alkali and alka- line earth metal ions, transition metal ions, and quater- nary ammonium organic cations. In comparison to the unsubstituted 18C6, complexation of bivalent cations with DCH18C6 (except for Ba2+) yielded enhanced stability constants, for the cis-syn-cis isomer, but no effect was observed on the selectivity for univalent cations. Sub— stitution in the cis-anti-cis conformation lowered the stability of complexes with univalent cations (except in the case of Na+ and Ag+) but showed little change with the bivalent cations. Izatt interprets the results as an indication of the cis-syn-cis isomer's cavity being better defined and less sensitive to conformational changes due to the ring rigidity attributable to the cyclohexyl groups. A comparison of the stability constants of DB18C6 complexes to those of 18C6 complexes shows a more dramatic effect. In methanol solutions the complexation of 18C6 with Ba2+ is an order of magnitude greater than with K+. However, if DB18C6 is used as the ligand, the K+ binds more strongly than the Ba2+. Lamb et a1. (23) have dis- cussed the above results and concluded that the bulkiness 14 of the ligand isolates the cation from the solvent by the hydrophobic groups, which prevents the regain of energy lost in desolvating the cation. The desolvation energy required for a dipositive ion is greater than for a unipositive ion resulting in a destabilization of the complexes of the former species. In solution studies of ligand-cation complexation, a competition for the cation exists between the ligand and the solvent. Because of this competition and because of the varying properties of different solvents, it is not surprising that the stability constant of a given cation- ligand complex depends upon the environment created by the solvent medium. This dependence may also cause a change in the selectivity of a particular ligand. Arnett and Moriarity (32) observed such change in selectivity in their calorimetric study. Heats of complexing were obtained by measurement of the partial molar heat of solution (ASS). Two methods were employed: first, the salt in pure solvent and in a crown ether-solvent solution were measured; and, secondly, by measuring Afis of the crown in pure solvent and then in a salt- solvent solution. It was explained that if there is no ion pair interactions (between the uncomplexed cation and the associated anion), it may be assumed that the formation constant, as well as the enthalpy of for- mation, for the complex is determined solely by the 15 competition between the ligand and solvent. They ex— plained that the single ion enthalpies of transfer vary in different solvents being greatest for small ions of high charge density. The greatest apparent selectivity of DCH18C6 was found in dimethyl sulfoxide (DMSO) and water, which strongly solvate small cations. With less polar solvents, the fit of the cation in the ligand cavity is more important, with less affinity for the solvent the selectivity factor is reduced. Although the dielectric constant of the solvent has been shown to influence complex stability, it is not the only parameter which affects the stability. Matsuura and co- workers (33) have reported formation constants for alkali metals with DBl8C6 in DMSO, dimethylformamide (DMF), and propylene carbonate (PC) and found that the values obtained were in the order of DMSORb>Cs>Na. In their study of DB18C6 complexes, thermodynamic values of the enthalpy change, entrOpy change, and free energy change due to complexation are presented. The validity of these quantities is questionable since their value is dependent on the extraction constants, which in the mathematical treatment used, did not consider the dissociation of the complex ion pairs. Mitchell and Shanks (55) studied the extraction of sodium into chloroform, by crown complexation, forming the ion pair complex (Na°DCHl8C6)+Ph4B-. The accurate determination of quantities in the order of ug Na+/g sample was made possible by thermal neutron irradiation of standard and sample solutions, followed by comparison 27 of the sodium isotope activity. Analysis of the inter- ference effects of other alkali and alkaline earth metal ions, a few transition metals, and various anions were performed, producing little effect except in the case of K+. Primary standard sodium oxalate solutions and silicon dioxide samples showed high precision in both separation and determination of sodium. Jaber and coworkers (56) conducted studies designed to investigate the complexing abilities of nonylphenoxypoly (ethyleneoxy) ethanols (NP), a series of nonionic surfactants used in industry, with alkali and alkaline earth metal ions. The rigid cyclic arrange- ment of donor atoms, as the crowns possess, is not a necessity for complexation of a neutral molecule and metal cation. These workers examined the complexing tendency of the open chain structure of NPs in hopes of using these molecules as ion-selective electrode sensors. The extraction technique, between methylene chloride and water, was utilized to assess the extent of metal ion complexation by NP. The order of stability for the alkali metal ion-NP complexes was found to be Rb>K>Cs> Na>Li, while for the alkaline earth metal ions the stability decreased Be>>Ba>Sr>Ca>Mg. A number of papers have been recently published dealing with the coextraction of water with alkali and alkaline earth metal hexanitrodiphenylaminates (HNPA) 28 into nitrobenzene (57-59). Since the cations are strongly hydrated, the water molecules are transferred to the organic phase with the cation and with HNPA as water- separated ion pairs. The authors found that in the presence of a number of different crowns the cationic hydration number decreases, implying partial replacement of water molecules by the oxygen atoms of the crown. In several cases the number of released water molecules almost equaled the number of donor atoms of the crown. Danesi and coworkers (60) performed several extractions using DB18C6 in various nitrobenzene-toluene mixtures, against aqueous solutions of the alkali metals. They found the same selectivity sequence, K+>Rb+>Cs+> Na+>Li+, for all diluent compositions. Independent of the relative amounts of nitrobenzene and toluene in the organic phase, the same value for the selectivity con- . O O + O stant (w1th1n error) of the react1on Cs + MLP1c ———> org r— CsLPicorg + M+ was observed for all the alkali metals, implying an apparent isostericity of the complexes. These results were attributed to the presence of water molecules shielding the cation, within the crown, from surrounding solvent medium. To substantiate this line of reason, Danesi et a1. (61) conducted a study of Na+ and K+ complexes with DB18C6 in solvent mixtures that were carefully dried. In these studies they were able to detect differences in the selectivity constant due 29 to the dissimilar interionic attractions in the ion pairs of Na+ and K+. These studies showed that the additional shielding the water molecules supplied the complexed cations was enough to considerably reduce any difference in the closest approach distance allowing for isosteric behavior of the complexes. Danesi and coworkers (62) also prepared a sub- stituted DBlBC6 macrocycle in which four neutral alkyl t—butyl groups were placed on the benzene ring, in hopes of improving the solubility of DB18C6 in the organic phase. Studies in various nitrobenzene-toluene mixtures, with the cesium ion, indicated an enhancement in solu- bility; however, a slight decrease in cation complexation was also observed most probably due to steric effects. In a more recent publication of the use of the extraction technique with polyethers by Marcus and Asher (63), comparative extractions of alkali metal halides are discussed. Studies were conducted in a variety of solvents and an attempt was made to correlate their find- ings with solvent properties. They report the order of extractibility among the halides as F>I>Br>C1. Sekine et a1. (64) extended the use of the extraction procedure to include thallium cation (I and III) complexation in benzene and chloroform. The quantity of Tl+ extracted into the organic phase by DBlBC6 was determined by atomic absorption. The extraction 30 equilibrium constant for the complex was calculated and compared to those of alkali metal ions in benzene. The equilibrium extraction constant they reported for thallium was log K = 4.49, as compared to sodium 2.20, potassium 4.65, rubidium 3.75, and cesium 3.07. Sulfur-substituted polyethers have also been studied by the extraction technique. Pedersen (43) per- formed extractions of K+, Na+, and Ag+ with several of these ligands in methylene chloride. The picrate ex- tracted (the indicator of complex formation) was less than 6% for K+ or Na+ with the sulfur substituted crowns;, however, excellent extraction efficiency was obtained with the silver ion. Sevdic and Meider (65, 66) studied the complexing abilities of 1,4,8,ll—tetrathiacyclotetradecane (TTP) and 1,4,8,ll,15,18,22,25-octathiacyclooctacosane (OTO) for silver (I) and mercury (11), chloride and perchlorate, by extraction between water and nitrobenzene. They have shown the enhanced complex strength of the sulfur con— taining crown molecules with Ag+ and ng+. For example, they have reported extraction equilibrium constants log Ke 11.11 for Hg(ClO4)2 complexation with TTP and log Ke 12.13 for the ligand OTO. Raman Studies of Solvation and Ion Association The existence of a weak complex between 18C6 and the solvent acetonitrile (ACN) has served as an 31 intermediate in the purification of the crown, as described by Gokel and coworkers (67). However, it has not been determined if the complex produces any observable change in the Raman spectrum, compared to the spectra obtained for pure 18C6 and pure ACN. Both infrared and Raman spectrosc0pic methods are of great value since they provide information at the molecular level, allowing even weak local intermolecular forces to produce rather easily observable frequency changes in the intramolecular vibrations (68). Several studies have been conducted in which IR or Raman, or both, techniques have been used to study solvation and ion association in nonaqueous solutions. These spectroscopic methods have also been used in 2+ (68), and Na+ (68, 69) per- studies of Li+ (68), Mg chlorates, of Na+ and K+ DB18C6 complexes (20), and of Li+ and Na+ cryptates (70) in several nonaqueous solvents. Even more specifically to the case to be presented here are studies performed using the solvent acetonitrile and observing its characteristic CEN (2253 cm-1) and C-C (919cm-l) stretching frequencies change, as a function of electrolyte concentration. Balasubrahmanyam and Janz (71) studied the AgNO -CH3CN system at various 3 concentrations of AgNO3. Analysis of the concentration dependence of the relative intensities and frequencies obtained of the CEN and C-C bands in AgNO3-CH3CN 32 solutions helped in understanding the solute-solvent interactions occurring in the solution. These authors reported the appearance of bands at 928 cm"1 and 2272 cm-1 (in addition to the 2253 cm.1 and 919 cm-1 bands) which were attributed to the C-C symmetric stretch and CEN symmetric stretch of the complex, respectively. Chang and Irish (72) extended the above studies by further characterizing the system, measuring the formation constant of the ion pair species Ag+NO3-. They also provided evidence for multiple ion aggregates and the solvation of the Ag+ and Ag+NO3 in solution. Nuclear Magnetic Resonance To a solution chemist, or in fact to any indi— vidual interested in studying a chemical reaction occur- ring in some solvent, a knowledge of what species are present and the influence of solvent properties is of utmost importance. However, this is only the beginning of the complex behavior of the different species which may be present. A variety of interactions, including solute-solute, solute-solvent, and solvent-solvent type, add tremendous complexity to any solution. The result of solute-solute interaction is the formation of ion pairs, which may be a single anion—cation attraction or be composed of several ions forming higher aggregates, depending on the nature of the ion. Two properties of the solvent come into play here: 33 (a) The dielectric constant of the solvent (b) The solvent donor capabilities (solvating ability) Solvent-solvent interactions of a given solvent may also change from solution to solution depending upon the nature of the ionic species introduced. In addition to the reaction under study, these solvent-solvent inter— actions may be important. Many spectroscopic techniques have been used in hopes of understanding the variety of interactions occur- ring in a simple solution. One of the more recently utilized techniques is nuclear magnetic resonance (NMR), which has proved to be very sensitive to these solution interactions. Alkali metal NMR has been studied extensively in this laboratory, not only in relation to the study of ionic association and solvation of alkali metal salts, but also in complex formation with a variety of ligands, particularly tetrazoles and macrocyclic crown ethers. An older but very excellent review of NMR as an investigative tool for the study of ions in different solvents has been published by Hinton and Amis (73). A more recent review of specifically alkali metal NMR has been presented by Popov (19). POpOV discussed the importance of this technique in studying the immediate chemical environment of alkali metal ions in solution. The NMR studies of solutions of alkali salts, alone and 34 in the presence of a complexing ligand, were shown to be helpful in better understanding both solvation and com- plexation of these cations. Cesium-133 NMR A number of investigations of ionic association and solvation have been reported using nuclear magnetic resonance of 133Cs. Those reported by DeWitte (74), Hsu (75), Mei (76), and Hourdakis (77) have characterized ion pair formation of the cesium salts and cesium com- plexation with crowns and cryptands, in several non- aqueous solvents, through chemical shift studies. The physical properties of the 133Cs nucleus are shown in Table 1. Different chemical shifts arise from the fact that the magnetic field experienced by the nucleus depends upon its environment and is not the same as the applied field. The screening of a nucleus, represented as the shielding constant (6) incorporates several terms. The diamagnetic shielding factor (6d) arises from the induced magnetic field due to the circulation of elec- trons around the nucleus. The paramagnetic term (6p) is a deshielding factor caused by the interaction of ground state with excited electronic states in the presence of the magnetic field. The shielding of the 35 Table 1. NMR Properties of the 133Cs and 205T1 Nuclei. 133Cs 205Tl Resonance frequency at 14.09 kgauss field (MHz) 7.8709 34.6113 Natural abundance (%) 100 70.48 Relative sensitivity for an equal number of H at constant field 0.0474 0.192 Nuclear spin (in units of h/2n) 7/2 1/2 Nuclear magnetic moments (in units of nuclear magnetons, eh/41TMc, with diamagnetic correction) 2.579 1.627 Electric quadrapole moment (in multiples of barns, 10’24cm2) -0.003 Relaxation time (T1, in secs) 13 0.4 36 resonant nucleus by other atoms is represented by do. The relative importance of the paramagnetic term increases as the number of electrons around the atom increases (78). For nuclei larger than 7Li, 6p is the dominant factor in determining the shielding constant (79). 133Cs chemical shift of DeWitte (74) studied the several cesium salts in different nonaqueous solvents and in water as a function of concentration. His results showed only a small concentration dependence of the chemical shifts for solutions of cesium perchlorate and tetraphenylborate but comparatively large dependence of similar solutions of cesium thiocynate, iodide, bromide, and chloride. The constancy of chemical shift with a changing concentration implies no ion pair formation whereas a concentration dependence is evidence of contact ion pair formation. From the data obtained, asSociation constants (ion pair formation constants) were determined. Hsu (73) reported several formation constants for complexes of different cryptands with cesium in a number of solvents. A temperature dependence study was also conducted for complexes in pyridine. Mei's (76) work was concerned with the complex- ation of polyether crown and cryptand compounds with cesium salts. Complex formation constants were obtained for all systems studied and thermodynamic data (AG, AH, 37 and AS values) was calculated from a temperature depen- dence study of cesium tetraphenylborate with 18C6 and C222. Complexation studies of the dilactam of C222 by 133Cs NMR were reported by Hourdakis (77). The historical section of these theses amply cover the literature through 1977 and will not be pre- sented again. Thallium-205 NMR The alkali metal ions have been shown to fulfill at least three functions in biological systems: (a) Neutralization of ionic charges and maintainance of macromolecular conformations; (b) Enzyme activation; (c) Maintainance of a membrane potential for nerve and muscle function. Several studies have shown that thallium (I) may substi- tute for alkali metal ions performing a number of these same functions (80). As a result of these findings, the study of the properties of thallium have proved interest- ing and seems to be a logical extension of alkali metal complexation studies. Table 2 shows some physical properties of thallium (I) nucleus and of the alkali metal cations. The ionic ‘ . . . . + radius of T1+ is quite Slm1lar to that of K but the 38 mm I HO mm.o Nw.H N.m vm.a HE mm I mm.o mn.o vm.m em.H mo cm a >~.o mm.o mm.H mm.a mm um I om.o mm.o mm.H «v.H M boa: mm.o mm.o av.c NH.H oz vmal om.o mm.o mo.o wm.o flu anthem... a a. a a a . coupowam >uw>flummmcouuomam muflHfiQmeumHom Aflomm owcoH 2 co saaccucm mcoflumo Hmpwz flamxad map can AHV Esflaamne mo mmfluummoum HMUMmhnm .N manna 39 molecular polarizability of Tl+ is considerably greater than that of any alkali ion. Therefore, the nature of bonding and, in general, solution behavior of Tl+ should be quite different from those of the alkali ions. From the enthalpy of formation, shown in Table 2, one may see that Tl+ is only weakly hydrated in solution. Several workers have used 205T1 NMR as a tech- nique to study some problems of biological importance, using thallium as a probe of the role of alkali ions. For example, Reuben and Kayne (81, 82) have found 205Tl NMR suitable for the investigation of the binding of thallium ions to biological macromolecules and have utilized it in a study of the complex between Tl+ and rabbit muscle pyruvate kinase as well as of substrates of this enzyme. It had previously been shown that both a monovalent and divalent ion must be present for the pyruvate kinase enzymatic reaction to occur, but the role of the monovalent ion had not been understood. These workers determined the number of monovalent bind- ing sites in the enzyme through substitution of Tl+ for K+. Thallium-205 NMR has also become an important technique in examining solvation, solution structure, and complexation of thallium salts. The physical 205T1 nucleus are shown in Table 1. 205 properties of the The solvent dependent chemical shift for T1 may be 40 as great as 2600ppm, depending on the particular thallium 23 salt and solvent under study, while 7Li is 5ppm, Na is 39 133 23ppm, K is 35ppm, and Cs is 120ppm. This greater 205Tl a better range of chemical shifts should make probe of smaller changes in the environment of the nuclei (83). The spin-lattice relaxation time has also been shown to be strongly solvent dependent (84). Freeman et a1. (85, 86) have reported chemical shift data for different thallium salts in aqueous solutions. They found that at low concentrations the chemical shift varied nonlinearly with concentration of anion, which was attributed to ion pair formation. In solutions of higher concentration the variation of chemical shift with concentration was linear, which Freeman explained as being due to the effect of ions on the hydration atmosphere of the ion pair. The shift extrapolated to zero anion concentration is character— istic of the free solvated thallium (I) ion. In the second publication, these workers reported chemical shifts and line widths of solid thallium (I) and (III) compounds, as well as for aqueous solutions of thallium (III) salts. Line widths of the solid compounds were quite broad so that the precision with which the chemi— cal shifts were measured was low. Line broadening in the solutions was explained by assuming varying con- tributions from spin-spin coupling between both the T1+ ion and the anion and between unlike thallium isotopes. 41 Several studies concerning the solvent and anion dependence of chemical shifts have been performed by Dechter and Zink (e.g., 81 and 87). They reported chemical shifts extrapolated to infinite dilution in a number of nonaqueous solvents. The aim of the inves- tigation was to determine whether or not the change in chemical shift was related to specific donor atoms or groups or corresponded to the Lewis basicity of the solvent. The results showed that there were two regions of shifts, one related to nitrogen coordinating and the other to oxygen coordinating solvents. An attempt to find a correlation between the chemical shifts and some bulk solvent property or some measure of solvent polarity was not possible. In contrast to Dechter and Zink's results, Hinton and Briggs (88) reported a reasonable correlation of the extrapolated infinite dilution resonance frequencies with solvent Gutmann donor number;* the resonance frequency increased with increasing donor number. They also found that the lower donor number solvent, DMSO, solvated the thallium ion better than did pyridine (py). This was also found 23 by POpov and coworkers (89) in Na NMR studies and explained in terms of solvent structure effects. Hinton *Defined as the negative of the enthalpy of interaction of a base with SbC15 when the two are dis- solved in equimolar amounts in an inert solvent, 1,2- dichloroethane. 42 and Briggs (90), in agreement with Popov's reasoning, explained that this phenomenon was due to a large solvent structural effect or to a steric effect related to the size of the py molecule which favors solvation of T1+ by DMSO. Hinton and Briggs have done several preferential solvation studies of thallous ion occurring in binary systems of all combinations of water, formamide, N- methylformamide, and N-ethylformamide (91) as well as water, pyridine, and dimethyl sulfoxide (90). From their 205T1 and the solvent Gutmann chemical shift data of donor numbers, they calculated needed donor numbers by a linear least squares analysis. They derived mathemati- cal expressions for their binary systems to calculate the chemical shift due to donicity effects and assigned the difference, between this value and the observed chemical shift, to the shift caused by structural effects. They concluded from their findings that steric and structural effects are quite important in liquid ion— solvation processes, in fact, often more so than donicity effects. Early studies of aqueous thallium-DB18C6 com— plexation were conducted using ultraviolet spectroscopy by Pedersen (1) and Shchori et a1. (92). Koryta and Mittal (93) calculated stability constants of alkali metal ions and thallium ion for DCH18C6 by electrochemical 43 reduction of the metal complexes at a dropping mercury electrode. Rodriguez and coworkers (94) reported a kinetic study of the complexation of Na+, K+, Rb+, Tl+, and Ag+ cations with 15C5 in aqueous solutions by ultra- sonic absorption using a laser acousto-optic technique. The NMR technique has been employed recently for studies with both macrocyclic polyethers and cryp- tands. Srivanavit and coworkers (95) developed a method of measuring stability constants of several univalent cations relative to the thallium ion, with a number of polyethers, by 205T1 chemical shifts. Measurements were obtained in methanol, dimethylformamide, and dimethyl sulfoxide. Concentration, anion, and solvent effects are discussed in conjunction with the method. Complexation of thallium with cryptand C222 has been studied revealing the stability constant of Tl+ in water to be greater, by an order of magnitude, than any of the alkali metals (96). The alkali metal ion com- plexes with C222 show a good correlation between the ionic radius and the stability constants; however, the stronger complexation with thallium implies that other effects must also be important here. CHAPTER III EXPERIMENTAL MATERIALS, INSTRUMENTS, AND PROCEDURES Materials Reagents Hydrochloric acid (Mallinckrodt), picric acid (J. T. Baker), phosphorus pentoxide (Fisher), calcium chloride (anhydrous, J. T. Baker), sodium carbonate (anhydrous, Mallinckrodt) as well as all hydroxides of the metals, lithium (LiOH-H20, Ventron-Alfa), sodium (Drake Brothers and Fisher), potassium (Fisher and Matheson, Coleman and Bell), and cesium (Ventron-Alfa, 99.9%), were used as received. Potassium hydrogen phtha- late (Matheson, Coleman, and Bell), used in the standardi- zation of the hydroxides, was dried at 110°C for at least two hours prior to use. Potassium acetate (J. T. Baker) was dried in an oven at 110°C for 24 hours. While cesium perchlorate (Alfa Inorganics, 99%) necessitated drying at 110°C for a minimum 24 hours, the other cesium salts used required additional treatment. Cesium acetate (Alfa, technical) 44 45 was recrystallized from methanol and vacuum dried over P205 for 48 hours. Cesium thiocynate (Rocky Mountain Research) was recrystallized from absolute ethanol and vacuum dried. Cesium tetraphenylborate* was precipitated from a metathetical reaction of sodium tetraphenylborate and cesium chloride in a mixed solvent system of tetra- hydrofuran and water. The precipitate was filtered and washed with conductance water followed by drying under vacuum for 48 hours. Both thallium nitrate (Ventron-Alfa) and thallium perchlorate (K and K) were dried at 110°C for 1.5 hours, while thallium acetate (Ventron-Alfa, ultrapure) was dried under vacuum for 48 hours. Thallium thiocynate was prepared by the addition of thallium acetate to sodium thiocynate solution (J. T. Baker) in methanol and dried under vacuum. Ligands The complexing ligands used in this work are macrocyclic crowns 18C6, DBlBC6, and DT18C6. The 18C6 (Aldrich) was purified by forming a complex with aceto- nitrile (Matheson, Coleman, and Bell), according to the method developed by Gokel and coworkers (67). In addition, Shih (97) suggested that the procedure be * Synthesis performed by Adamantia Hourdakis. 46 followed by dissolving the 18C6 in a minimal quantity of acetone (Drake Brothers) and vacuum filtering to remove any remaining impurities. The recrystallized 18C6 was dried under vacuum for 48 hours. Dibenzo-18C6 (Aldrich) was dried over P205 under vacuum for 48 hours. The melting point agreed well with the literature value of 164°C (1). The sulfur-substituted DT18C6 was successfully synthesized following Bradshaw's procedure (44) to pre- pare l-thia-9C3; DT18C6 is a solid by-product of this synthesis. An oligoethylene glycol dichloride was allowed to react with sodium sulfide as shown in the reaction /-\ 95% Cl CZHSOH NaZS + NaOH + -——-——+ C1 reflux \__/ O + C + 2NaCl l—Thia-9-crown-3 LSJ 1,10-Dithia-18-crown-6 A solution composed of 48g sodium sulfide non- ahydrate (Mallinckrodt) and 0.96g sodium hydroxide (Drake Brothers), in 95% ethanol, was prepared and placed in a large three-necked round bottom flask. Also 47 prepared was a solution containing 37g bis-(2—chloro- ethoxy)ethane (BCEE) (Eastman) in 300ml of 95% ethanol. This solution was added to the NaZS-9HZO/NaOH/EtOH solution in a dropwise fashion from a separatory funnel. The reaction flask was equipped with a mechanical stirrer as well as with a nitrogen gas inlet. In order to optimize the yield of the desired product, it was necessary to keep the reaction mixture at high dilution, vigorously and continuously stirring, under a nitrogen atmosphere. Once the two solutions were completely mixed, the resulting solution was refluxed for approxi- mately 8 hours and finally allowed to cool to room tem- perature. The highly basic solution, initially made so to facilitate the reaction, was neutralized with hydro- chloric acid. A solid precipitate, (NaCl) which had formed during refluxing, was filtered off from the solution and the ethanol was removed using the Rotovap-R (Bfichi). The oil, which remained, was extracted four times with ether. Bradshaw reported that the DT18C6 formed in the ether extract and must be removed before any attempt to continue the 1-thia-9C3 recovery. How- ever, no product was found in the ether fraction, rather, it was found crystallized among the precipitate filtered previously. The crystals were manually separated from the precipitate. A second synthesis performed reacted more typically. The product, DT18C6, was recrystallized 48 several times using a 50-50 mixture of chloroform (Fisher) and hexane (J. T. Baker). Following the purification of the DT18C6, by recrystallization, a melting point was measured on the Fisher-Johns Melting Point Apparatus and was found to be 89-90°C. Literature values are 90-91°C reported by Dann and coworkers and 89-90°C reported by Bradshaw. A proton NMR of the compound was also obtained using the Varian T-60 NMR (Figure 2). A 0.14M solution of DT18C6 in deuterated chloroform (Aldrich Diaprep, 99.9%) was used. Tetramethylsilane (Merck & Co.) was the internal standard. The chemical shifts obtained are shown in Table 3 and compared to those obtained by Bradshaw (44). The proton NMR is very characteristic for these compounds and this spectrum seemed to verify that the desired compound was prepared. A direct syn- thesis of DT18C6 was first attempted using the method proposed by Dann, Chiesa, and Gates (33). A solution of Na S°9H 2 2 pared and pure BCEE was added dropwise. The reaction 0 in a 50% ethanol water mixture was pre- mixture was refluxed for 88.5 hours, since the workers reported a greater yield with longer refluxing times. After the completion of refluxing a yellow oil formed, which should contain the product. It was separated from the solution, dissolved in hot ethyl acetate (Fisher), cooled to 7°C in a dry-ice-acetone 49 Accmn mowm mcflccwmm u mmmv .mocmummmu Accumucfl coca lamawsumEmuuou nuflz ShoonoHSU omumumusmo cw muwaao mo mzz cououm .N .mHm 2mm o.o o.H o.~ o.m o.v o.m o.o o.h o.m q — ,4 a _ _ _ _ mmm mmm plhihld H.WJ.MEPLPUAWI§IIIIIII . 1 mo.o am>ma Hmzom .m.m a Hmuaflm um oom nuofl3 mmmzm com omm mEfiu mmm3m mmm ow mums mcascwmm .¢.m m.m um ca mosuflamsm Hmummucw mza o.~ wHQEmm m.m mosuflamam Esuuowmm mm o ummmmo nuoflz mmo3m muwumamumm mzz owIB :maum> 50 bath, and the solution was quickly filtered to remove the white wax polymer which formed as a reaction by- product. The alcohol-ethyl acetate filtrate was evapor- ated to a small volume. At room temperature the product should form as a white solid, however, no crystals were obtained and the second by—product, an oil, was left. The polymer was formed in such a large amount that it was felt the reaction went predominantly in the direction of that product. Table 3. The Chemical Shifts, Both Observed and Litera- ture Values, Obtained for DT18C6/Deuterated Chloroform Proton NMR. Protons Splitting Literature Observed Responsible Pattern Chemical Shift Chemical Shift SCH2CH2o triplet 2.82 2.37 OCHZCHZO singlet 3.63 3.55 OCHZCHZS triplet 3.72 3.64 The method used by Bradshaw, although less direct, did produce the desired product and therefore was used for all subsequent synthesis of DT18C6 used in this work. In one study the DT18C6 used was obtained from Parish Chemical Co. and required vacuum drying for 48 hours. For the Raman studies of the 18C6-acetonitrile complex, the crown, 18C6, was purified prior to use and 51 acetonitrile (ACN) was used as purchased. The 18C6-ACN complex was obtained by combining the crown and solvent, cooling in an ice-acetone bath and quickly vacuum filter- ing the precipitate. Solvents The nonaqueous solvent used in the extraction studies, methylene chloride (Drake Brothers and Mallin- ckrodt), was purified according to a method by Mathews (98). The process included first washing the methylene chloride with an equal volume of distilled water and then with an approximately 0.1M aqueous sodium carbonate solution. The washing procedure was carried out in a separatory funnel. The organic solvent was then separ- ated, dried over anhydrous calcium chloride for 48 hours, and fractionally distilled. An additional precaution was taken, based on a known chemical reaction between the related solvent, chloroform. The reaction CH C1 + O ——9 COC1 + HC1 3 2 2 is known to occur when chloroform is Open to atmospheric oxygen. A simple chemical test for the detection Of the decomposition involved an extraction Of the solvent with distilled water; if the chloride ion was present (due to HC1 formation), it would be transferred to the aqueous phase. Addition Of a concentrated silver nitrate solution 52 to the aqueous phase would result in the visible precipi- tation of silver chloride. However, no such precipitate was formed and decomposition was not a problem. Toluene (Mallinckrodt), employed for a single extraction, was fractionally distilled before used. Solvents used in the NMR studies included aceto- nitrile (Matheson, Coleman, and Bell), nitromethane (Aldrich), dimethylformamide (Mallinckrodt), pyridine (Fisher), and dimethyl sulfoxide (Fisher). Acetonitrile was refluxed over calcium hydride and fractionally distilled. Nitromethane and dimethyl- formamide were fractionally distilled, under vacuum, over phosphorus pentoxide. Pyridine was refluxed over granulated barium hydroxide and fractionally distilled in a nitrogen atmosphere. All solvents were dried over molecular sieves for at least 2 days prior to use.* Dimethyl sulfoxide was allowed to sit over freshly activated molecular sieves (8-12 mesh, Davison) for 24 hours, followed by vacuum distillation onto new sieves, which had been thoroughly washed prior to acti- vation. Molecular sieve activation entailed heating them at SOD-700°C in a tube furnace (Lindberg) for 24 hours with a constant flow of nitrogen (bubbled through * These solvents were prepared by members of the research group for community use. 53 sulfuric acid). The automatic Karl Fisher Titrator (Aquatest II, Photovolt) was used to test water content of the solvent, whenever possible, in order to assure the maintenance Of water below lOOppm. Following the purification and drying steps, all solvents were placed in a dry box dry nitrogen atmosphere to prevent recon- tamination by water. Instruments and Procedures Extraction Studies Direct preparation of a known concentration Of an aqueous solution of picric acid was impossible due to the necessity of always keeping the solid slightly wet in order to avoid explosive decomposition. The concentration of the solution, therefore, was determined potentiometrically. The pH was measured at various increments of a standard NaOH solution with a Beckman Combination Electrode and a Heath Servo Digital pH/Volt Meter. The basic procedure for the preparation Of any of the systems for extraction measurements were identical. The crown compound studied was dissolved in the methylene chloride and a series of dilutions was made to Obtain varying concentrations of the crown solution. The aqueous phase consisted Of the particular cation Of interest, picric acid, and a hydroxide. The extractions themselves were carried out by shaking equal volumes of 54 the organic and aqueous solutions for a predetermined time, to ensure establishment of equilibrium. The solutions were shaken in the Wilkens-Anderson Shaker Bath. The systems were then allowed to set until a definite boundary between the two phases developed (approximately 1/2-1 hour). Ultraviolet absorbance Of the picrate component of the aqueous phase was measured both prior to and following the extraction procedure. The initial absorb- ance value provided the total picrate concentration before extraction. The absorbance after extraction was a measure Of how much picrate was extracted into the organic phase with the metal cation. Measurements were Obtained on systems containing a variety of component concentrations. Based upon calculations using an equation derived by Frensdorff (22), the dissociation constant (Kd) Of the complex in the organic phase as well as an overall extraction equilibrium constant (Ke) were obtained. A more detailed look at the equilibria in solution and equilibrium constant calculations will be presented in Chapter IV. All solutions were prepared using volumetric glassware and dilutions were made from a master solution delivered in various amounts from a buret. Flasks were wrapped with teflon tape and parafilm and the time of exposure to the atmosphere was minimized as much as 55 possible to eliminate the difficulties arising from solvent evaporation. All absorbance measurements were Obtained on a Cary 17 UV-Vis Spectrophotometer at ambient room tem- perature, using rectangular one centimeter quartz cells. Any measurements made on organic phase components were made in cylindrical quartz cells with a ground glass stopper and wrapped with teflon tape. The Unicam SP.800 UV-Vis Spectrophotometer was used for some qualitative runs. An aqueous solution of K2CrO4 (dried at 110°C for 2.5 hours) in approximately 0.05M KOH was used as a standard to periodically check the spectrophoto- meter's performance (99). Infrared Absorption and Laser Raman Infrared spectra of the 18C6 ligand in aceto— nitrile solution, at room temperature, were obtained with a Perkin-Elmer 237B and 457 Grating Spectrophotometers. The samples were measured using sodium chloride discs (Barnes Engineering Company). Spectra were recorded of pure ACN, of pure nujol, and of 18C6 in nujol, so that the frequencies at which the bands were observed could be compared and assigned to a correct component. The spectra collected for pure 18C6, ACN, and the 18C6-ACN complex were also Obtained on the Spex Ramalog 4 Laser—Raman System which includes the Spex 1401 56 Double Spectrophotometer. The Spectra Physics-265- Excitor Argon Laser with the Ar-ion 5145A line was utilized as the excitation source. In all cases the pulse counting mode was employed. Typical instrumental parameters are listed in Table 4. Table 4. General Instrumental Parameters Used for the Collection of Raman Spectra. Laser 5145A Ar source Laser power 1 Watt Photomultiplier power 1940 Volts Chart speed 5 cm/min Pen period 0.5 Slit height 2 mm The Raman spectra Of the crown and the solvent- crown complex were obtained from a solid sample, whereas the ACN spectrum was Obtained from a liquid sample. All samples were placed into l.6-l.8 x 90 mm capillary tubes and sealed using tightly sealing parafilm. Once the complex had been formed and placed in the tube, it was necessary to store it in an ice bath because the complex noticeably broke down at room temperature. Since the laser beam did not heat the samples being measured, there was no concern Of complex decomposition during spectrum collection. 57 NMR Specrometer 133 205 Measurements Of Cs and T1 resonances were Obtained with a Varian DA-60 spectrometer with a magnet Operating at a field of 14.09kG. An external proton lock at 60MHz was used to maintain a stable field. The DA-60 is equipped with a wide-band probe (100), capable of accommodating different tunable inserts, allowing for multinuclear studies. Time averaging of spectra and Fourier transformation Of data were made possible with an interfaced Nicolet Instrument Corporation 1083 com- puter. Wright and Rogers (101) have discussed the com- puter controlling, Of necessary rf pulses and time delays for the particular nuclei under study, which the system possesses. The glassware used in sample preparation was carefully cleaned and dried at 110°C overnight. All samples were prepared from stock metal solutions and various amounts of the complexing ligand under nitrogen atmosphere of the dry box. Upon complete dissolution of the ligand the solutions were transferred to 10mm OD NMR sample tubes (Wilmad), capped, and wrapped with teflon tape to prevent both contamination by atmospheric water and solvent evaporation. Chemical shifts Obtained for the different metal complexes were referenced to 0.5M aqueous solution of 133 CsBr for Cs and to 0.3M solution Of TlNO3 in water 58 205T1 measurements. Cesium-133 chemical shifts for reported were finally referenced to infinite dilution, Of the nuclei in water (a correction factor of -9.59ppm for 0.5M CsBraq reference). All chemical shift data were Obtained at ambient probe temperature. A positive value for a chemical shift is indicative of an upfield (diamagnetic) shift, while a negative value implies a downfield (paramagnetic) shift. The chemical shifts are corrected for bulk sus- ceptibility (102, 103) of the solvent in relationship to the water reference. Live and Chan (103) discussed the expression _ ref _ sam Scorr - 6Obs + 2/3(xv Xv ) for nonsuperconducting spectrometers, which corrects for the difference between the reference and sample solvent. The var1able Sobs 13 the Observed chemical sh1ft; acorr is the chemical shift corrected for bulk susceptibility; ref and XV and Xiam refer to the volume magnetic suscepti- bilities of the reference and sample solutions, respec- tively. Table 5 contains the corrections applied as well as some of the important solvent properties to consider in discussion Of solution behavior. The collection of chemical shifts as a function of the relationship of crown to metal for a given solution, measured in terms of mole ratio of crown to 59 .Uomm um omcflmpno wum3 mumnuo Ham “Doom um oousmmwfi mucwEOE OHOQHQ .4. cc.o omc.o- o.mm cm.ma sm.H scum: mm.ou mec.o- H.mm oe.me sm.m acacaasm ac.ou Hmm.ou a.~ a.mm acm.m cccccmsouuaz em.c- moc.ou m.m~ cc.cc m.m ccaxomaem Hacuwsao Hm.o- mum.cu c.c~ Hs.cm cm.m measmeuooascumsao mm.cu amm.ou H.6H m.sm ce.m meauuacoucoa AEmmv cofluomuuou A onv >uflaflnflummumsm wonfizz Hocoo ucmumcoo Asmw mHIoHv chasm Hmoasoco c cauccseao> escapee oaauocecac wmmmmm ucm>aow .xuoz wage :a com: mcofluomuuou muflaanflumoomsm xasm cam mOAuuomoum ucm>aom meow .m manna 60 metal, allowed for determination of complex formation constants. A nonlinear curve fitting program, KINFIT, written by Dye and Nicely (104) and the CDC 6500 Computer system were used for this purpose. CHAPTER IV SOLVENT EXTRACTION TECHNIQUE FOR THE DETERMINATION OF MACROCYCLIC LIGAND- METAL EQUILIBRIUM CONSTANTS Introduction Liquid-liquid extraction has proven to be an extremely valuable technique in chemical analysis. The movement of species across the interface, Of the immisc- ible solvent pair, may be as uncharged molecules or as ion pairs. Solvent extraction with different complexing macrocyclic ligands involves transfer, into the organic phase, as ion pairs (M+Pic-). This method was one Of the initial techniques employed to qualitatively compare the complexing abilities of these ligands. If an aqueous solution of a metal cation, picrate anion, and an excess quantity of hydroxide ion (as com- pared to picrate) is shaken together and equilibrated with an immiscible organic solvent containing the macro- cyclic ligand, the extent tO which the picrate is extracted may be taken as a measure Of metal-ligand complexation. Although the magnitude of extraction 61 62 is not solely dependent upon the complexing equilibrium, it is useful in ranking the complex strengths of dif- ferent ligands with a cation. Other system properties which may exert an effect upon the extraction include solubilities and partition coefficient of the various uncomplexed and complexed species. Several different extraction systems will be pre- sented here and with so many components to continuously note the following shorthand will be utilized for easy reference: organic + macrocyclic / picric metal salt solvent ligand acid or hydroxide + water ORGANIC PHASE AQUEOUS PHASE Distribution Equilibria Analysis of the extraction equilibrium constant for the metal-ligand complexation reaction involves a number of constituent equilibrium which must be con- sidered. Frensdorff (22) discussed the equilibria con- trolling the extraction processes and developed an expression for the calculation of the extraction equil— ibrium constant. The reaction which describes the total extraction process, for which the equilibrium constant is desired, is: 63 + . - + . - M aq + Pic aq + Lorg ::: ML P1corg + . - Kg=: [ML Plc ]Org (l) [n+1aqtpic'1aqlL10rg where ML+Pic- is the ion pair, cation-ligand complex (ML+) with the associated picrate (Pic-) anion, formed in the organic phase. Since the organic solvents amend- able to the extraction process are moderately polar and the concentrations of the components are quite low, the ion pairs in the organic phase are partially dissociated, MLP1corg __1 ML org + P1c org K = [ML+]org[Pic:lorg (2) d [MLP1c]org In addition one should consider the partition Of the ligand between the two phases _ L a Lorg 3:: Laq Pe I (L133; (3) as well as the formation of the complexes in the aqueous phase + + + [ML ]aq M + L ——> ML K = 4 aq aq <—-' aq s [M+] [L] ( ) aq aq Frensdorff determined that the concentrations of hydroxide ion, uncomplexed metal cation, and picrate 64 in the organic phase can be neglected in equilibrium calculations. He took into consideration the activity coefficient of the aqueous metal ion; however, all other activity corrections were neglected, since the other components were either neutral species or ions at very low concentrations. It is implicit in the mathematical treatment of the data that equal volumes of each phase be used in the extraction procedure, although the concen- trations Of the phase components may vary widely. It is also assumed that the ion pair complex involved is in a 1:1:1 stoichiometry. Considering the various equil- ibrium reactions occurring within the extraction system, the following equation may be used to determine the value of extraction equilibrium constant (Ke): 2 5 [23 + Kd - (Kd + 4KdE) ] [1 + Pe + PeKSf(CM — 3)] K = (5) e 2 f(cM - E) (cPic - E)(CL - E) M’ Pic’ and CL are the total concentrations of the metal ion, the picrate, and the ligand, respectively; where C C f is the single ion activity coefficient of the aqueous cation, and E is the concentration of the picrate ion extracted into the organic phase. The derivation of this expression may be found in Appendix A. Before any attempt to determine the overall extraction equilibrium constant may be made, due to the complexity of component behavior within the extraction 65 system, the complex stability constant in water (Ks) as well as the partition coefficient (Pe) must be known. The values of CM’ C , and CL are Of course known, that Pic of f is Obtained from the Debeye—Huckel equation, and that of E from absorbance measurements. Characterization Of Picric Acid Ultraviolet Absorption The picric acid (2,4,6—trinitrOphenOl) anion was utilized in these extraction studies as a monitor of metal-ligand complexation. Extraction efficiency is high since the anion is large and easily polarized. Because the aqueous absorbance change of picrate, before and after extraction, was the basis Of this technique, a clear view Of picrate absorption characteristics was necessary. Figure 3 is a Beer's law plot for aqueous solutions of picric acid of maximum absorbance at 356 nm. The calculated molar absorptivity is (1.44 t 0.01) x 104 M-lcm-l. Several spectra were run after different time periods to check the stability of picric acid solutions. At a given concentration neither the spectral shape nor the absorbance changed. The extractions had to be performed in a basic system, consequently it was Of interest to explore spec- tral changes of aqueous picric acid solutions upon addition of a strong acid or a base. Since picric acid has a pK value of 0.29, at 25°C, it should be largely ABSORBANCE O {o A = 356 66 A A A A A A A I A j A A 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.010.011.012.0 PICRIC ACID CONCENTRATION, M (x105) Fig. 3. Beers law plot for picric acid in water, nm. __ 67 dissociated in aqueous solutions. In fact the spectra showed only a peak at 356 nm. It would be expected that the addition Of a strong acid to an aqueous solution of picric acid would shift the equilibrium towards the for- mation of undissociated acid. A solution, 1.6M in hydro- chloric acid, produced a spectrum in which the peak was shifted from 356 nm to 341 nm. This peak corresponds to undissociated picric acid (105). Several spectra were taken of solutions with addition of strong base in which there appeared to be no change in the spectra. Later studies, at higher metal hydroxide concentrations, did cause a shift of the wavelength maxima, as great as 35 nm, to higher values, with a corresponding increase in peak absorbance for a given picric acid concentration. This phenomenon was reported by Abe (106). He also found the absorbancy Of the picrate peak slightly increased at low concentrations of sodium hydroxide up to approximately 0.4g and then decreased at higher concentrations. Identi— cal results were Obtained in this study. Abe attributed the new band to the formation of a complex between the picrate anion and hydroxide ion. Possible Sources of Error in Extraction Procedures The problems in Obtaining accurate extraction data arise from various sources including: 68 (a) Extraction Of picrate anion into the organic phase due to factors other than ion pair for- mation; (b) The removal Of metal ions (Na+, K+) by the ligand from the surface of the glassware; (c) Dissolution of some of the organic solvent in the aqueous phase; (d) Measurements obtained before the system has reached equilibrium. Several experiments were initially performed to determine the magnitude Of these difficulties and to deve10p the technique tO alleviate them. It was shown by measurement of an extraction system with all compo- nents except the ligand that no picrate, and hence no metal ion, was transferred to the organic phase and; therefore, interaction with the ligand was responsible for the extraction. The possibility of ligand complexation with metal ions originating on container surfaces was negated by a comparative study of two extraction systems: NOH + PicH + H O (a) CH 4 2 2Cl2 / Bu NOH + PicH + H 0 C1 4 2 (b) CH + DB18C6 / Bu 2 2 Tetrabutylammonium hydroxide was used as the hydroxide ion contributor because it is a reagent for ion pair extrac- tions (107). Some extraction was desirable to enhance 69 detection of differences between the two systems. The first system was used as a blank, while the second system would show an enhanced extraction if any metal ion-ligand complexation occurred; however, no change in absorbance was Observed. Pure water was used as a reference in absorbance measurements; however, the aqueous phase after extraction could contain traces of the organic solvent. A UV spectrum of pure water extracted with methylene chloride, referenced to distilled water, was identical to a pure water versus pure water spectrum. Identical samples Of the extraction system CH C1 0 2 2 + 18C6 / KOH + PicH + H 2 were prepared and measured as a function of time, that the two phases were together, to demonstrate the estab- lishment Of equilibrium. Figure 4 contains the results of the study. The plot Of picrate concentration in the organic phase versus time shows that no trends occurred as a function of time implying that the differences are due to experimental error and that equilibrium has been established by two minutes. 70 .meu mo cofluocsm 6 mm OUAHOHco mcwa>£uma one“ oouomuuxm :ofi OOMMOMQ mo coflumuucmocoo m5» m0 coflumcfleumumo Eswunflaflsvm cofiuomuuxm .v .mwm AmmuDCHEV mzHB oa a m n m m e m H _ p p _ _ — _ _ _ a . _ . . . . a . r l.m.m .0 I. .Im.m . II j.h.m .0 I_m.m l.m.m (SOIX) fi 'NOIIVHINHONOO NOI HIVHOId 71 Ligand Partition Coefficient Studies Dibenzo-18C6: Spectrophoto- metric Determination The studies Of partitioning Of DB18C6 between equal volumes of methylene chloride and water were per- formed using ultraviolet spectroscopy. Dibenzo-18C6, in methylene chloride, absorbs at 277 nm. The DB18C6 solutions' absorbances were measured both before and following extraction with distilled water. The mean partition coefficient was found to be 0.09 i 0.07. In this particular solvent system the value Of the partition coefficient is not very significant in the extraction equilibrium constant calculation, therefore, the large standard deviation was tolerable. 18C6 and 1,19-Dithia-18C6: Gravimetric DeterminatIOn Since both 18C6 and DT18C6 cannot be studied spec- trophotometrically, a gravimetric determination of each was performed. Several solutions Of varying concen- tration, in methylene chloride, were prepared and a given volume, prior to and after the extraction process, were delivered into pre-weighed vials. Following evapor- ation to dryness under vacuum, the difference between each blank and the corresponding organic phase after extraction was taken as the weight of ligand which was transferred into the water. The partition coefficient Obtained for 18C6 was 0.23 1 0.04. In the case Of DT18C6, 72 no partitioning was detected, which indicates that the upper limit of Pe is 0.01. As shown later, the calcu— lations of Ke are relatively insensitive to the exact values of Pe. The literature values for the complex stability constants are presented in Table 6. Table 6. Stability Constants (Ks) in Water for the Metal- Ligand Complexes. Mafiiggygic Miggl Log Ks Reference 18C6 K 2.03 12 Cs 0.99 12 T1 2.27 12 DBlBC6 K 1.67 92 DT18C6 Tl 0.93 23 Extraction Systems The picrate extraction data needed to calculate Ke for the systems studied is derived from absorbance measurements Of both the total picrate and that remaining in the aqueous phase after extraction. The concentration Of picrate extracted into the organic phase (B) was found by difference between the initial picrate concentration aq) . (C ) and the unextracted concentration ([Pic] Pic 73 Extraction Studies Utilizing the DB18C6 Ligand The complexing ability of DBlBC6 macrocyclic ligand with the potassium ion was studied for the system CH Cl 2 2 + DB18C6 / KOH + PiCH + H 20 TO enable the calculation Of an accurate extraction equilibrium constant, it is mandatory that absorbance data be collected at varying concentrations of all com- ponents within the system. Table 7 contains the results Of these studies. Figure 5 contains the composite graph Of each set of data points. The results are presented as percent picrate extracted, (E x 100)/CPiC, as a function of the ligand to picrate (CPic) mole ratio. For a particular set (each curve represents one set) the metal and picrate concentrations are constant. Aliquots are extracted with the organic phase containing different ligand concentrations. Since the picrate is extracted in a 1:1 relationship with the amount of metal complex- ation, the graphs actually relate the concentration of the complex formed to the concentration of ligand intro- duced in the organic phase prior to extraction. The curves prove helpful in allowing visual comparison of the effect of increasing potassium and DBlBC6 concen- trations and that the amount Of picrate extracted into the organic phase, relative to the initial picrate 74 m.om vm.m hm.m th.o mm.m H.vm mm.v mm.v mom.o mm.v m.mv hv.v wh.v mmm.o mh.m n.mm m~.m mm.m mmm.o mm.a m.¢m mm.m mm.w moo.a mm.o I . I can I . I x :mucH x mm a I 666 Smao e u mac m.mH o>.H am.m mmm.a mm.H m.mH oa.a mm.m Hmm.H mm.H v.HH ma.a mm.m omv.a no.a m.h mm.o mm.oa mmv.a Hm.o m.v wm.o mm.oa mmm.a Hm.o ~.m mm.o mm.oa mmm.~ mH.c . I Own I . I M EVIOH x HNH H I you SOHO o I moo m.oa ow.o Ho.m Hmh.o bh.m m.m om.o wo.m m~>.o me.~ v.m 5v.o vH.m ovu.o em.H m.o mm.o mm.m mmn.o HH.H m.v ¢~.o hm.m m>>.o mm.o H.m nH.o vv.m mm>.o mm.o I . I can I . I x Emloa x Hm m I woo Smoo o I moo I I . can A OOHOMHUXM Amoaxv Z AmOHXV Z COMWWMMNHWmMMHwfl A UMU\DHOUV m»MH0flm unmoumm mono—UnaH OMHOAmH wumuoflm msomsqfl Oaumm Odo: .ONm + moan + mom \ cOmHma + NHONmO smumsm we» mo mcsum coauomuuxm ssaunaaasvm .a manna 75 Om OHMHOAA msomsvd m.>n mm.m mm.~ nmm.o mm.H >.o> Hm.> hm.m Hhv.o N~.H m.om ow.m mm.v omm.o mm.o m.mv oa.m mo.o mnm.o Hm.o o.ma we.H mh.m oov.H ma.o I . I Dad I . I x zeucfi x mHA H u emu seam c n mac m.an mA.m me.m mvm.o mv.m m.mm on.v mm.m mmm.o m~.m o.mv mm.v ov.e mmm.o mm.m n.mv mo.m em.¢ Hah.o vm.a o.mm m>.N mm.m vvm.o mm.a II o .I Dang I. o I +v~ zmloa x mm m I Ono Eomo o I Ono >.mm v>.h nm.m mmv.o em.v m.mo mo.h mo.v mmm.o Hm.m m.me om.m eh.m wmm.o mo.~ m.Hm bv.m no.5 mmo.a mo.H H.ma Ho.m mo.m mom.H mm.o II o In Uflm II o .I VH EVIOH x oaa a I gnu 2mmo o I moo I I can A omuomnpxm A oaxv z Amoaxv z GOHWWMMMHWmMMumd A wmo\muoov ouMHOHm accoumm mHOAOHmA AOAOH owumm OHOZ .owscflucoo h OHQMB 76 1.00.ch 90.04) 80.04 I} cK+ = 0.849_M_ C + = 0.090M K .— o . 3.1 70°01 CK... = 0.086M O 2 ‘I is m 60.00 I'd a a I 3 CK... = 0.04324— 9 50.0" a Z 8 " a: m m 40.0" 30.0“ V 20 003' 4)» CK+ = 0.0102! 10.0... CK+ = 0.005fl p, '/ 1.0 2.0 3.0 4.0 5.0 6.0 org a MOLE RATIO (CL /CPIC) Fig. 5. Extraction of potassium picrate into methy- lene chloride by DBlBC6 at various cation and ligand con- centrations. 77 . . + concentratlon, 1ncreased at a faster rate as the K concentration was raised. Extraction Studies Utilizing the 18C6 Ligand A number Of studies using 18C6 as the complexing macrocycle were performed with potassium, cesium, and thallium cations. Although the stability of these metal— crown complexes have been well characterized, it was of interest to find the extraction equilibrium constants in methylene chloride, as well as provide a basis of comparison for the later studies using a sulfur-substi— tuted macrocycle. The system components used for the potassium and cesium extractions were, CH Cl + 18C6 / MOH + PicH + H O 2 2 2 however, the thallium studies were conducted using a thallium salt, either acetate or thiocynate, and lithium hydroxide, rather than T10H. A few systems were measured in which the hydroxide source was LiOH, rather than from that of the metal Of interest. Using LiOH presents the danger Of the lithium ion and the thallium ion competing for complex formation with the ligand. However, the extensive work gone into the prOperty Of cation selectivity demonstrated by the crown ligands has shown that 18C6 does not accommodate 78 the lithium because of its small size and high degree of solvation by the solvent (for example 24). In this work, it was found that if the lithium concentration was maintained below approximately 0.004M in the aqueous phase, no extraction could be attributed to lithium ion— ligand complexation. The picrate absorbance data and calculated results of the extent of complex formation are shown in Tables 8- 10. Figures 6—8 show the corresponding curves Of percent picrate extracted versus mole ratio of ligand to initial picrate. The same relationship described for the DB18C6°K+ data are Observed with these as well; that is, greater extraction is Observed as the ligand and/or metal concentration is increased. In all systems the curves level off at high mole ratios, as maximum complexation is approached for a given metal concentration. A comparison Of the 18C6-K+ extraction curves with those Of DBlBCG'K+ demonstrates the higher efficiency of the 18C6 ligand in extracting the metal cation and, hence, implying a stronger complex formed in this system. Extraction studies have been conducted in a number Of organic solvents. Low dielectric constant solvents are used since uncomplexed alkali metal picrates are scarcely extracted into the solvents and their non- ionizing property prohibits the complex ion pair from dissociating to any great extent. Pedersen noted that 79 o.mm vm.m mh.o voa.o om.m m.mm mo.m no.0 mmH.o oa.m m.¢m em.fi mv.a vo~.o mv.m m.m> ma.h Hm.a mh~.o mm.m m.mm m~.m mm.m hoe.o ow.H H.mv om.e m>.¢ mmm.o mm.o H.mm mv.m Hm.m mom.o mm.o m.om vw.a mm.h ovo.H m~.o I . I Dang I . I +M ZmIOH x we m I Ono Emoo o I gnu m.mm mm.h Hm.a mmH.o mm.m m.~m em.h mm.H mmm.o mo.m H.om ma.> hn.H mmm.o mv.¢ v.>> mm.m Ho.m omm.o ow.m b.mh mm.m vm.m 5mm.o mm.m o.wm mo.m mm.~ oav.o mm.~ o.mv mm.v vm.v emo.o mm.a v.va n~.H mo.h mmo.a mm.o I o I Uflm I. o I +VH EmIoa x om m I gnu Emoo o I goo I I own A as. a 1.2.. . .owwmwwha. 1:22... mumuoam usmouom mHOAOflmL OMAOHQH wumuowm msomswd owumm mac: .Omm + mead + mom \ cOmH + maOmmO amummm can mo mcsum ceauomuuxm ssflunaaasvm .m manta 80 A.mm hv.m mo.o omo.o mo.m ~.om Am.m mm.o mNA.o mv.m m.wm mm.h m~.A omA.o vm.~ A.hh mo.> mo.~ oom.o mm.A m.vm hm.v ma.v emm.o mm.o m.mv hm.m mA.m mm>.m mm.o m.m~ mo.~ mo.h on.A m~.o I o I UHAH I o I +V~ EmIOA x 0A m I Ono Zomo o I you m.mm mA.m mm.o omo.o ~v.w A.Am oo.m mh.o MAA.o mA.m m.mm mm.h om.o omA.o mv.v o.mm m>.> mo.A AmA.o mm.m m.mm Am.h hN.A mmA.o >~.m o.Am AA.> hm.A ovm.o wm.m ~.~m m¢.m mm.m mhv.o mm.A m.mA mw.A NA.h mmo.A mm.o I o I “VA-Hm I o I +VH ZmIOA x we m I vmo Zon o I Umu I I 0Am A omuomuuxm AmOAxv S Amoaxv z GOAWWMMMHWmMMumd vmo\muoov mpmuowm ucmouwm mHOAOAmA OMAOAAH OAMHOAA msomsvd oAumm onz .OOSCAUGOU .m OAAMB 81 h.mm mo.m mm.o NvH.o ov.N o.mm mm.h vm.H mmm.o om.H N.om mn.m mw.m mmm.o Hm.o m.hv hm.¢ mo.m hmh.o mm.o N.mN N¢.N mN.b mvo.H mN.o I . I GHQ I . I +M ZmIcA x we a I 660 ammo c I 660 I I cam a cmuomuuxm A OAxV z A OAxc z :ofluomnuxm umuma A 963300v mumnoam unmoumm m Foam. m Aoama wocmnHOmnd Oaumm OAOE . who . Wm . oumuowm msomsvd . .ccscaucoo .m mAncs 82 m.ow me.m mm.m mom.o on.m m.mm >>.e mA.v mom.o mA.> m.ev mm.m oo.m omn.o mm.v m.om Am.m em.m MAm.o hm.m m.m~ no.m mm.m mom.o mm.m m.mA mm.A hm.h hvo.A mA.A o.mA mv.A mm.h mmo.A mm.o m.m mm.o 0A.m nmA.A mm.o I o I Uflm I o I +mo ZmIOA x mm m I Ono Emoo o I Ono e.om mq.v ~¢.e >mo.o mh.m n.me mm.m mo.m e~>.o AN.> >.em mo.m mm.m mmm.o hm.v o.mm mm.m mm.w NAm.o mm.m m.mm mo.m mm.m Amm.o ov.m w.eA om.A Am.h bmo.A mA.A o.AA mm.o mm.h va.A om.o A.> mo.o m~.m mmA.A mm.o I . I OE I . I +8 ZmIOA x Am m I Ono Smoo o I vmo OAm A cwuomuuxm AmOAxv z AmOAxc z coflwwmmwmmeMum< A 663300v ounuoam usmoumm muOAOAm_ quOAmA wumuOAm msooowm Owumm onz .on + mOAm + mOmO \ mOmA + NAOmmO Emumxm map mo wosum cowuomsuxm EOAHAAAAOUA .m OAAMB 83 m.m~ cc.n ~m.~ acm.o mA.m H.5c mm.c mA.m «me.o m>.c 5.5m me.m mo.e Amm.c cm.e m.ae mn.e An.e mmc.c cm.m ~.Ae ~m.m oc.m mom.o a~.~ H.5m mm.~ em.c MAc.A AA.A c.- mo.~ ma.> Anc.A am.c m.mA NN.A em.m cmA.A cm.c I . I oAa I . I an SmIcA x mm a I ecu Some c I +660 I I can A cmuomuuxm AmOAxc z AchxV z acawWMMMMWmMMumm A 666\maoov mncaofla accuses mquoAaA OonAeA mnmuoaa meocnce oAucm 6A0: .cwscflusoo .m OAAMB m.hm Hm.m mN.o hmo.o vm.m 84 m.om vm.m mm.o bvo.o mm.o m.vm no.m om.o who.o h~.v m.mm em.m mm.o omo.o vA.m h.mm mv.w mo.A mmA.o OA.~ o.~h mm.m mm.m mmm.o vo.A m.mm hm.m om.m mAm.o mh.o m.mm ov.m 5A.m mmm.o vm.o I HQ I UHm I HQ. 0 I + c o I . o I Seoo o I woo SmIOA x mm m I you Son o I +vmu N.hm mm.m om.o mmo.o mm.m m.om mm.m mm.o hvo.o mm.m m.mm mm.m vv.o moo.o A¢.v m.~m vm.m mh.o «OHIO m~.m m.wm vo.m NN.A mhA.o hA.~ m.mo mm.m mm.~ vAv.o mo.A m.mm ve.m Am.m mvm.o Aw.o A.mm hm.m m~.m mom.o mm.o I . I +HA I . I Oflm I . I HE Zeoo o I gnu SmIOA x mm m I Omu Smoo o I +Umu I I 0AA A it... a 1.... . SHEEN... 1:22... mpmHOAm ucoouom mquOAm_ OQAOAmL wumuuAm mSOOSOA OAumm onz .ONm + moflm + mead + Nemmmer \ cemA + NAONmo scumam can we Acsum ceauocauxm ssuAEAAAscm .cA manta 85 0.00 00.0H 00.0 000.0 00.0 H.00 00.0H 00.0 000.0 00.0 0.00 H0.0H 00.0 000.0 00.0 0.00 mN.0H 00.0 000.0 00.N 0.00 00.0 00.H 00H.0 m0.H .0.00 00.0 H0.N 000.0 N0.0 0.00 00.0 00.0 000.0 00.0 N.00 00.0 00.0 H00.0 00.0 I . I +._...H I . I Guam I. . I +HB 2600 o I 660 zeIcA x emc A I emu gave 0 I ecu I can a coucmuuxm AmOAxc z Amoaxc z ccAWWMMMHMmmmhma ccO\maooc ouMMOAm unmoumm mquOAmH 0mAOAmA mumuowm msomsw< Oflumm 0A0: .wwssAuCOO .OA mAAma 86 100.0» 900% «b 80.0“ 70.0«- C3 [11 E-I E C. / B 1P s 60.0 m d E D .E.’ 50.0» 04 E-I .. Z 8 0.1 m 40.0.. 04 30.0“ I CK+ (g) 'I I .0.052 00.020 20.0.. , > I I0.018 I 00.08 10‘0" A0.005 1.0 2.0 3.0 4.0 5.0 6.0 org aq MOLE RATIO (cL /CPic) Fig. 6. Extraction of potassium picrate into methy- 1ene chloride by 18C6 at various cation and ligand concen- trations. 87 100.0 II 80.0" 7 . In 0 0 c 5+ = 0.020M 60.0‘I CCs+ = 0 .009M 50.0 CCS+ = 0 .0051! PE RCENT P ICRATE EXTRACTED 30.0" 10.0 “ A A ' 1 1.0 2.0 3.0 4.0 5.0 6.07.0 8.0 9.0 10.0 org aq MOLE RATIO (CL /CPic) Fig. 7. Extraction of cesium picrate into methy- lene chloride by 18C6 at various cation and ligand concen- trations. A I A I A I A A V v ' v ‘ v r 1 100.0 90.0‘ 80.0‘ 70.01 60.0‘ 50.0 PERCENT PICRATE EXTRACTED 30.0 -‘ 88 40.0 “ 20.0 10.0 Fig. 8. CT1+ (M) I 0.008 A 0.016 0 0.048 J A 1 4L A v T W 0 f 1.0 2.0 3.0 4.0 5.0 6.0 7.08.0 9.010.0 j A t A J A V I f r org aq MOLE RATIO (CL /cPic) Extraction of thallium picrate into methy- lene chloride by 18C6 at various cation and ligand concen- trations. 89 as the dielectric constant of the organic solvent is lowered the extraction efficiency drops, and the selec- tivity of the ligand increases. The lower efficiency Of extraction when toluene (dielectric constant, e = 2.38) comprised the organic phase, in relation to methylene chloride (8 = 8.93) was shown here. Figure 9 shows the extraction curves for 18C6°K+ systems in which toluene (curve A) and methylene chloride (curve B) were used. At the potassium concentrations shown, one can see the tremendous reduction in the extent Of picrate extraction. A comparative study was conducted for the systems CH C1 2 2 + 18C6 / TlsCN + LiOH + PicH + H O 2 and H O + LiOH + PicH + H O CH C1 2 3 2 2 2 2 + 18C6 / T1C in order to determine the effect Of changing the thallium (I) cation's counterion. The studies were performed using the same series of ligand solutions and the con- centration of thallium ion in the two aqueous phases were similar. The results, in Figure 10, show that, at least, in the case Of TlsCN and T1C2H302, anion does not affect extraction data. a change in the PERCENT PICRATE EXTRACTED 90 90 .0 {P (- METHYLENE CHLORIDE 80 0 0 CK-I- = 0.008_M_ (b B 70.()1- 60.0" 0 50.0‘* " E 5.0I 5:: 40.0 ; 4JJI m A u a 3.0 g TOLUENE 30.0 '6'. 2.0‘ cK+ = 0-010fl S " an 1.0‘ .A 8’ 2000 1p (14 v 4 f T 4 t i 1.0 2.0 3.0 4. " org aq MOLE RATIO (CL /CPic) 10.0" 1b 1.0 2.0 3.0 4.0 5.0 6.0 MOLE RATIO (cgrg/Caq ) (b j V 1P A v in I I. 1 Pic Fig. 9. Comparative potassium picrate extraction efficiency into methylene chloride and toluene. 91 90.0" 80.0.» 70.0.. D 'I' E E 60.0Ib B x v m a 50.04. U H 1- 9.4 S 40.0" m 2 _ m .. CT1+ - 0.005M G4 30.0. I '1‘1c211302 A new 20.0‘ 10,0. 1.0 2.0 3.0 4.0 5.0 org aq MOLE RATIO (cL /CPic) Fig. 10. Comparative extraction of thallium picrate using two different counterions. 92 Extraction Studies Utilizing the 1,10-Dithia-18C6 Ligand Picrate extraction studies were conducted with the metal cations potassium, cesium, and thallium. Tables 11-13 contain the data collected for these sys- tems with the corresponding percent picrate extracted versus ligand to picrate mole ratio graphs in Figures 11- 14. The magnitude Of the extraction, and, hence, metal- ligand complexation, was drastically reduced for potassium and cesium as compared to those studies with 18C6. Much higher concentrations were necessary in order to measure any extraction. Much larger extractions were Obtained when thallium (I) was used as the metal cation. At metal. and ligand concentrations required for cesium and potassium systems to extract less than 20% of the total picrate con- centration, the thallium system was extracting up to 95%, however, not with the same efficiency as in the Tl+-18C6 systems. Extraction Equilibrium Constant Determination Once the extraction data have been collected, calculation of the equilibrium constant for the complex formation becomes the primary goal. A Fortran computer program was written which used the picrate absorbance measured, complex stability constant in water (KS), and the ligand partition coefficient (Pe), to determine the overall extraction equilibrium constant (Ke), according 93 A~.A 00.0 mmA.A N0.AOA 00.A NA.0 00A.A 00.00A 00.0 0m.0 mmA.A 00.mmA m 0 m0.0 0m.0 00~.A mA.00 0 0 00.0 mm.0 0N~.A 0A.m0 0 m mm.0 00.0 AmN.A 0N.Am 0.0 0m.0 00.0 00~.A m~.0m I HA I Dam I M o I + . o I . o " zaIcA x me 0 I vco EmIOA x Am a I 0co zmcm c 060 0.0 00.0 mm.0 00~.A N0.AOA 0.0 «0.0 00.0 0AN.A 00.00A 0 0 m0.0 mm.0 mm~.A 00.mmA 0 0 mm.0 m0.0 mv~.A mA.00 0 0 m0.0 00.0 0m~.A 0A.m0 A 0 0m.0 00.0 00~.A 0N.Am 0.m 0~.0 00.0 m0~.A mm.0m I . I AA I . I OAm I . I & EeIcA x me e I +cco EmIOA x 0A a I ecu Ecmo c I 060 0AA A A... z 1.2.. I. 5.000%... 1.6.8... wumHOAm ucmonmm mquOAmL wwAOAmH wumuOAm m900504 OAumm 0A0: .Omm + meaq + m m m m m . mofla + O m om \ cOmAeo + Au mo swummm 0:» mo mosum coAuocauxm ssAAnAAAswm .AA mAnme 94 A.mA mh.A mh.m Aca.A cc.AmA m.mA A.A c>.m mam.A me.ch m.MA mm.A cm.m cma.A mm.meA >.AA am.A AA.¢A Ame.A 0A.>~A 0.0A -.A mm.OA ena.A co.ch m.m mo.A cm.OA mma.A mm.om m.m mm.c om.cA mAm.A co.~n m.c an.o AA.cA mam.A mm.cm I . I 0AA I . I m EVIOA x maA A I 060 Emma c I 080 m.AA mA.A «5.0 hm~.A oa.cA~ c.AA mA.A m~.m om~.A mo.mmA m.oA no.A mm.m cp~.A mc.meA m.m ea.o ca.m Am~.A no.50A ~.m Am.c mc.m mom.A cc.c~A m.c mc.o -.m mmm.A mo.ch 6.0 mo.o Am.m mmm.A eo.em 0.6 ce.c ea.m ccm.A No.~e I . I 0AA I . I x EmIOA x cm a I ecu zAmA o I 00 O m m N O m + moan + mos \ cOmAea + Au no I I 0Am A cmuomuuxm AmoAxv S AmoAxv 2 ckowmmwmmeMumd A Umu\0nouv mumuowm usmoumm mquOAm_ vaOAm_ mumuowm msomsv< Owumm 0A0: .cmscAucoo .AA mAnma 95 0.00 AA.0 00.0 000.0 00.000 0.00 00.A 00.0 000.0 A0.00A 0.00 00.A 00.0 0A0.A 00.00A A.0A 00.A 00.0 000.A 00.00 0.0A 00.A 00.0 00A.A 00.00 I . Dam I . I M EmIcA x 0m 0 I 060 zAma c I mmo I I Oflm A cwuomuuxm AmoAx0 z AmoAxv z ckowmmmmwmmwum< A 0m0\0nouV mumuowm ucmoumm mquOAmH WMAOAmH mummoam msowswd oflumm 0A0: .cmscflucoo .AA mAQMB 96 m 0 0.0 A0.0 000.A 00.00A 0.0 0.0 00.0 000.A 00.00A 0.0 0.0 00.0 m00.A A0.00A 0.0 0.0 00.0 000.A A0.00 0.0 0.0 00.0 000.A 00.00 0.0 0.0 00.0 A00.A 00.00 0.0 A.m 00.0 0Am.A 00.00 . I HA I . I can I . I mo 0I0A x 00 0 I +000 SmIoA x 00 0 I wmu £000 0 I +vmo 0.0 0.0 00.0 000.A 00.00A 0.0 0.0 00.0 0004A 00.00A 0.0 0.0 00.0 000.A 00.00A 0.0 0.0 00.0 A00.A A0.00 0.0 0.0 00.0 000.A 00.00 0.0 0.0 00.0 000.A 00.A0 0.0 0.0 00.0 000.A 0A.00 o I HA I o I GHQ I o I +mU 0I0A x 00 0 I +000 20I0A x A0 0 I 000 2000 0 I 000 0AA A cmuomuuxm A00Ax0 m AmoAxv m cOAWWMMMMWmmw000 A 003980v mUMHOAm usmoumm mquOAmH vaOAmH mumuowm m90050< OAumm 0A0: m m m m N .ONm + moAq + 0000 + O m 000 \ 000Ae0 + HO mu Ewumxm msu 00 009am GOHuOMHuxm Edflunwaflsvm .NA manMB 0.00 00.0 A0.0 000.0 00.0H 97 m.mm mm.m 0~.0 AAO.o oo.AA 0.0m 00.0 A0.0 000.0 00.0 c.00 Am.0 mo.m Nm0.c om.0 A.A0 00.m m0.m 0A0.c om.m m.mm 0m.m om.0 m00.c oo.0 0.00 00.0 mA.0 Amo.A co.~ I o I +fl-H I o I OH“ I o I +IH-H— 20IoA x mA 0 I wmo EmIOA x mm 0 I 080 2000 o I 000 m.0m 0A.0 No.0 NAo.A A0.cA 0.0m 00.m mm.0 0mo.A 00.0 m.Am cm.m 00.0 cAA.A A0.0 m.0~ om.m oo.0 mmA.A «0.0 0.00 m0.~ 00.0 ANN.A A0.0 0.0m mm.~ 00.0 om~.A m0.m 0.0A A0.A 0m.0 ~0A.A A0.A I . I +HA I . I GHQ I . I +HB :0IOA x mA N I 060 :0I0A x cNA A I 00o z0~o c I 08o 0A0 A 1.8. a 1.2.. A 5.00000... 1.3.... mMMAOAm unmouwm mquOAmA vaOAmH muMHOAm 0:00:00 oAumm 0A0: m m N m m .Omm + mOAA + moAm + o m OAB \ 0O0Aeo + AU no 500000 0:0 00 00:00 :oAuomAuxm ssAAAAAAsmm .mA mAnme 98 m.mm nm.m ov.o mmo.o hm.ova «.mm mm.m mv.o moo.o om.mmH m.¢m nm.m om.o mno.o mv.moa m.mm ha.m ow.o bmo.o mm.¢m H.mm oo.m hh.o HHH.o wv.¢o o.mm on.m no.H mma.o Ho.vv m.am oo.m h>.H mmm.o vm.m~ I . I +Hq I . I can I . I +HB Eona x mv m I wmu :mIOH x mm m I vmo zHom o I vmu m.mm no.m mv.o mmo.o mm.oma >.¢m oo.m om.o who.o nv.mma H.vm vm.m mm.o mmo.o mv.moa o.mm mm.m hm.o hmo.o hm.hm o.am «w.m ow.o vma.o mm.om h.hm mm.m 5H.H mmH.o w~.mv 0.0m oo.n om.H mnm.o Hm.e~ I o I +..H._H o I GHQ I o I ATHB ZvIoa x mv w I wmo mloa x cm m I wwu Saba o I vmu Dam A cmuomuuxm Amoaxv m Amoaxv m :OHWWMMMHWmMMum< A vmo\muouv mumuoflm unwoumm muofiofimfl vmfioflma wumuofim msomsvfi Ofiumm mac: .cwscflucou .mH magma 99 m.mm mo.~a om.o hmo.o mn.hma m.¢m mm.HH on.o Hoa.o mm.mma m.mm hm.HH Hm.o ~HH.o hm.maa m.mm qh.aa «m.o oma.o Hq.mm m.mm mm.HH om.a mmH.o o¢.mo H.nm mo.HH mm.a om~.o mm.hv m.m> mo.oa mw.~ mnm.o ¢~.mm I . I +fla I . I cam I . I Ida zmIOH x mo H I van szoa x mom H I vmu z~¢a o I vmu N m m m o m + moan + mean + ozHa \ oomaea + H0 mo m.ha -.m ¢~.o mmo.o ~¢.HHN m.mm 5H.m mm.o mvo.o ~m.HmH ~.om oa.m mm.o mmo.o -.~ma «.mm mo.m «v.0 ¢mo.o ~m.-H m.em m¢.m vm.o mho.o No.mm H.Nm H>.m mn.o moa.o mq.mo H.0m ¢H.m Hm.H omH.o mm.mm . I +Hq I . I can I . I +Ha vIoa x fie o I vmu ZmIoa x ow m I van zmvm o I vmo I I ofim A Umuomuuxm Amoaxv z Amoaxv z COHWWMMMHWmMMumd A wm0\muoov mumuoflm usmoumm muomoflmH wm_oflm_ mumuoflm mnomsw< oflumm mac: .owscflucoo .MH magma 100 30.0‘I 4b 0 cK+ = 0.45m a m 'I a D g 20.0 *- m ‘_ E g 0 cK+ = 0.22% g I. m ‘_ B = Z CK+ 0.20991 :3 I g 151 C = O. M 9+ 10.0 a K“ - II A . CK+ = 0.096fl A. I. «lb 1.0 2.0 org aq -2 MOLE RATIO (CL /CPic ) x10 Fig. 11. Extraction of potassium picrate into methylene chloride by DT18C6 at various cation and ligand concentrations. 101 7.0L 4» CCS+ = 0.072! II 6.0" v C) a 5.0w 5 I} = E . CC5+ 0.03491 In 4.0" ' E‘ I. g $ I. H Q-I g 3.0“ . ‘3 I m b m 94 2.0" V 1.0" I 0.5 1.0 1.5 2 org aq - MOLE RATIO (CL /CPiC) x10 Fig. 12. Extraction of cesium picrate into methylene chloride by DT18C6 at various cation and ligand concentrations. 102 100.0 90.0‘I 80.0” r 70,0. 60.0‘L V 50.0' 40.04? PERCENT PICRATE EXTRACTED O = .024M CT1+ 0 -— 30.0‘ 20.04 V 10.07 J A A A A V V L A_._ A .1 A U V V v A A ‘ I v 1 T v 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 org aq MOLE RATIO (CL /CPic) Fig. 13. Extraction of thallium picrate into methylene chloride by DT18C6 at cation concentrations of 0.024g and 0.079g with various ligand concentrations. 102 100.0 90.01? 80.0” f 70.0. 60.0" — 0.07% V 50.0‘ 40.00 PERCENT PICRATE EXTRACTED 0 *3 H + I 0.0245 30.0“ 20.0‘ V 10.0” A J A A 4 AL 1 v V V V I 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 org aq MOLE RATIO (CL /CPic) ‘— 4 A J I v *1 v v 1 ‘ Fig. 13. Extraction of thallium picrate into methylene chloride by DT18C6 at cation concentrations of 0.024g and 0.079g with various ligand concentrations. 103 100.0 0 0 90.0.. 'IL O m 0 [-I U 5 I» E-' X an ID [13 E4 g 80.0.. 0 3 I Mi 3 I 0 142 m * II . U I g o 0.171 O-I 0 a 0.201 “I 00.248 7000‘, 4» III I 1.0 2.0 org aq -2 MOLE RATIO (CL /CPic) x10 Fig. 14. High efficiency extractions of thallium picrate into methylene chloride by DT18C6 at various cation and ligand concentrations. 104 to equation 5. The program used is included in Appendix B. The user supplies a numerical range for Kd’ as well as the number of Kd values to be employed within that range. Each Kd value yields an average equilibrium constant for a complete data set at a single metal ion concentration. A plotting subroutine plots the deviation in the calcu— lated average extraction equilibrium constant versus each Kd' The Kd for the system is found at the minimum value in the standard deviation plot. The average Ke calculated using this Kd is the extraction equilibrium constant for the complex. Extraction constants were calculated for all sys- tems studied except for the complexes of K+ and Cs+ with DT18C6, because of the low picrate extractions observed. In order to calculate a reliable constant, data should be available in the range of 10-90% extracted. Table 14 contains the results of these calculations as well as the corresponding Kd values for ion pair dissociation. The complexing strength of 18C6 for the metal cations decreases in the order K+>Tl+>Cs+ while DT18C6 decreases Tl+>K+>Cs+. Although the actual extraction equilibrium constants for (DT18C6'K)+ and (DT18C6'Cs)+ complexes were not calculated, the percent picrate extracted imply this order. The lower the partition coefficient, the more is the hydrophobic character of the ligand. Since K+-ligand complexation for all three macrocycles were studied, an 105 Table 14. Extraction Equilibrium and Complex Dissociation Constants. Complex log Kd log Ke K+'DB18C6 -4.61 i 0.13 4.79 i 0.22 K+'18C6 —4.85 i 0.31 6.38 1 0.06 Cs+-18C6 -4.67 i 0 32 5.56 1 0.41 Tl+-18C6a -2.89 5.81 i 0.05 Tl+°DT18C6b -4.08 i 0.00 3.80 i 0.01 aThe standard deviation could not be obtained for Kd since it was calculated from only a single curve. bThe standard deviation shown for Kd reflects the precision obtained for all curves (Figure 13); however, the author acknowledges that an error value exists below 0.01. 106 interesting comparison may be made. The partition coef- ficients for 18C6, DBlBC6, and DT18C6 reported values were 0.23, 0.09, and DBlBC6> 18C6. The magnitude of the ligand hydrophobic character should have no effect on the stability of the complex, but it will have a slight effect on the value of Ke' It is apparent from this work, as well as of other investi- gators, that less hydrOphobic ligands facilitate the extraction of the alkali metal ions into the organic phase. Since the overall extraction equilibrium is much too complex to deduce the specific molecular processes, the equilibrium may be studied more effectively by analy- sis of three constituent equilibria (22): + + Maq+Laq——=MLaq KS —A P org <—— aq e ML+ + Pic- ——> MLPic P aq a <—— org c Since these three equilibria may be summed to give the original equilibrium expression, the overall extraction equilibrium constant is given by (7) (8) (9) (10) 107 where Pc is the partition coefficient of the complex from separate ions in the aqueous to ion pairs in the organic phase, and Ke’ Pe’ and KS retain their usual meaning. Table 15 contains the equilibrium constants of the component equilibria from the extraction systems presented in this work. The partition coefficient, Pc' was calculated from equation 10. The data presented shows that the log PC values, for the 18C6 complexes of K+ and Cs+ are similar enough not to be the important factor in determining Ke' The tremendous difference in the stability constant, Ks’ appears to be the deter- mining factor. Thallium and potassium possess comparable extraction equilibria constants, however, for the complex (18C6°Tl)+ the Ks value is substantially greater than that of (18C6‘K)+. The less efficient extraction of Tl+ may be attributed to the smaller value of Pc‘ The order of stability of metal complexes has been an important area of study. One theory proposed to explain the preference in binding is based on Pearson's concept of hard-soft acid-base interactions (108). "Hard" species are small and only slightly polarizable, while "soft" species are larger and more polarizable. Pearson suggested that the stability of "acid-base" complexes is determined by the rule that hard acids prefer to bind to hard bases and vice versa. The hard alkali and 108 .wmsam> Umezmmm mum coflpmasono ca poms mucmfloflwmmoo coflufluuwm i bm.~ hOH x m¢.> Hoooo.o mOH x wm.m pm.m 00H x we.» Hooo.o mOH x em.m nm.m mOH x m¢.n Hoo.o MOH x vm.m hm.v vOH x o¢.> Ho.o Hm.m mIOH x H~.m mOH x ma.m Ha momHeo mH.¢ vOH x Hm.H mm.o m.me mIOH x mm.H moH x be.o He Hm.¢ 03 x me.o mm.o n>.m mIOH x um.m mOH x m¢.H mo mm.¢ 00H x mh.m m~.o ~.nOH mIOH x Hv.H mOH x oe.m m oomH pH.e H.OH x hv.H mo.o nn.m¢ mIOH x Ho.~ ¢OH x mH.w m momHmo on 00H on «mm mm ox 0x :oHumo ocmmHg .mmumuoflm mo :ofluomuuxm mnu Hem mucmumcou Eswunflawnvm .ma manna 109 alkaline earth and early transition metal ions preferen- tially bind in the order O>N>S while the softer species (greater number of d electrons), favor ligands S>N>O. This theory considers hard-hard interactions as pri- marily electrostatic in nature, while the soft-soft interactions are due more to covalent character in the bond. The polarizability of d electrons is important for the soft-soft interactions, with metals containing d10 configuration being very soft acids. By substitution of sulfur donor atoms for some, or all, of the oxygen atoms in the macrocyclic ring, the ligand's binding sites are becoming softer, therefore, DT18C6 would be a softer base than 18C6. The potassium ion is considered a hard acid while the thallium ion, with an electron configur- ation of [Xe]4f145d10632, is classified as a soft acid. Based upon this theory, one would expect K+ to complex better with hard base 18C6 than with soft base DT18C6 and Tl+ to complex more strongly with DT18C6 than with 18C6. The results reported in this thesis support the theory in the case of K+ but not Tl+. The extraction equilibrium constant of (DT18C6-Tl)+ is less than (18C6-Tl)+ by a factor of 100, although complexation between DT18C6 and Tl+ is substantially greater than with the alkali metals ions. CHAPTER V INFRARED AND RAMAN STUDIES OF 18C6'ACETONITRILE COMPLEX Introduction The infrared (IR) and Raman studies of 18C6 in acetonitrile (ACN) were conducted to investigate the weak binding interaction between the crown and the solvent. Solid 18C6 is soluble in ACN up to approximately 0.20fl, however, beyond this concentration a precipitate forms which has been shown to be the 18C6°ACN complex. Gokel and coworkers (67) reported that depending on experimental conditions the complex may be of variable stoichiometry. Two possible factors the authors proposed would favor formation of the "host-guest adduct": (a) The substantial size and lack of rigidity of the lB-membered ring might prefer the formation of the solid complex, stabilized by the ACN molecule within the crystal lattice; (b) The presence of so many electronegative atoms in the ring would make it capable of interacting with and further ordering the ACN in the lattice. 110 111 Infrared and Raman spectroscopic methods were chosen to study this interaction since both are sensitive to small changes in intermolecular forces and because ACN molecules produce distinctive bands (C-C and CEN symmetric l and 2253 cm_l, respectively) affected stretches, 928 cm- by molecular interactions (71). One would expect either a shift in the pure solvent bands and/or formation of new bands due to vibrations within the complex. Infrared and Raman Spectral Data Infrared and Raman spectra were obtained for pure samples of ACN and of 18C6 so that the various bands arising in the complex spectrum could be identified as belonging to either the crown or the solvent. Since the complex is known to be weak, one would expect the spectrum to be predominantly the sum of the spectra of the two components. The interaction of 18C6 and ACN could be either through the carbon or nitrogen atom of the solvent, therefore, the solvent's C-C and CEN symmetric stretch regions were particularly investigated. Table 16 contains the IR and Raman band frequen- cies of pure 18C6, of pure ACN, and of the crown-solvent complex. The Raman spectrum was obtained from the solid complex. The infrared spectrum of the solid, however, could not be obtained because the physical stress applied when preparing a sample in nujol or as a KBr pellet 112 Table 16. Infrared and Raman Band Frequencies of 18C6 and ACN Samples (Regions: IR 600-4000 cm"1 and Raman 840—3010 cm'l). ACN 18C6 18C6'ACN Complex IR Raman IR Raman IR Raman 750 749 832 860 867 841 860 902 894 912 921 921 944 938 959 945 985 988 1039 1032 1043 1047 1042 1064 1068 1092 1081 1112 1117 1106 1130 1134 1125 1138 1138 1159 1238 1242 1252 1260 1254 1278 1287 1274 1298 1292 1337 1351 1357 1367 1374 1376 1375 1375 1419 1391 1429 1415 1409 1449 1447 1455 1475 1477 1479 1495 I 2195 2250 2254 2252 2243 2294 2294 2290 2291 2414 2411 2414 2622 2626 2682 2735 2727 2769 2783 2844 2884 2939 2943 2942 2996 3003 2993 3005 3154 3168 113 caused the complex to decompose. The spectrum of 18C6 in ACN solution was, therefore, recorded. Considering small variations in the instrument's frequency calibration for the different spectra, the bands of the complex (in both solution and as the solid) do appear to be a combi- nation of the frequencies of the individual components. Raman spectra of ACN, of 18C6, and of 18C6-ACN complex in the C-C and CEN symmetric stretch regions of the ACN molecule were separately studied under identical instrumental settings, at a low scan rate (10 cm-lmin-l) in order to help Optimize the electronic response time. The solvent's C-C stretch band appeared at 921 cm.1 in pure ACN and in the complex, at 920 cm-1. In this region pure 18C6 possessed a very weak band at 938 cm"1 and a moderate sized band at 989 cm-1, however, the com- plex only showed a band at 943 cm-1, of intensity com- parable to the 989 cm.1 band. In the region of higher frequency three bands 1 1 appeared in pure ACN, 2294 cm- , 2254 cm- (CEN symmetric stretch) and 2205 cm-1. The Raman spectrum of 18C6 showed no bands in this region, however, the solid 18C6'ACN complex produced bands at 2290 cm.1 and 2244 cm-1. The disappearance of the band at 2205 cm.1 could be attrib- uted to a lower concentration of the ACN in the complex. It is difficult to determine the significance of these small shifts, but they were repeatedly observed. 114 These studies do show the sensitivity of IR and Raman spectrosc0py in detecting the existence of the crown-solvent interaction and indicate that a more thorough study, using these spectrosc0pic techniques, could provide for better characterization of the nature of the binding and possibly lead to identification of the binding site. CHAPTER VI DETERMINATION OF CESIUM- AND THALLIUM-DT18C6 COMPLEX STABILITY CONSTANTS IN NONAQUEOUS SOLVENTS BY NMR Introduction A second technique employed to help characterize the complexing abilities of DT18C6 was nuclear magnetic resonance. Specifically, studies of the DT18C6 complex- ing with cesium and thallium ions were performed using 133 205 Cs and T1 NMR. A shift of the resonant frequency from that of free metal ions is caused by the disturbance of its electrons. Since the 133Cs and 205 T1 chemical shifts are dominated by the paramagnetic term, if the electron density around the nuclei is increased by an interaction with an anion, a ligand, or solvent molecules, a downfield (paramagnetic) chemical shift will occur. However, if the electron density is decreased then the shift is upfield (diamagnetic). Polarization of the cesium and thallium ions are such that chemical shifts may be quite large (especially for the thallium ion) 130 ppm and 2600 ppm, respectively. The wide range of 115 116 chemical shifts provides for high sensitivity to the variations in the immediate environment of the nuclei, whether they be due to ionic interactions, complexation, or both, and, therefore, metal NMR should be a very useful technique for the study of DT18C6 complexes. Concentration Formation Constants from NMR Chemical Shift Measurements A sample prepared for studying a metal ion-ligand complex in a particular solvent will consist of two forms of metal ions (assuming that the ion is not totally com— plexed), the free metal ion and the complexed metal ion, which will give rise to resonance at two different fre— quencies. If the exchange between the free solvated metal ion and the complex is fast on the NMR time scale a single resonance is observed (population-average resonance) and the observed chemical shift (ppm) is given by the expres- sion (109). éobs = Xféf + Xc6c (l) where Xf and Xc are the mole fractions of the free cation and complexed cation and 6f and 6c are the chemical shifts of the free and complexed cations, respectively. Given that X =———- (2) 117 M M . Cf and Ctotal are the concentrations of free and total metal ion, the following expression may be written 0 = -————— (0 total f - dc) + 6c (3) Since the ultimate goal is to obtain the formation con- stant, assuming 1:1 stoichiometry + + M+L—>ML (4) the concentration equilibrium constant is expressed as + C f + M L (M )(L) Cf cf (5) where CC is the concentration of the complex and C? and C2 are the concentration of the free metal and the free ligand. The expression can be rewritten in the form M M K = Ctotal — Cf (6) f M L M M Cf(ctotal " Ctotal + Cf) The final expression 6 = [(K CM - K CL - l) + (KZCL2 obs f total f total ‘ f total + KZCM2 - 2K2CL CM + 2K CL f total f total total f total 0 0 M % f- c + 2KfCtotal + 1) ][2K CM ] + 60 (7) f total 118 can be derived by combining equations 3, 5, and 6. Measurements of dobs were obtained for a series of solutions, possessing a constant metal concentration but a varying concentration of the ligand. Since 0f may be measured, using the metal salt solution with no ligand, equation 7 contains two unknown quantities, 6c and Kf. A nonlinear least—squares program KINFIT was used to solve the expression by supplying the experimental L M total’ total and 6c until the allowed error in curve-fitting was parameters 00b5, 6f, C and C and adjusting Kf obtained. The expressions above are valid when there is no ion pairing in solution, or if the ion pair formation constant is very small compared to Kf. It is possible that both the ligand and anion compete for the metal ion in solution, and the change in the observed chemical shift is the summation of anionic association and com- plexation by the ligand. In this situation the expres- sion for the observed chemical shift must include a term accounting for the ion pairing interaction 0 + x 6 5 (8) éobs = Xf f c c + Xip ip where Xf, XC, and Xip are the mole fractions and 0f, 0c, and sip are the chemical shifts for the free, complexed, and ion paired metal ion, respectively. Derivation of 119 the final expression used to fit the experimental chemical shifts yields L M _ [M] KfC [M] KILC [M16313 sobs _ L 6f + M 6c + M (9) C C (Kf[M]+1) C (Kip[M]+1) with Kip representing the ion pair formation constant. The values of Kip' CM, 6ip' and 6f must be supplied for analysis by KINFIT while Kf and 5c are adjusted until the statistical error has been minimized to the allowed limit. In both situations the data supplied was weighted such that the more accurate measurements would be relied on more heavily in determining the formation constant and limiting chemical shift of the complex. Cesium-133 NMR Study Although 133Cs NMR appears highly suitable for the study of cesium complexation reactions, it is not without difficulties, particularly the limited solubility of the metal salts in nonaqueous solvents. The range of a cesium salt solubility is typically 0.01M, therefore, a greater number of scans must be performed to obtain a reasonable NMR signal (a respectable signal-to-noise ratio). However, the sensitivity provided by the narrow NMR lines helps in resolving the signal from background noise. Another difficulty encountered was the ligand solubility. In propylene carbonate and acetone, it was 120 found that DT18C6 was insoluble, or at most, soluble to a very small extent, impractical for NMR measurements. Cesium-133 chemical shifts were determined as a function of DT18C6 to cesium ion mole ratio for the formation of the (Cs-DT18C6)+ complex in nitromethane (NM), pyridine (py), dimethylformamide (DMF), and aceto- nitrile (ACN). The cesium salt used, for a particular solvent, was chosen for maximum solubility in the solvent of interest, and minimal ion pairing in solution, there- fore, different anions were utilized. The chemical shift results, shown in Table 17, possess an estimated error in the measurements of i0.10 ppm. The plot of chemical shift versus ligand-to- metal mole ratio (Figure 15) shows that (Cs°DT18C6)+ complexation in all four solvents yielded shifts, para- magnetic in nature, resulting from an increase in the electron density about the cesium ion. In all cases, the plots show little curvature, implying very weak metal-ligand interaction. Before any attempt could be made to numerically access the strength of the (Cs‘DT18C6)+ complex, it was of utmost importance to characterize the solute—solute and solute—solvent interactions in solution. These interactions will be dependent upon the solvation number of the ions, the solvating ability of the solvent, the solvent's dielectric constant, and the ion pairing 121 Table 17. Cesium-133 Chemical Shifts of (Cs’DT18C6)+ Complexes in Various Solvents (ambient probe temperature 32-33°C). Nitromethanea Dimethylformamideb CL/CCS+ 6 (ppm) CL/CCS+ 5 (ppm) 0.00 60.27 0.00 0.96 0.88 56.25 0.54 0.72 0.99 56.02 0.87 0.57 1.26 54.70 1.06 0.49 1.71 54.32 1.46 0.34 2.47 49.82 2.69 0.18 2.59 49.90 3.79 -0.28 2.66 49.12 2.89 49.04 . . c . . d Acetonitrile Pyridine CL/CCS+ 6 (ppm) CL/CCS+ 6 (ppm) 0.00 -35.29 0.00 41.36 0.34 -35.84 0.34 39.00 0.79 -36.41 0.93 35.96 1.08 -36.84 1.56 32.18 1.46 -37.40 1.63 31.93 1.72 -37.52 1.93 30.26 2.39 -38.27 2.27 28.52 2.82 -38.70 2.43 27.96 3.70 -39.26 2.86 26.28 4.02 -39.63 3.51 22.81 4.60 -38.83 3.62 22.87 4.50 19.96 a[CsC104] = 0.00445 b _ [CsC104] - 0.0494M c[CssCN] = 0.03455 d[CsBPh4] = 0.01485 122 .coflumEHOH xmamfioo +Hmomaao.mov How oflumu mHoa Hmuwe ou ocmmflH mcmuw> muwflnm Hmoflfiwno MMHIEsHme .mH .mwm H+mooxqoc oHeam age: 0.4 o.m o.~ o.H AZUmmUv 20¢ I1IIII'I1I L.O.ONI «u nu : a m A.o.o.nl flu AVOHUmUv .mza W I I A. .d 7w ..0.0H : v 9 H nmmmov >0 1 .d d m. AVOaUmUV £2 1 10. oo 123 capabilities of the cation and anion. In solvents of intermediate to low values of dielectric constant and of donor ability, the association of the cation and anion would be strong, producing a concentration depen- dence of the chemical shift, implying contact-ion pair formation. Alternatively, in solvents of high dielectric constant and donor ability contact-ion pair formation is not expected. However, of the alkali metal ions, cesium is the most poorly solvated, owing to its large size and low charge density; hence, it is possible that it may undergo contact-ion pair formation, even in polar- solvating solvents (110). The concentration dependence of 133Cs chemical shifts as a function of salt concentration have previously been established by DeWitte et al. (110). Ion pairing was shown to be significant in cesium tetraphenylborate- pyridine solutions. The ion pair formation constant was reported to be 370 i 20 with the ion pair limiting chemical shift (dip) of 40.0 ppm. Cesium thiocynate- acetonitrile solutions were also shown to be concentration dependent, however, to a much lesser extent. Hsu (75) found the ion pair formation constant to be 13.2 i 3.8 and a value of —36.7 ppm as the limiting chemical shift of the ion pair. Cesium perchlorate solutions of both dimethylformamide and nitromethane were shown to exhibit no ion pairing, by DeWitte (74) and Hsu (75), respectively. 124 The formation constants for the complex reaction in the different solvents are listed in Table 18. The curves (Figure 15) showed no observable break at any mole ratio so the constants were calculated based on a complex stoichiometry of 1:1. It should be noted that the equilibrium constant is expressed in concentration units and, therefore, it cannot be considered a thermo- dynamic constant. The calculated formation constants, for cesium complexation with DT18C6, in each solvent, were of comparable magnitude. The influence of solvent properties on metal-ligand complexation has been estab- lished. Generally one would expect that in solvents of high donicity the complex stability should decrease due to the competition between the ligand and the solvent molecules for the metal ion. The results obtained in this work do not strictly fit this generality. A more valid comparison of the effects of solvent properties on complex formation would be in studies in which the same counterion was used. The complex (CS°DT18C6)+C104' was more stable in NM, a poor donor solvent, than in DMF, a solvent of comparable dielectric constant but larger Gutmann donor number. Complexation was greatest in ACN, which possesses a medium donor number and a high dielec- tric constant; however, the formation constant of the complex in py was surprisingly high. Pyridine possesses a high donor number, yet it does not solvate the cesium 125 Table 18. Complex Formation Constants for Cesium Com- plexes with DT18C6 in Four Nonaqueous Solvents. Limiting Chemical Solvent Counterion log Kf Shift NM c104’ 0.65 1 1.09 -155.8 1 4.77 DMF 0104' 0.56 1 0.29 -2.10 1 1.27 ACN SCN' 0.90 1 0.27 —44.10 1 3.03 py BPh ’ 0.84 1 0.15 -127.8 1 49.8 4 126 ion as well as expected, leading to a stronger cesium- DT18C6 complex. An explanation, based on the hard-soft acid-base theory, contends that since py is a nitrogen donor and, therefore, a soft base, it does not strongly solvate the alkali metal ion, a hard acid (111, 112). In the case of CsClO in NM and DMF the formation constant 4 values should be considered only as estimates of the true concentration constant because of the large error involved in the curve fitting; however, they do provide insight into the complexing abilities of DT18C6 toward the alkali ion, Cs+. Sulfur-substituted crowns have been shown to possess complexation strengths much lower than their polyether analogue. Complexation studies of cesium with 18C6 were performed by Mei (113) in the solvents pyridine, acetone, prOpylene carbonate, acetonitrile, dimethylforma- mide, and dimethyl sulfoxide. Regardless of the solvent used, 1:1 and 2:1 (ligand:metal) complexes were observed. The weakest complex was found in DMSO and possessed formation constants, Kl = 1100 i 100 and K2 = 1.0 i 0.4. Thallium-205 NMR Study 205T1 resonance Chemical shift measurements of the frequency of solutions containing thallium perchlorate and DT18C6 were obtained in DMSO and py, The chemical shift was measured as a function of ligand-to-metal mole ratio. Table 19 contains the results 127 Table 19. Thallium-205 Chemical Shifts for (T1°DT18C6)+ Complexes in Two Nonaqueous Solvents. Pyridinea Dimethyl Sulfoxideb CL/CT1+ 6 (ppm) CL/CT1+ 6 (ppm) 0.00 -596.72 0.00 -317.92 0.43 -540.87 0.84 -316.07 0.80 -493.22 2.01 -3l4.49 1.06 -467.98 2.82 -312.72 1.34 -434.70 3.18 -315.19 1.81 -385.10 3.80 -314.31 1.98 -369.23 4.67 ~316.16 2.27 -339.47 6.94 -312.37 2.52 -322.50 7.87 -313.25 2.79 -299.35 9.03 -312.81 3.30 -257.69 3.63 -231.46 a[T1C104] = 0.0574M (probe temperature = 34°C) b[T1C104] 0.0098M (probe temperature = 29°C) 128 of the two complexation studies, which are graphically presented in Figure 16. In DMSO, a solvent of high donicity and rela- tively high dielectric constant, no evidence of complex- ation was observed, implying that the solvent strongly solvates the T1 (I) ion blocking any ion-ligand inter- action. Srivanavit and coworkers (95) have studied the complexation of T1ClO4 and 18C6 in DMSO, and obtained information indicative of a strong complex. The chemical shift versus mole ratio plot possessed an extrapolated break at 1:1 stoichiometry (as found in several solvents) and obtained the limiting chemical shift of the complex at a mole ratio of four. The complex formation constant was found to be greater than 103, implying that 18C6 could successfully compete with the solvent, therefore, ligand complexation rather than metal ion solvation resulted, in contrast to the results of DT18C6. Figure 16 also contains the plot of the results for T1C1O4 and DT18C6 in py. The chemical shift is up— field (diamagnetic) and extremely large, spanning from -600 ppm for the free solvated Tl (I) ion to -230 ppm at a mole ratio 3.6. A very slight curvature in this plot implies formation of a weak (T1‘DT18C6)+ complex. Just as noted with the cesium complexation in py, it is evi- dent that the solvent does not solvate the T1 (I) ion as effectively as DMSO even though it possesses greater 129 .eoHumeuom memeou +xoomHea.Hev How oHumh 0Hos H8008 on eemmHH msmhm> mumHsm H80H80go momIesHHHmne .mH .mHm H+moo\HoC oHeam mac: o.m o.m 0.5 o.® o.m o.v o.m o.m otn mwml mmmI vomI 3 «:7 H 3 W I D ova W S H movI I 3 Tu on? o. w, vva m leoHoHec omze lll’ll.I AvoHoHeC we N M N I 130 donor ability. However, in the case of the T1 (I) ion, which is classified as a soft acid, the previous expla— nation based upon the hard-soft acid-base theory does not seem appropriate. Computer analysis of the chemical shift data determined the formation constant of the complex, in py, to be log K 0.34 i 0.02 with a limiting chemical shift of 632.4 1 44.3 ppm. The constant was calculated based on a 1:1 complex stoichiometry with no ion pair consider- ations. However, Briggs and Hinton's (90) NMR studies, of thallium salts, have shown that ion pairing is sig- nificant in py. The authors referenced their chemical shifts to infinite dilution aqueous T1+N03-. For the system TlClO4 in py they reported the chemical shift of the free solvated Tl (I) ion and the limiting chemical shift of the ion pair, as -781 ppm and -660 ppm, respec- tively. These workers did not report any ion pair for- mation constant (Kip)’ however, from their data, our KINFIT program was utilized to calculate Kip' An attempt to take into consideration activity effects did not result in an acceptable fitting of the data, however, the concentration ion pair formation constant fitting was very good. For TlClO4 in py, Kip = 715 t 69 indicative of a significant amount of cation-anion interaction. Both ligand and anion compete for the thallium ion in solution and the change in the observed chemical shift (Table 19), 131 for this system, is partly due to changes in the degree of anionic association with the thallium ion upon com- plexation with DT18C6. Re-evaluation of the complex formation constant, taking into consideration the extent of ion pairing, resulted in a more realistic value, Kf — 20.28 + 9.10 with the limiting chemical shift of the complex of 1125 i 444 ppm. CHAPTER VII COMMENTARY The studies presented in this thesis demonstrate the weak complexing strength of DT18C6 for the alkali metal ions, potassium and cesium. The substitution of sulfur atoms for the oxygen atoms in the 18C6 macrocyclic ligand is responsible for the reduction in binding com- pared to the polyether. Although DT18C6 prefers binding with the thallium (I) ion, rather than the alkali metal ions studied, a decrease in the complex strength is still observed when the sulfur-substituted ligand is used, instead of 18C6. The two techniques employed to determine the binding preference of DT18C6, distribution equilibria studies and metal NMR measurements, led to the same order of complex strengths (T1'DT13C5)+ > (K'DT18C6)+ > (CS'DT18C6)+ regardless of the solvent the complexation reaction took place in. It would be interesting to determine, by these techniques, the complexing capabilities of nitrogen substituted 18C6 ligands for these same metals. 132 133 Preliminary 205 T1 NMR studies, in our laboratories, indicate that 1,10-diaza-18C6-metal ion complexes are much stronger than those using the DT18C6 ligand. A more thorough investigation of these ligands as well as the monosubstituted sulfur and nitrogen 18C6 molecules are necessary in order to explain the binding preferences. Pearson's (108) hard-soft acid-base theory does not appear to fully explain the complex strengths observed. In accord with the theory, the hard acid, alkali metal ion, prefers to bind to the ligand possessing donor atoms in the order O>N>S, which is the order of hard to soft donor atoms, but the soft acid, thallium metal ion, shows the same preference, O>N>S, contrary to the theory. However, the hard-soft acid-base theory does apply in + . one sense, for example K 18 complexed more strongly by 18C6 than is 11*, but less strongly by DT18C6. APPENDICES APPENDIX A EXTRACTION EQUILIBRIUM CONSTANT DETERMINATION FROM DISTRIBUTION EQUILIBRIA STUDIES APPENDIX A EXTRACTION EQUILIBRIUM CONSTANT DETERMINATION FROM DISTRIBUTION EQUILIBRIA STUDIES The determination of the overall extraction equil- ibrium constant (Ke) for the complexation of a metal ion and macrocyclic ligand in the organic phase was discussed in Chapter IV. The derivation of the final expression used to calculate Ke follows. Pic’ and CL are the total concentrations of the metal ion, picrate ion, and the ligand, respec- If CM, C tively, then for a univalent metal ion the following relations may be defined: + [M I = f(CM - E) (1) [Pic] = cPic - E (2) cL = [L]org + Pe[L]org + PeKSf(CM - E) + E (3) where the brackets ([]) represent equilibrium concentra— tions of the particular species, f is the single ion activity coefficient of the aqueous cation, and E is the concentration of the picrate ion extracted into the 134 135 organic phase. The equilibrium constants, Pe and Ks’ represent the ligand partitioning between the phases and the complex stability constant in water, respectively. In equation 3 the first and fourth terms represent the free and complexed ligand in the organic phase. The second and third terms account for the uncomplexed ligand and the complexed ligand in the aqueous phase. The equilibrium concentration of the complex ion pair is 2 15 d + 4KdE) ]/ 2 (4) [MLP1c]org = [2E + Kd - (K where Kd is the dissociation of the ion pair in the organic phase. Equation 4 is obtained by considering the equilibrium expression for Kd' the stoichiometry relation E = [MLP1c]org + [Pic]org (5) and the neutrality condition . + [PIC] = [ML ] (6) org org Substitution of equations 1-4 into the equilibrium expression for the overall extraction process [MLPic] K = org <7) [M+]aq[PiCIaq[L] org 136 gives a single expression .. 2 ’2 ._ [2E + K (K + 4KdE) ] [1 + Pe + PeKSf(CM E)] d d e 2f(CM - E)(CPic - E)(CL - E) containing two unknowns, K8 and Kd' APPENDIX B COMPUTER PROGRAM EXTRACT APPENDIX B COMPUTER PROGRAM EXTRACT The Fortran computer program written to determine the extraction equilibrium constant for the metal-ligand complexation studies is presented below. 99069.! rutonc' 1 905000 cxtnacv cxuout. flavour. vaacso c uoIt tan gonunw o¢?o. loot. sncao.looI. HSYEOo RECngn I7I. YAEEIO: 923:3}! Dl'twstnu nxcsaI It? 5 I s0 I I 5 °‘?Oio °lcng;fg Is she ' 0 7 DEV 50 o IVVEISOI. Cotnjgynj. ‘3‘ 's 20 «on r I! NIEGto'tnotbo. vies. '"' D" “9:56 3 l. Tlpf:: I 6 10 c ntao cansrnnvs. 9: Is than» 90:71; N r €3°"?s'°3€‘&1"§.°2:20891 41.: I. 25§°"§"5"5§5""" " " ""‘ I -‘ kiétIs ten [27 an? s nr x «can: 15‘“uaEELO"h "'I” ‘5 1"! LnlEn I S [9 nuun n or [stéos 70 n: IKEN BE IE u «nuIN ann KM 01 t O MAXIUIN km [5 Nu 9 9 OF SE15 I 5 Na vars: Vlkusg r oz. «5. 8 : QEDI:I:E?E.IO :. IS. [95. K “In. KDHSRoa N§TE°¢ LSEY ’0 c 26 g 23°6789E50.3$;:3 cans.» 70 cu HI ‘9‘ c s '. wot ls gut nuance or on . 9,, ”$335) 0" "us mu 8 3 23?}:"éevil°‘.éi‘é.885°éo£ééarai?.‘“‘ "' 3“ I"? I° k 0 O N‘ R ,5 E '2: eggs" 037: onlut. to 0:70 taunts 4211.38“.5‘2033E2m‘“‘ on:vcc .8 25‘szgggblo‘.‘ I’ICTOT ‘1‘! “707 III. CRYOT III. lflS III. . C 5'49? Incorv N1: an! s r In 3° .9 732 J :.4§=* *.§"‘ ..5 °. . C - sou“ . 0.: I «on a oxonluI/NSTED c I:‘3:.§c°&8 run [A a TI 9 H | 4| )5 (9.320 x I 1":915 c 0"! c O 6;: . ;£ 95' o. I gonzcutgtgx|I/tl.o:ontcutotcx11ii 032.0?! onxc I-SOSYIDKI I00 06 OOKIJIOD xc:- .. s : Ia: :s;::.° :I :,;I 'z': =I°4':I.:.°Ic' . - O . EK‘JO“ . .‘I2."‘3: C‘ o E 0. HA I IOTCKI OICII UNI - :xII.xI . ‘00 U”§?:'§ EKIJ0K1°92u0 SUHSOK I 05 c §~CULO E AVEDIGC Inn 0 v A! r .IE".Ii'.'.82:;Iéi""s3I'-° E04..4 3.. ?,*s.I'.'I‘? "L“' "' '°- 5" SEIIS‘IIRQEXI'I3‘¥:$6£I"” . ... | | 50 8°.?‘C,L.'o:o::§gfi’ I 0 V“ o SUDEVIJ'Q QSDIJIO J. IOU§TE3| . O I . \ o 555 éit' gta'SDILI .. Iii) :.I;=:.; 4.. . 766 §gnv'hmt ' ' ° ' an? (S I: re ..v.1III?IQI PISI ‘° .33 5°“ZIIIIE'248‘ . z . I I ' .Iuo"nos :1. £I~.&.':625¢I°" It ' " r"‘°' ' °"" ""’° 137 50980011"! P107 IO IE ?C 2‘ 3t 3‘ as 5' 9A 6% 7° 7% 89 9F V‘cvdd--- DIOD’fiFVJ- 3‘ 7. AI 7’ 7‘ 7| 5: 61 no «I 5‘ AI “I a: 00 On ‘1138 BUBQOUTINE 9L0! _ IHE~SI0N ICIIOII VCISIIIICHAaIIZIIISYN AI2II CONNnNc KI’OIIOOII H700l00'oNPI IR COODI7I.YAPFAOIIAPEAIINSEI , L7 "NhANHHHXQH‘3/ "UM IND ulNIHU" X ‘ND V V‘LU£§ REV “1NI‘00‘70‘? ll‘l‘soigolh “‘N“”oifio|l IllIPOI?Jo‘° u-uuunu—dfloooouo. O d. O. G. .. -—-- Ml—IuidlZO’--’ v<¢4vxouoz 4 z<~<~u-— r::~d--Q‘-d‘d--~-. MW-o-o-o—olu—uuuufl LAIE DANGFS OF I AND V “x-xNAx-Ifl IN :NGEstqu-VIIN i‘t 9L0? Llu 13 I 010 LARGER IHAN RANGES In‘chqu-o. o Mfr: xnlNAIuIN-c.es:DAuGr: val IIVUAIOO. osooAungv VNINIYUlN-fl. ofiouAN §gic999...:-,.,g,,: ENCOEIENIS vsIEO-Ivqu-vulnIlso ES“8lIS? x.v AonA Av I8 10 I0 I xOIJIl INOIJ-|‘°!SYE° :8(K\Iv:1IS:K—IIOYSTFP GLANK ADDA 74 rill-It" 1 CONS AIDA/ll‘ l I t T Iu-IIHIIZAISH-oo-IIoIHII 833§3£228 -u-411urva m- ‘hnao fioo L‘IOIOU *‘Ll'06t0‘c‘lLL.‘O‘NDO“‘:LQLL‘OL'O'C‘LLL.|.‘. 00 YO 7S and "“" °6 D I I v I n I I ‘!;.LL..L‘O'C‘L"0‘"0."“L.LL'IB'OYC'L.'".60 70 7? '0 I I 8‘ A '7‘.” V ““9 7“ I:7*ISVUO:I?I ' 6° '0 ni|.:"I§VIIIl?II l I N” u n. :h-n -fi0 J :— ACYFO L NF oIIAJI 6 AI OII CHAOIAI.I-:oIOOI o I 6 9 ‘H‘ii‘flWh- lh O‘Dollolfl GTIVCI | I CHOPC" N V ‘égzsIo.J§IR-IIuIIInIAI:'nl'° may 8% GI .100 ' nADISAI-‘yunq 0U?! luuf On°IEYF 'HEP HPIIIIIAOEAI.AI F0 AIIIH .IAI.IH1.29I§u---IIIIHII ghltcchOEAI.n0I n HAYIIn .IAIAIHOIonoscIHOI 4!!! Ia TE‘Yl9£6|o°0‘XHINoXCI70‘OI “::O‘C‘“°.O'C‘aO‘O‘C“°“‘ F «AIIIN Ihl‘OIIEI9o3|A/ll| v ItIIADEAI.AIIIMIN.anx.vu u. HA: I lift!" .v'INIHUI x lg ¢.£ . ISI.¢-A1quu I Is c.£Ih 1.11.! IIN IS .OE‘.O’C“O’.“‘“U~ IS ’0‘..070 LIST OF REFERENCES l. 10. ll. 12. 13. 14. LIST OF REFERENCES C. J. Pederson, J. Amer. Chem. Soc., Q9, 7017 (1967). B. Dietrich, J. M. Lehn, and J. P. Sauvage, Tetra- hedron Lett. 2885 (1969). B. Dietrich, J. M. Lehn, J. P. Sauvauge, and J. Blanzat, Tetrahedron, 23, 1629 (1973). H. K. Frensdorff, J. Amer. Chem. Soc., 9g, 600 (1971). C. J. Pedersen, Fed. Proc., 23, 1305 (1968). W. H. Delphin and E. P. Horwitz, Anal. Chem., 59, 843 (1978). E. Blasius and P. G. Maurer, J. Chromatog., léé, 511 (1976). B. E. Jepson and R. A. Schwind, First Symposium on Macrocyclic Compounds, Provo, Utah Aug. 15-17 (1977). A. Knochel and R. D. Wilken, J. Radioanal. Chem., 33, 345 (1976). B. C. Pressman, E. J. Harris, W. S. Jagger, and J. H. Johnson, Proc. Nat. Acad. Sci. U.S., Q9, 1949 (1967). 7 H. K. Wipf, L. A. R. Pioda, Z. Stefanac, and W. Simon, Helv. Chim. Acta, 5}, 377 (1968). R. M. Izatt, R. E. Terry, B. L. Haymore, L. D. Hansen, N. K. Dalley, A. G. Avondett, and J. J. Christensen, J. Amer. Chem. Soc., 9@, 7620 (1976). R. M. Izatt, R. E. Terry, D. P. Nelson, Y. Chan, D. J. Eatough, J. S. Bradshaw, L. D. Hansen, and J. J. Christensen, J. Amer. Chem. Soc., 9@, 7626 (1976) . J. M. Lehn, Acc. Chem. Res., 1}, 49 (1978). 139 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 140 C. J. Pedersen and H. K. Frensdorff, Angew. Chem., Int. Ed., 11, 16 (1972). N. Matsuura, K. Umemoto, Y. Takeda, and A. Sasaki, Bull. Chem. Soc. Jap., 49, 1246 (1976). J. Smid, Angew. Chem., Int. Ed., 11, 112 (1972). M. Bourgoin, K. H. Wong, J. Y. Hui, and J. Smid, J. Amer. Chem. Soc., 93, 3462 (1975). A. I. POpOV, Solute-Solvent Interactions, New York: Marcel Dekker, Inc., J. F. Coetzee and C. D. Ritchie, eds., 1976, pp. 271-230. D. N. Reinhoudt, R. T. Gray, F. DeJong, and C. J. Smit, Tetrahedron, 39, 563 (1977). A. T. Tsatsas, R. W. Stearns, and W. M. Risen, J. Amer. Chem. Soc., 93, 5247 (1972). H. K. Frensdorff, J. Amer. Chem. Soc., 99, 4684 (1971). J. D. Lamb, R. M. Izatt, J. J. Christensen, and D. J. Eatough, Thermodynamics and Kinetics of Cation-Macrocycle Interaction, in press. J. J. Christensen, J. 0. Hill, and R. M. Izatt, Science, L73, 459 (1971). J. M. Lehn, Pure Appl. Chem., 43, 857 (1977). J. M. Lehn and F. Montavon, Helv. Chim. Acta, 63, 67 (1978). C. J. Pedersen, J. Amer. Chem. Soc., 93, 386 (1970). D. Bright and M. R. Truter, Nature, 225, 177 (1970). M. A. Bush and M. R. Truter, Chem. Comm., 1439 (1970). C. J. Pedersen, J. Amer. Chem. Soc., 93, 391 (1970). D. J. Cram, R. C. Helgeson, L. R. Sousa, J. M. Timko, M. Newcomb, P. Moreau, F. DeJong, P. W. Gokel, D. H. Hoffman, L. A. Domeier, S. C. Peacock, K. Madan, and L. Kaplan, Pure Appl. Chem., 43, 327 (1975). E. M. Arnett and T. C. Moriarity, J. Amer. Chem. Soc., 99, 4908 (1971). 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 141 J. R. Dann, P. P. Chiesa, and J. W. Gates, J. Org. Chem., 26, 1991 (1961). L. A. Ochrymowyca, C. P. Mak, and J. D. Michna, J. Org. Chem., 39, 2079 (1974). F. Vogtle and E. Weber, Angew. Chem., 89: 126 (1974). D. St. C. Black and I. A. McLean, Tetrahedron Lett., 3961 (1971). W. Rosen and D. H. Busch, Inorg. Chem., 9, 262 (1970). W. Rosen and D. H. Busch, J. Amer. Chem. Soc., 91, 4694 (1969). M. D. Glick, D. P. Gavel, L. L. Diaddario, and D. B. Rorabacher, Inorg. Chem., 15, 1190 (1976). J. J. Christensen, D. J. Eatough, and R. M. Izatt, Chem. Rev., 73, 351 (1974). N. B. Tucker and E. E. Reid, J. Amer. Chem. Soc., 55, 775 (1933). J. R. Meadow and E. E. Reid, J. Amer. Chem. Soc., 56, 2177 (1934). C. J. Pedersen, J. Org. Chem., 3&, 254 (1971). J. S. Bradshaw, J. Y. K. Hui, B. L. Haymore, J. J. Christensen, and R. M. Izatt, J. Hetero. Chem., 19, 1 (1973). J. S. Bradshaw, J. Y. K. Hui, Y. Chan, B. L. Haymore, R. M. Izatt, and J. J. Christensen, J. Hetero. Chem., 1}, 45 (1974). J. S. Bradshaw and J. Y. K. Hui, J. Hetero. Chem., 11, 649 (1974). J. S. Bradshaw, R. A. Reeder, M. D. Thompson, E. D. Flanders, R. L. Carruth, R. M. Izatt, and J. J. Christensen, J. Org. Chem., 41, 134 (1976). N. K. Dalley, J. S. Smith, S. B. Larson, K. L. Matheson, J. J. Christensen, and R. M. Izatt, Chem. Comm., 83 (1975). B. Metz, D. Moras, and R. Weiss, J. Inorg. Nucl. Chem., 36, 785 (1974). 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 142 M. Jawaid and F. Ingman, Talanta, 23, 91 (1978). S. Kopolow, Z. Machacek, U. Takaki, and J. Smid, J. Macromol. Sci.-Chem., A143), 1015 (1973). S. Kopolow, T. E. Hogen Esch, and J. Smid, Macro- molecules, Q, 133 (973). A. Sadakane, T. Iwachido, and K. Toei, Bull. Chem. §oc. Jap., 43, 60 (1975). T. Iwaichido, A. Sadakane, and K. Toei, Bull. Chem. §oc. Jap., 53, 629 (1978). J. W. Mitchell and D. L. Shanks, Anal. Chem., 43, 642 (1975). A. M. Y. Jaber, G. J. Moody, and J. D. R. Thomas, Inorg. Nucl. Chem., 33, 1689 (1977). M. Kawasaki, K. Toei, and T. Iwachido, Chem. Lett., 417 (1972). T. Iwachido, M. Kimura, and K. Toei, Chem. Lett., 1101 (1976). T. Iwachido, M. Minami, A. Sadakane, and K. Toei, Chem. Lett., 1511 (1977). P. R. Danesi, H. Meider-Gorican, R. Chiarizia, adn G. Scibona, J. Inorg. Nucl. Chem., 33, 1479 (1975). P. R. Danesi, R. Chiarizia, C. Fabiani, and C. Dome- nichini, J. Inorg. Nucl. Chem., 33, 1226 (1976). Lj. Tusek, P. R. Danesi, and R. Chiarizia, J. Inorg. Nucl. Chem., 33, 1538 (1975). 1 Y. Marcus and L. E. Asher, J. Phys. Chem., 83, 1246 (1978). T. Sekine, H. Wakabayashi, and Y. Hasegawa, Bull. Chem. Soc. Jap., 53, 645 (1978). D. Sevdic and H. Meider, J. Inorg. Nucl. Chem., 33, 1403 (1977). D. Sevdic and H. Meider, J. Inorg. Nucl. Chem., 33, 1409 (1977). G. W. Gokel, D. J. Cram, C. L. Liotta, H. P. Harris, and F. L. Cook, J. Org. Chem., 33, 2445 (1974). 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 143 I. S. Perelygin and M. A. Klimchuk, Russ. J. Phys. Chem., 43, 1138 (1973). M. S. Greenberg and A. I. P0pov, J. Soln. Chem., 5, 653 (1976). Y. M. Cahen and A. I. Popov, J. Soln. Chem., 4 599 (1975). “ K. Balasubrahmanyam and G. J. Janz, J. Amer. Chem. Soc., 93, 4189 (1970). T. G. Chang and D. E. Irish, J. Soln. Chem., 3, 161 (1974). J. F. Hinton and E. S. Amis, Chem. Rev., 6], 367 (1967). W. J. DeWitte, Ph.D. Thesis, Michigan State Uni- versity, East Lansing, Mich. (1976). L. L. Hsu, Masters Thesis, Michigan State University, East Lansing, Mich. (1976). E. H. Mei, Ph.D. Thesis, Michigan State University, East Lansing, Mich. (1977). A. R. Hourdakis, Ph.D. Thesis, Michigan State Uni- versity, East Lansing, Mich. (1978). N. F. Ramsey, Phys. Rev., 23, 699 (1950). W. G. Schneider and A. D. Buckingham, Discuss. Faraday_Soc., 33, 147 (1962). A. G. Lee, Coord. Chem. Rev., 3, 289 (1972). F. J. Kayne and J. Reuben, J. Amer. Chem. Soc., 93, 220 (1970). - J. Reuben and F. J. Kayne, J. Biol. Chem., 3&3, 6227 (1971). J. J. Dechter and J. I. Zink, J. Amer. Chem. Soc., 93, 2937 (1975). J. F. Hinton and R. W. Briggs, J. Mag. Res., 23, 379 (1977). R. Freeman, R. P. H. Gasser, R. E. Richards, and D. H. Wheeler, Molecular Phys., 3, 75 (1959). 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103.. 144 R. Freeman, R. P. H. Gasser, and R. E. Richards, Molecular Phys., 3, 301 (1959). J. J. Dechter and J. I. Zink, Chem. Comm., 96 (1974). J. F. Hinton and R. W. Briggs, J. Mag. Res., 13, 393 (1975). R. H. Erlich, M. S. Greenberg, and A. I. Popov, J. Amer. Chem. Soc., 93, 4989 (1970). R. W. Briggs and J. F. Hinton, J. Soln. Chem., 1, l (1978). R. W. Briggs and J. F. Hinton, J. Soln. Chem., Q, 827 (1977). E. Shchori, N. Nae, and J. Jagur-Grodzinski, J. C. S. Dalton, 2381 (1975). J. Koryta and M. L. Mittal, J. Electroanal. Chem., 36, App. 14 (1972). L. J. Rodriguez, G. W. Liesegang, R. D. White, M. M. Farrow, N. Purdie, and E. M. Eyring, J. Phys. Chem., 83, 2118 (1977). C. Srivanavit, J. I. Zink, and J. J. Dechter, J. Amer. Chem. Soc., 93, 5876 (1977). J. M. Lehn and J. P. Sauvage, Chem. Comm., 440 (1971). J. S. Shih, private communication. J. H. Mathews, J. Amer. Chem. Soc., 43, 562 (1926). M. G. Mellon, Analytical Absorption Spectrosc0py, New York: John Wiley & Sons, Inc., 1950, p. 261. D. D. Traficante, J. A. Simms, and M. Mulcay, J. Mag. Res., 13, 484 (1974). D. A. Wright and M. T. Rogers, Rev. Sci. Instr., 43, 1189 (1973). J. R. Zimmerman and M. R. Foster, J. Phys. Chem., 83, 282 (1957). D. H. Live and S. I. Chan, Anal. Chem., 43, 791 (1970). 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 145 J. L. Dye and V. A. Nicely, J. Chem. Ed., 43, 443 (1971). C. M. French and I. G. Roe, Trans. FaradayfiSoc., 43, 791 (1953). T. Abe, Nature, 131, 234 (1960). A. Brandstrom, P. Berntsson, S. Carlsson, A. Djurhuus, K. Gustavii, U. Junggren, B. Lamm, and B. Samuelsson, Acta. Chem. Scand., 23, 2202 (1969). R. G. Pearson, J. Amer. Chem. Soc., 83, 3533 (1963). E. T. Roach, P. R. Handy, and A. I. Popov, Inorg. Nucl. Chem. Lett., 2, 359 (1973). W. L. DeWitte, L. Liu Hau, E. Mei, J. L. Dye, and A. I. Popov, J. Soln. Chem., Q, 337 (1977). A. Hourdakis and A. I. POpOV, J. Soln. Chem., Q, 299 (1977). E. Mei, L. Liu Hsu, J. L. Dye, and A. I. Popov, J. Soln. Chem., Q, 771 (1977). E. Mei, A. I. Popov, and J. L. Dye, J. Phys. Chem., 83, 1677 (1977). NI LIBRARIES MIIIIIIIIIIIIIIIIINIIHIIIIIIIIII | NIH 3 1293 01061 77 44