Complexation of f-Elements with

1 downloads 0 Views 2MB Size Report
Apr 9, 2004 - Van Pelt, Dr. Michael G. Bronkowski and Dr. William J. Crooks for help getting started ... Burrgt for their help with laser fluorescence experiments and ...... Typical experimental solutions were 10.00 or 20.00 mL with ... dissolution of a sample of known mass in deuterium oxide. ...... 0.434 and –0.260 e.Å. -3 ...
The Florida State University

DigiNole Commons Electronic Theses, Treatises and Dissertations

The Graduate School

4-9-2004

Complexation of f-Elements with Aminopolycarboxylate Ligands Glenn A. Fugate Florida State University

Follow this and additional works at: http://diginole.lib.fsu.edu/etd Recommended Citation Fugate, Glenn A., "Complexation of f-Elements with Aminopolycarboxylate Ligands" (2004). Electronic Theses, Treatises and Dissertations. Paper 4387.

This Dissertation - Open Access is brought to you for free and open access by the The Graduate School at DigiNole Commons. It has been accepted for inclusion in Electronic Theses, Treatises and Dissertations by an authorized administrator of DigiNole Commons. For more information, please contact [email protected].

THE FLORIDA STATE UNIVERSITY COLLEGE OF ARTS AND SCIENCES

COMPLEXATION OF f-ELEMENTS WITH AMINOPOLYCARBOXYLATE LIGANDS

By GLENN A. FUGATE

A Dissertation submitted to the Department of Chemistry in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Degree Awarded: Spring Semester, 2004

The members of the Committee approve the dissertation of Glenn A. Fugate defended on 9 April 2004.

__________________________ Gregory R. Choppin Professor Directing Dissertation

__________________________ William M. Landing Outside Committee Member

__________________________ Naresh Dalal Committee Member

__________________________ Kenneth A. Goldsby Committee Member

The Office of Graduate Studies has verified and approved the above named committee members.

ii

ACKNOWLEDGEMENTS I would like to thank my major professor, Dr. Gregory R. Choppin, for his patience, aid and guidance throughout this project. I would like to thank Dr. Kenneth L. Nash of Argonne National Laboratory for past, present and future guidance, for his help and ideas on this project and for being my friend. I would like to thank all the past and present members of the Choppin group for their help, guidance and camaraderie. I would like especially like to thank: Dr. Craig Van Pelt, Dr. Michael G. Bronkowski and Dr. William J. Crooks for help getting started in the laboratory, Dr. Serguei I. Sinkov for his help and insight, Dr. Ivan Laszak, Dr. Isabel Paiva and Dr. Clara Comuzzi for their friendship, Dr. Alfred Morgenstern for helping me leave the laboratory more often, and Dr. Marian Borkowski for games of pool, knowledge, support and friendship. I appreciate help I have received from staff members of the Department of Chemistry. I would especially like to thank: Dr. David Gorman and Dr. Burt van de Burrgt for their help with laser fluorescence experiments and equipment, Dr. Ronald J Clark for determining the crystal structures in this work, Dave Parker, Charlie Betts, Gary Poplin and Tom Wages for help with electronic and computer problems, Randall Pelt for making glass replacement parts quickly, Lamar Norman and Keith Collins for help in repairing mechanical problems, and Dr. Joseph B. Vaughn, Steve Freitag and Dr. Thomas Gedris for help with the setup and operation of the NMR experiments in this work. Dr. Gedris is due additional thanks for running some of the NMR spectra as well as being a great gardener, an excellent cue maker, a fantastic pool player and teacher and a generally remarkable human being. iii

I would like to thank Dr. Paul Rickert of Argonne National Laboratory for the synthesis of two of the ligands studied in the work. I would like to thank the friends who have helped me through this process: Dr. Joseph Lehnes for his love of football, video games and late night dinners, Dr. Jennifer Llewelyn for being a good listener, Dr. Elizabeth “Libby” Mayo, Amy Aldridge, David Gilmore, Dr. David Sunseri, Doug Hattaway, Dr. Thomas Gedris, Dr. Marian Borkowski and all the other members of “The Village Idiots” for a lot of laughs during a little pool. Lastly, I would like to thank my family for their help, love, support and understanding.

iv

TABLE OF CONTENTS

List of Tables ................................................................................................ix List of Figures ................................................................................................xi List of Abbreviations and Symbols ....................................................................xii Abstract ......................................................................................................xiii 1. INTRODUCTION ..........................................................................................1 1.1 Statement of Problem............................................................................1 1.2 Aminocarboxylate Properties................................................................2 1.3 Lanthanide(III) Elements ......................................................................3 1.3.1 General Properties........................................................................3 1.3.2 Comparison of Lanthanide Ions (4f) with Actinide Ions (5f) ......4 1.3.3 Separation of Lanthanide(III) and Actinide(III) Ions ..................4 1.3.4 Spectroscopy of Lanthanides .......................................................5 1.4 Objectives of Present Work ..................................................................6 2. EXPERIMENTAL..........................................................................................14 2.1 Chemical Reagents................................................................................14 2.1.1 Lanthanide(III) Perchlorate Stock Solutions ...............................14 2.1.2 Sodium Perchlorate Stock Solutions............................................15 2.1.3 Ligands.........................................................................................15 2.1.3.1 Chelidamic Acid ..............................................................15 2.1.3.2 Piperidine-2,6-dicarboxylic Acid.....................................15 2.1.3.3 2,6-Dicarboxypipepridine-N-acetic Acid ........................16 2.1.4 Potentiometric Titration Solutions...............................................16 2.1.4.1 Sodium Hydroxide Standard Solutions............................16 2.1.4.2 Perchloric Acid Standard Solutions .................................16 2.1.4.3 Ligand Solutions ..............................................................17 2.1.4.4 Lanthanide(III) Complex Solutions .................................17 2.1.5 Laser Induced Fluorescence Solutions.........................................17 2.1.6 Hypersensitivity Spectroscopy Solutions ....................................18 2.1.7 Nuclear Magnetic Resonance Solutions ......................................18 2.1.7.1 Solutions for Adjusting pH ..............................................18 v

2.1.7.2 Ligands Solutions.............................................................19 2.1.7.3 Lanthanide(III) Complex Solutions .................................19 2.1.8 Crystals ........................................................................................20 2.1.9 Calorimetry Solutions ..................................................................20 2.1.9.1 Standard Solutions ...........................................................20 2.1.9.2 Ligand Solutions ..............................................................20 2.1.9.3 Lanthanide(III) Perchlorate Solutions..............................21 2.2 Equipment .............................................................................................21 2.2.1 Mass Determination .....................................................................21 2.2.2 Potentiometric Measurements......................................................21 2.2.3 Potentiometric Titrations .............................................................21 2.2.4 Laser Induced Fluorescence.........................................................22 2.2.4.1 Fluorescence Spectra .......................................................22 2.2.4.2 Fluorescence Lifetime Measurements .............................23 2.2.5 Hypersensitivity Spectroscopy.....................................................23 2.2.6 NMR ............................................................................................23 2.2.7 Crystallography............................................................................24 2.2.8 Calorimetry ..................................................................................24 2.3 Procedures.............................................................................................25 2.3.1 Potentiometric Measurements......................................................25 2.3.2 Potentiometric Titrations .............................................................25 2.3.2.1 Standardization ................................................................25 2.3.2.2 Acid Dissociation Constant Determination .....................26 2.3.2.3 Stability Constant Determination.....................................26 2.3.3 Laser Induced Fluorescence.........................................................26 2.3.3.1 Fluorescence Experiments ...............................................26 2.3.3.2 Fluorescence Lifetime Experiments ................................27 2.3.4 Hypersensitivity Spectroscopy.....................................................27 2.3.5 NMR ............................................................................................27 2.3.5.1 pH Titrations ....................................................................27 2.3.5.2 Coalescence Experiments ................................................28 2.3.5.3 Nuclear Overhauser Effect Experiments .........................28 2.3.5.4 13C 1H Decoupled Spectra................................................29 2.3.5.5 1H Homonuclear Decoupled Spectra ...............................29 2.3.6 Crystallography............................................................................29 2.3.7 Calorimetry ..................................................................................30 2.3.7.1 Heat of Dilution ...............................................................30 2.3.7.2 Standardization ................................................................31 2.3.7.3 Heat of Protonation ..........................................................31 2.3.7.4 Heat of Complexation ......................................................31 3. CALCULATIONS..........................................................................................32 3.1 Potentiometric Titrations ......................................................................32 3.1.1 Electrode Calibration ...................................................................32 3.1.2 Acid Dissociation Constant Determination .................................33 vi

3.1.3 Lanthanide(III) Complex Stability Constant Determination .......34 3.2 PSEQUAD ............................................................................................34 3.3 Laser Induced Fluorescence..................................................................35 3.3.1 Fluorescence Lifetime Measurements .........................................35 3.3.2 Fluorescence Spectra ...................................................................35 3.4 Hypersensitivity Spectroscopy..............................................................36 3.5 NMR ................................................................................................37 3.5.1 Acid Dissociation Constant Determination .................................37 2.5.2 Nuclear Overhauser Effect...........................................................39 3.6 Calorimetry ...........................................................................................40 3.6.1 Change of Enthalpy......................................................................40 3.6.2 Change of Entropy .......................................................................41 4. RESULTS……. ..............................................................................................42 4.1 Acid Dissociation Constants .................................................................42 4.2 Lanthanide(III) Stability Constants.......................................................43 4.3 Calorimetry ...........................................................................................43 4.4 Hypersensitivity Spectroscopy..............................................................44 4.4.1 PDA..............................................................................................44 4.4.2 DPA and CA ................................................................................44 4.4.3 Oscillator Strengths......................................................................46 4.5 Eu(III) Laser Fluorescence ...................................................................46 4.5.1 Excitation Spectra ........................................................................46 4.5.2 PDA..............................................................................................47 4.5.3 DPA..............................................................................................47 4.5.4 CA ................................................................................................48 4.6 NMR ................................................................................................49 4.6.1 Acid Dissociation Constant Determination .................................49 4.6.2 1H NMR Peak Identification of PDA ..........................................49 4.6.3 1H NMR Peak Identification of DPA ..........................................50 4.6.4 NMR of Lanthanide(III) Complexes ...........................................51 4.7 Crystallography.....................................................................................52 5. DISCUSSION.. ...............................................................................................74 5.1 Acid Dissociation Constants .................................................................74 5.2 Lanthanide(III) Stability Constants.......................................................75 5.2.1 PDA..............................................................................................75 5.2.2 DPA..............................................................................................75 5.2.3 CA ................................................................................................76 5.3 Calorimetry ...........................................................................................77 5.3.1 Ligand Protonation.......................................................................77 5.3.1.1 DPA..................................................................................77 5.3.1.2 CA ....................................................................................77 5.3.2 Change in Enthalpy......................................................................78 5.3.2.1 PDA..................................................................................78 vii

5.3.2.2 DPA..................................................................................78 5.3.2.3 CA ....................................................................................79 5.3.3 Change in Entropy .......................................................................81 5.3.3.1 PDA..................................................................................81 5.3.3.2 DPA..................................................................................81 5.3.3.2 CA ....................................................................................82 5.4 Hypersensitivity Spectroscopy..............................................................82 5.4.1 PDA..............................................................................................82 5.4.2 DPA..............................................................................................83 5.4.3 CA ................................................................................................83 5.5 Eu(III) Laser Fluorescence ...................................................................84 5.5.1 PDA..............................................................................................85 5.5.2 DPA..............................................................................................85 5.5.3 CA ................................................................................................85 5.6 NMR….. ...............................................................................................86 5.6.1 PDA..............................................................................................86 5.6.2 DPA..............................................................................................86 5.6.3 CA ................................................................................................87 5.7 Crystallography.....................................................................................87 5.8 Conclusions...........................................................................................88 5.8.1 PDA..............................................................................................88 5.8.2 DPA..............................................................................................88 5.8.3 CA. ...............................................................................................89 5.8.4 Steric Hindrance of Aminopolycarboxylate Ligands ..................90 5.8.4.1 Piperidine Ring .........................................................................90 5.8.4.2 CA. ............................................................................................90 5.9 Future Work ..........................................................................................91 APPENDICES A Hypersensitive Spectroscopy ................................................................104 B Eu(III) Laser Induced Fluorescence......................................................127 C 1H NMR Spectroscopy..........................................................................135 D Crystal Structures..................................................................................141 E Previously Published Data From This Work. ......................................154 REFERENCES

................................................................................................170

BIOGRAPHICAL SKETCH ..............................................................................176

viii

LIST OF TABLES

Table 4.1: The concentration acid dissociation constants for PDA, T = 25.0°C, NaClO4 media......................................................................54 Table 4.2: The concentration acid dissociation constants for DPA, T = 25.0°C, NaClO4 media ......................................................................55 Table 4.3: The concentration acid dissociation constants* for CA, T = 25.0°C, NaClO4 media ...................................................................................56 Table 4.4: Stability constants of select lanthanide(III) ions with DPA, T = 25.0° C NaClO4 media.......................................................................57 Table 4.5: Stability constants* of select lanthanide(III) ions with CA, T = 25.0° C, NaClO4 media.....................................................................58 Table 4.6: Thermodynamic values for the DPA complexation with Ho(III) or Nd(III), pH = 5.30 to 5.50, [MES] = 0.100 M, T = 25.000°C, I = 0.50 M NaClO4. ................................................................................59 Table 4.7: Thermodynamic values for the CA complexation with Ho(III) or Nd(III), pH = 5.30 to 5.50, [MES] = 0.100 M, T = 25.000°C, I = 0.50 M NaClO4 .................................................................................60 Table 4.8: Values from the linear regressions of Figure 4.1 for the change of the oscillator strength of Nd(III) when complexed with DPA and CA, I = 2.00 M NaClO4, pH = 4.50 - 6.00, T = 23°C.......................62 Table 4.9: Values from the linear regressions of Figure 4.2 for the change of the oscillator strength of Ho(III) when complexed with DPA and CA, I = 2.00 M NaClO4, pH = 4.50 - 6.00, T = 23°C.......................64 Table 4.10: The oscillator strengths of Nd(III) and Ho(III) and their complexes with DPA and CA, I = 2.00 M NaClO4, T= 23°C ..........65 Table 4.11: Summary of the peaks observed by fluorescence peaks of Eu(III) for the various PDA, DPA and CA, I = 0.50 and 2.00 M NaClO4, T = 23°C............................................................................................66 Table 4.12: The nuclear Overhauser effect of PDA at different pH, I = 0.50 M NaClO4 and T = 25.0°C ....................................................................69 Table 4.13: The observed nuclear Overhauser effect of DPA* at different pH, I = 0.50 NaClO4 and T = 25.0°C......................................................71 Table 5.1: Acid dissociation constant values for PDA, DPA and CA................92 ix

Table 5.2: Acid dissociation constant values for IDA, NTA, dipicolinic acid and hydroxybenzene ..........................................................................93 Table 5.3: The lanthanide(III) stability constants of PDA and DPA, T = 25 °C, I = 0.1 M KNO3 .....................................................................94 Table 5.4: Lanthanide(III) stability constants of IDA, NTA and dipicolinic acid, T = 25 °C, I = 0.1 M KNO3 .......................................................95 Table 5.5: Previously determined thermodynamic values of CA with the lanthanum(III) ion , T = 25°C, I = 0.5 M NaClO4 ............................96 Table 5.6: Thermodynamic values for the first protonation of IDA, NTA, dipicolinic acid, 4-hydroxypyridine and hydroxybenzene, T = 25 °C, I = 0.5 M NaClO4 ..................................................................97 Table 5.7: Thermodynamic values of lanthanide(III) complexation with IDA, NTA, and dipicolinic acid, T = 25 °C, I = 0.5 M NaClO4 ......98 Table 5.8: Nd(III) and Ho(III) oscillator strength values for NTA, IDA, dipicolinic acid and CA ....................................................................101 Table 5.9: Eu(III) fluorescence peaks for acetic acid, dipicolinic acid, CA, NTA and IDA ligand species............................................................102 Table 5.10: Eu(III) waters of hydration and ligand coordination numbers for acetic acid, dipicolinic acid, CA, NTA and IDA ligand species ...............................................................................................103

x

LIST OF FIGURES

Figure 1.1: Structures of NTA, EDTA and DTPA .............................................9 Figure 1.2: Structures of α-picolinic acid and dipicolinic acid ...........................10 Figure 1.3: Structures of PDA and DPA .............................................................11 Figure 1.4: Structural geometries and ring flip of the PDA and DPA.................12 Figure 1.5: Tautomer structures of chelidamic acid ...........................................13 Figure 4.1: The oscillator strength of Nd(III) complexes plotted versus the ligand to metal concentration ratio, I = 2.00 M NaClO4, pH = 4.50 - 6.00, T = 23°C ................................................................................61 Figure 4.2: The oscillator strength of Ho(III) complexes plotted versus the ligand to metal concentration ratio, I = 2.00 M NaClO4, pH = 4.50 - 6.00, T = 23°C ...............................................................................63 Figure 4.3: The chemical shift of the peaks in the 1H NMR spectrum of PDA versus –log of concentration of the deuterium ion, I = 0.50 M NaClO4 and T = 25.0°C. The peak legend refers to proton assignments given in Figure 4.5 .......................................................67 Figure 4.4: The chemical shift of various peaks in the 1H NMR spectrum of DPA versus –log of concentration of the deuterium ion, I = 0.50 M NaClO4 and T = 25.0°C. The peak legend refers to proton assignments given in Figure 4.6 ........................................................68 Figure 4.5: PDA proton atom designations used with 1H NMR .........................70 Figure 4.6: DPA proton atom designations used with 1H NMR .........................72 Figure 4.7: CA carbon atom designation used with 13C NMR ...........................73 Figure 5.1: The relationship between the residual enthalpy, δ∆H, of the first complex formation with Nd(III) and the total basicity of the nitrogen donor atoms of select polycarboxylate ligands ..................99 Figure 5.2: Correlation of ∆S1 of Nd(III) and the number of carboxylate groups with acetate (ac) and select polycarboxylate ligands ...........100

xi

LIST OF ABBREVIATIONS AND SYMBOLS

CA

chelidamic acid

CNL

calculated coordination number of the ligand

DPA

1,2-dicarboxypiperidine-N-acetic acid

DTPA

diethylenetriaminepentaacetic acid

EDTA

ethylenediaminetetraacetic acid

nH2O

number of water molecules in the inner hydration sphere of the Eu(III) ion

nOe

nuclear Overhauser effect

PDA

piperidine-1,2-dicarboxylic acid

PM

oscillator strength of the metal aquo ion

PML

oscillator strength of the ML complex

PML2

oscillator strength of the ML2 complex

xii

ABSTRACT Piperidine-2,6-dicarboxylic acid (PDA), 2,6-dicarboxypiperidine-N-acetic acid (DPA) and chelidamic acid (CA) were characterized using a variety of techniques. The acid dissociation constants of these ligands were determined by potentiometric titration and 1H NMR. The crystal structures of PDA and DPA were determined using an X-ray diffractometer. The stability constants of select trivalent lanthanide ions (La, Nd, Eu, Ho and Lu) with DPA and CA were determined by potentiometric titration. The oscillator strength of the Nd and Ho complexes of DPA and CA were determined by spectroscopic titration. These oscillator strength values were used to examine coordination effects of DPA and CA on Nd(III) and Ho(III). The 7F0 → 5D0 selective excitation was used to examine the inner coordination sphere of the Eu(III) ions upon coordination with DPA and CA. The number of Eu(III) hydrating water molecules and the calculated ligand coordination number were determined for all complex species. The ∆H and ∆S of protonation values and ∆H and ∆S values of Ho(III) and Nd(III) complexation were determined by calorimetric titration. The stability constants of PDA could not be determined by potentiometric titration. Oscillator strength measurements, calorimetric titration and laser fluorescence indicated only minimal bonding was occurring between the lanthanide(III) ions and PDA.

xiii

CHAPTER 1 INTRODUCTION

1.1 Statement of Problem Ions of the lanthanide(IIII) series tend to form similar coordination bonds to a ligand due to the similar charge and size of the cations. Molecules with a flexible structure can reorient the coordination site around the changing size of the lanthanide(III) ions and so have similar stability constants for these trivalent metal ions. A more rigid coordination structure could decrease the stability constant of the ligand but increase the lanthanide(III) size selectivity of the ligand due to varied coordination distances and the different radii of the ions. Such compounds could provide improved materials for the separation of lanthanide(III) or actinide(III) ions. Aminocarboxylate ligands have been studied for the separation of lanthanide(III), and actinide(III) ions1,2.

Typically these compounds are straight chain aliphatic

molecules with acetic acid functional groups such as nitrilotriacetic acid (NTA), ethylenediaminetetraacetic acid (EDTA) and diethylenetriaminepentaacetic acid (DTPA). Many aminopolycarboxylate ligands have a discontinuous trend in their lanthanide(III) series stability constants which would cause simultaneous elution of several lanthanide(III) ions from a cation exchange column2. Studies of lanthanide(III) and actinide(III) complexation with different variations of the structure of the coordination

1

site of the ligand can add insight into the nature of the aminopolycarboxylate coordination and may lead to more size selectivity in +3 f-element separations. 1.2 Aminocarboxylate Properties Compounds that contain at least two carboxylic acid groups bonded to an amine first appeared in Swiss patents between 1937 to 1940. Studies of alkaline earth and lanthanide coordination indicated strong metal bonding occurred between the cation and both the carboxylic acid groups and the amine3,4,5,6,7. These studies led to the use of IDA and NTA as elution reagents with cation exchange columns for the separation of lanthanide(III) ions8.

Aminopolycarboxylate ligands were found to have better

lanthanide separation compared to the previously used eluting agent, citrate9,10,11. The lanthanide coordination complexes of many of these compounds have been studied including NTA12,13,14, EDTA14,15,16,17 and DTPA14,18,19,20, shown in Figure 1.1. These molecules consist of an amine bonded to between one and three acetate groups with multiple units connected by an aliphatic carbon chain. Metal cations can coordinate with both the oxygen donor of a carboxylic acid and the nitrogen donor of an amine, giving aminocarboxylate ligands affinity for both hard and soft Pearson acids. Studies have examined the effects on lanthanide coordination caused by replacing the acetate group with a longer chain carboxylic acid. Studies have shown that bidentate lanthanide(III) coordination is most stable upon the formation of a 5-membered ring21,22,23. A metal coordinated by both the amine and the carboxylic acid forms a 5-membered ring in which the carboxylate is in an acetate group.

Other studies compared lanthanide(III)

coordination stability of EDTA analogs where the ethylene was replaced with longer aliphatic chains24,25.

Ethylene groups promote the formation of a 5 member ring

involving the coordination of both amine groups to the metal ion. Both substitution at the alkyl linking group and the carboxylic acid group demonstrate the importance of forming 5-membered ring structures in bidentate coordination.

Additionally, the effects of

replacing one or more of the acetate groups with a non-coordinating functional group such as methylene14 or hydrogen14 or coordinating functional groups such as alcohol14 or phosphonate26 have been studied.

2

Several studies have characterized the lanthanide(III) coordination chemistry of the aromatic aminocarboxylates α-picolinic acid14,27,28,29 and dipicolinic acid30,31, shown in Figure 1.2. All of the coordinating functional groups of these molecules have a fixed geometry due to the rigid, aromatic ring structure. Complexation of the aromatic ligands can not be compared directly to that of aliphatic ligands because the aromatic amine has increased electron density and forms shorter, stronger coordination bonds to lanthanide ions than aliphatic amines14. The rigid structure of dipicolinic acid does not increase the size selectivity across the lanthanide series compared to IDA, a straight chain aliphatic molecule with a similar coordination site. Although dipicolinic acid has a discontinuous stability constant trend across the lanthanide series, this ligand has been used for the separations of actinide ions from actinide, lanthanide and other metals ions32,33. 1.3 Lanthanide(III) Elements 1.3.1 General Properties Elements with an atomic number between 57 (lanthanum) and 71 (lutetium) are included in the lanthanide series. These elements have the gaseous electron configuration of [Xe]4fn5d06s2 except La ([Xe]4f05d16s2), Ce ([Xe]4f15d16s2), Gd ([Xe]4f75d16s2 ) and Lu [Xe]4f145d16s2 with the electron configuration of Gd reflecting the stability of the half filled shell34,35. The lanthanide ions are primarily in the +3 oxidation state because electrons in outer orbitals significantly shield the f electrons. Other stable oxidations states observed with lanthanide ions include +2 (Eu and Yb) and +4 (Ce and Tb), corresponding to vacant, half filled or filled 4f sublevels. The sizes of the +3 lanthanide ions are similar due to the lanthanide contraction and range from 1.216 Å for La(III) to 1.032 Å for Lu(III)36. Due to the decreasing size of the lanthanide ions across the series, the charge density (Z/rLn) of the ion increases across the series. The primary hydration sphere number changes from 9 (La(III) to Nd(III)) to 8 (Tb(III) to Lu(III)) due to the decreasing ionic radius and the increasing charge density with elements between Nd(III) and Tb(III) having either 8 or 9 waters of hydration37. Because of the similar charge and size, most lanthanide(III) ions behave similarly both physically and chemically. Lanthanide(III) ions are more similar to alkali and alkaline earth metals, coordinating as Pearson hard acids, than to d-block metals, preferring to bind to hard 3

acids like oxygen and fluorine donors than to soft donors like nitrogen and sulfur. The +3 lanthanides begin to hydrolyze above pH 6 and the hydroxide can readily compete for coordination of the metal ion at higher pH38. Lanthanide(III) ions can form both inner and outer sphere complexes.

An inner sphere complex forms through an Eigen

mechanism as the outer sphere complex replaces one or more primary hydration sphere water molecules39,40,41. The coordination is primarily ionic with a solution coordination typically ranging from 8 to 9, depending on properties of the ligand, solvent and concentration42. 1.3.2 Comparison of Lanthanide Ions (4f) with Actinide Ions (5f) The trivalent charge that is commonly observed for most lanthanide ions is not the most stable oxidation state for early (Th – Pu) actinides or the late actinide nobelium. The lanthanide(III) ions are similar in size and are commonly used as analogs for actinide(III) ions42,43. The 5f orbitals are significantly shielded from binding interactions similarly to the 4f orbitals. The actinide 5f orbitals have greater spatial distribution compared to the 7s and 7p orbitals than the spatial distribution of the lanthanide 4f orbitals compared to the 6s and 6p orbitals. The larger spatial distribution increases the possibility that actinide bonding has a small degree of covalent nature compared to lanthanide bonding44.

The increased covalency results in actinide elements having

slightly higher affinity for soft electron donors, such as amines. The increased affinity for soft donors has been used to separate lanthanide and actinide elements by use of a soft donor system like thiocyanate45 or a ligand with both hard and soft donor groups such as aminocarboxylates42. 1.3.3 Separation of Lanthanide(III) and Actinide(III) Ions The actinide(III) and lanthanide(III) ions are separated by the increased affinity of actinide ions for soft Pearson donors like amines. The slightly increased stability of the Am(III) complexes has been postulated to be caused by the increased covalency of 5f versus 4f elements. Calorimetry data has indicated that there is no significant enthalpy difference between amine coordination bonding of Am(III) and Nd(III) in aminopolycarboxylate compounds14.

Mössbauer spectroscopy measurements have

shown increased shift values as the actinide oxidation state is increased46. 4

These

increased shift values are taken to be qualitative evidence for an increased An-L covalent bonding compared to that of Ln-L, and has been attributed to mixing of 5d, 6d and 7s orbitals47,48.

Solvent extraction and synergistic solvent extraction studies have also

shown enhanced covalent nature in the actinide(III) ions49,50. A mixed donor ligand with improved size selectivity for lanthanide(III) ions can be used to separate actinide(III) ions and to separate actinide(III) ions from lanthanide(IIII) ions. The Am(III) and Nd(III) ions have similar charge and effective ionic radii, 1.106 Å and 1.109 Å for 8 coordinate ions, respectively36. A plot of log βAmL versus log βNdL for a series of aminopolycarboxylate ligands has been shown to be linearly related with a slope of 1.142. Actinide(III) ion elution can be predicted by the elution of similarly sized lanthanide(III) ions. The 4f and 5f ions can then be separated by the increased interaction of an actinide ion with soft donor groups. 1.3.4 Spectroscopy of Lanthanides Absorption bands of lanthanide ions arise from both ground state and excited state electronic transitions with different J-states of the same 4fn configuration, resulting from different electron distributions in the 4f orbital. The f-f transitions do not change parity and so are LaPorte forbidden. The f orbitals are well shielded and so the symmetry of the orbitals is generally not affected by ligand field as commonly occurs in d-d transitions. Since the ligand field does not relax the selection rules, lanthanide f-f transitions are weak with molar absorptivities less than 10 M-1 cm-1 and the bands are very narrow. Some f-f transitions are affected by the ligand field, exhibiting both spectral intensity and peak positions changes.

These transitions are termed hypsersensitive and have the

following characteristic change in term symbols: ∆S = 0 or 2, |∆L| ≤ 2 and |∆J| ≤ 251. They follow selection rules for electric quadrupole transitions but have intensities of electric dipole transitions and have been explained in terms of an inhomogeneous dielectric surrounding the lanthanide ion derived from the ligand field effect34. Hypersensitive transitions have also been explained by a charge transfer due to the covalency of the lanthanide-ligand bond52. Hypersensitive transitions have been used to distinguish inner sphere and outer sphere complexation53,54,55 and to determine the coordination number of lanthanide(III) ions in solution56,57. 5

Lanthanide ions also have characteristic luminescence spectra in the visible and near-IR wavelength regions58,59,60.

Luminescence radiation emissions occur because

lanthanide-ligand radiationless de-excitation processes are relatively ineffective. The luminescence of Eu(III) and Tb(III) have been studied extensively59. The emission bands of Eu(III) occur from the 5D0 lowest excited state to the 7FJ ground state where J = 0 to 6. Since both the excited and ground states of the 7F0 → 5D0 selective excitation (excitation from 578 to 581 nm and emission at 614 nm) are nondegenerate and can not be further split by the ligand field, the wavelength of the excitation occurs at different wavelength for each different Eu(III) environment. The change of fluorescence intensity and peak position are related to the number of Eu(III) coordination environments of the system61. The O-H oscillator is an effective quenching agent of Eu(III) luminescence. This phenomenon has been explained by an overlap of the ν = 3 O-H energy level with the 5D0 lowest excited state59.

The O-D oscillator does not quench the luminescence as

effectively and so species have prolonged lifetimes and increased intensity in D2O. The number of water molecules in the inner hydration sphere of the Eu(III) ion has been shown to be related to the difference of the fluorescence lifetimes of a species in water and D2O62. 1.4 Objectives of Present Work The goal of the present work is to study the interactions of a series of lanthanide(III) cations (La, Nd, Eu, Ho and Lu) with aliphatic aminocarboxylate ligands that have structural features that reduce the reorganization of the coordination site such as piperidine-1,2-dicarboxylic acid (PDA) and 1,2-dicarboxypiperidine-N-acetic acid (DPA), shown in Figure 1.3. These molecules have a coordination site that is sterically hindered by the piperidine ring structure which is directly bonded to the carboxylic acid functional groups.

Most aliphatic aminocarboxylate ligands previously studied are

straight chain molecules in which the carboxylic acid groups are bonded to the amine and have no other restrictions on the coordination geometry, as discussed in section 1.2. The acid dissociation constants and lanthanide(III) complex stability constants of PDA63,64 and DPA65 have been previously reported.

The size selectivity of PDA across the

lanthanide(III) series is lower than the straight chain coordination analog iminodiacetic 6

acid (IDA) while DPA was reported to have improved size selectivity across the lanthanide(III) series compared to the aliphatic structural analog NTA. The ligands and their lanthanide(III) complexes were not characterized structurally nor spectroscopically. The coordination geometry of PDA and DPA should be more organized and more rigid than straight chain aliphatic molecules due to the piperidine ring structure. The increased rigidity and pre-organization of the coordination site should increase the size selectivity of the ligands. The piperidine ring exists in either a boat or chair form similar to cyclohexane. The boat form can convert to the opposite boat form where the axial and equatorial positions are reversed, proceeding through twist-boat and boat geometries. This conversion, shown in Figure 1.4, is called ring flip. The interchange of chair and boat forms of the piperidine ring, ring flip, can cause nitrogen inversion that would reorient the coordination bonds of the complex, which should decrease the stability of the complex and decrease the size affinity of PDA and DPA. The coordination chemistry of chelidamic acid (CA) has been characterized with some divalent alkaline earth and d-block ions66,67,68 and trivalent lanthanum69. CA is a coordination analog of dipicolinic acid, but the ring structure has increased flexibility due to the tautomers, shown in Figure 1.5, that exist simultaneously in solution. The keto tautomer has been shown to distort the planar structure that is observed in dipicolinic acid70. The distortion is caused by increased sp3 orbital character of the nitrogen that causes the amine and the carboxylic acid groups to adapt non-planar orientations. The bond lengths of the CA piperidine ring vary according to the structure of the keto tautomer with alkene bonds being shorter and aliphatic bonds being longer than observed for dipicolinic acid. The increased flexibility of the CA coordination site should decrease the lanthanide(III) size affinity while the decreased aromatic nature of the CA amine should decrease the lanthanide(III) complex stability compared to dipicolinic acid. Straight chain IDA and NTA have similar coordination sites as the piperidine based PDA and DPA, respectively, but the ring structure should make the coordination sites more rigid and increase the lanthanide(III) size selectivity of the ligands. Dipicolinic acid has a similar coordination site as CA but the increased aliphatic nature of CA should decrease the stability and size selectivity of lanthanide(III) complexes. The 7

stability constants determined by this study should indicate changes in ligand size affinity by comparison with the stability constants of IDA, NTA and dipicolinic acid across the lanthanide(III) series. Comparison of the thermodynamic (entropy and enthalpy) values of complexation determined by this work with the values for aminopolycarboxylate ligands can provide evidence for changes in the type or length of coordination bonds or net changes of the systems caused by complex formation. The coordination compounds of PDA, DPA and CA were examined by Ho(III) and Nd(III) hypersensitive spectroscopy and Eu(III) laser fluorescence.

The coordination nature of ligands with similar

coordination sites can be compared to show any difference caused by the rigid coordination geometry or other effects caused by the ring structures. These structural and thermodynamic differences caused by the additional ligand ring structures can be used to predict the properties of new ligand structures and suggest new compounds that might have improved lanthanide(III) ion size selectivity or enable improved lanthanide(III) and actinide(III) ion separation than current compounds.

8

HO HO

O

N

N

O

OH

HO O

O

O

NTA HO

OH

OH

N O HO

O

EDTA O

O N

N

N

O

HO O

OH

OH OH

O

DTPA

*

Figure 1.1.

Structures of NTA, EDTA and DTPA . * NTA = nitrilotriacetic acid; EDTA = ethylenediaminetetraacetic acid; DTPA = diethylenetriaminepentaacetic acid

9

N

OH

HO

O picolinic acid

N O O dipicolinic acid

Figure 1.2. Structures of α-picolinic acid and dipicolinic acid.

10

OH

HO HO

OH

N O

O

O

O O DPA

PDA

Figure 1.3. Structures of PDA and DPA.

11

OH

N

OH

A H

H

H

A

A H

N

N

A H

H

N

A

H

H

H

H

H

A

H

PDA A H

A H

N A

N

A H

H

N

A H

H

A

A

A

H

DPA

A

H H

Figure 1.4. Structural geometries and ring flip of the PDA, top, and DPA, bottom. Protons from meta and para carbon have been omitted for clarity and carboxy groups have been shown as A.

12

O

HO

OH

OH

N O

H

HO

O

O

Figure 1.5. Tautomer structures of chelidamic acid.

13

OH

N O

CHAPTER 2 EXPERIMENTAL

2.1 Chemical Reagents Water that had been purified by a reverse osmosis system and then de-ionized (Epure, Barnstead) was used for preparing aqueous solutions in all experimental procedures. Concentrated perchloric acid (60%, Fisher Scientific) and sodium hydroxide (Fisher Scientific) were reagent grade. Standard hydrochloric acid (certified, Fisher Scientific) and sodium hydroxide (certified, Fisher Scientific) solutions were used without dilution. The solutions for the calibration of the pH meters were pH 4.00 (± 0.01), 7.00 (± 0.01) and 10.00 (± 0.01) standard buffers (Fisher Scientific). 2.1.1 Lanthanide(III) Perchlorate Stock Solutions Solutions of lanthanide perchlorates were prepared by dissolution of the oxides of lanthanum (99.999%), neodymium (99.99%), europium (99.999%), holmium (99.999%) or lutetium (99.99%) (Aldrich Chemical) in concentrated perchloric acid. The solutions were fumed to near dryness three or more times to remove excess perchloric acid before dilution with de-ionized water. The metal concentrations were standardized by titration with an ethylenediaminetetraacetic acid (EDTA) standard solution (0.0946 M, Aldrich Chemical) using xylenol orange (Aldrich Chemical) as an indicator in acetate buffered solutions of pH 4.5 that contained a few drops of pyridine71. The solutions were warmed on a hot plate to 50 to 60°C prior to titration to reduce error caused by the slow kinetics 14

involved in the competition between the EDTA and the xylenol orange complexes. Typical solutions of lanthanide perchlorates were 100.0, 250.0 or 500.0 mL with known concentrations that ranged from 0.101 to 1.01 M. The lanthanide perchlorate stock solutions contained an unknown amount of excess perchloric acid from the dissolution procedure. The perchloric acid concentration were determined by titration with a standardized sodium hydroxide solution using a calibrated pH electrode to monitor the acidity change. An endpoint, pH 6.0, was used for the perchloric acid titration to avoid lanthanide hydrolysis that typically begins around pH 6.0 to 6.5. The perchloric acid concentrations in the lanthanide perchlorate stock solutions were calculated as the average value of five or more titrations and were reported with three significant figures. 2.1.2 Sodium Perchlorate Stock Solutions Stock solutions of sodium perchlorate (reagent grade, Aldrich Chemical or Acros Organics) were prepared in volumetric flasks. Samples of sodium perchlorate, weighed on an analytical balance, were dissolved in de-ionized water. The solutions were filtered through a 0.45 µm nylon filter membrane (Whatman) before use. Typical stock solutions were 500.0 mL or 1.000 L of known concentrations of sodium perchlorate that ranged from 6.00 to 8.50 M. 2.1.3 Ligands The purity of the ligands was checked by determination of the concentration of a solution by potentiometric titration with a standard base solution.

The calculated

concentrations indicated that the purity of each ligand was ≥ 99%. 2.1.3.1 Chelidamic Acid Chelidamic acid, CA, (97%, Aldrich Chemical) was used without further purification. 2.1.3.2 Piperidine-2,6-dicarboxylic Acid Piperidine-2,6-dicarboxylic acid, PDA, was synthesized by Dr. Paul Rickert at Argonne National Laboratory using the procedure of Anderson and Saito72 that consists the hydrogenation of pyridine-2, 6-dicarboxylic acid using a platinum oxide catalyst. The material was used without further purification. 15

2.1.3.3 2,6-Dicarboxypipepridine-N-acetic Acid 2,6-Dicarboxypiperidine-N-acetic acid, DPA, was synthesized by Dr. Paul Rickert at Argonne National Laboratory using the procedure of Kundra and Thompson73 that consists of the reaction of chloroacetic acid with PDA. The material was used without further purification. 2.1.4 Potentiometric Titration Solutions 2.1.4.1 Sodium Hydroxide Standard Solutions A 4.95 M sodium hydroxide stock solution was prepared by dissolving 396.11g of sodium hydroxide in a 2.000 L volumetric flask with de-ionized water. All other sodium hydroxide solutions were made by dilution of this 4.95 M stock solution. To prepare the sodium hydroxide solutions, sodium perchlorate stock solution was added so that the sum of the concentrations of sodium perchlorate and sodium hydroxide would give the desired ionic strength. The sodium hydroxide solutions were standardized by titration with potassium hydrogen phthalate (A.C.S. acidimetric standard, Aldrich Chemical) that had been dried at 70.0˚C in an oven for over a week. Phenolphthalein (A.C.S. reagent, Aldrich Chemical) was used as the indicator. Stock solutions were typically 500.0 mL or 1.000 L with a known concentration ranging from 0.020 to 0.100 M sodium hydroxide with an ionic strength of 0.10, 0.50 or 2.00 M. All standardized sodium hydroxide solutions were stored under nitrogen gas (DOC grade, Airgas). Prior to use, the nitrogen was bubbled through solutions of 4.00 M sodium hydroxide solution and sodium perchlorate solution at the ionic strength of the standardized solution. The sodium hydroxide prevented carbonate contamination of the base solution by removing carbon dioxide from the nitrogen gas. Bubbling through a sodium perchlorate solution saturated the nitrogen with moisture in an effort to reduce evaporation caused by the flow of the gas through the titrant and other experimental solutions. 2.1.4.2 Perchloric Acid Standard Solutions All perchloric acid stock solutions were made from concentrated perchloric acid by dilution with de-ionized water in a volumetric flask.

The perchloric acid stock

solutions were standardized by titration with standardized sodium hydroxide solutions 16

using phenolphthalein as the indicator. The stock solutions were typically 250.0 or 500.0 mL of perchloric acid with known concentration ranging from 0.200 to 0.500 M. 2.1.4.3 Ligand Solutions Stock solutions of CA, PDA or DPA were prepared in volumetric flasks by dissolution of a sample of known mass. A stock solution was typically 100.0, 250.0 or 500.0 mL with known concentration ranging from 5.00 to 20.00 mM for PDA or 3.00 to 5.00 mM for CA or DPA. Experimental solutions of each ligand were made by dilution from the corresponding stock solution. Aliquots of a sodium perchlorate stock solution were added to obtain the desired total ionic strength. A typical experimental solution was 10.00 or 20.00 mL with a ligand concentration that ranged from 2.00 to 10.00 mM PDA or 2.00 to 5.00 mM CA or DPA with an ionic strength of 0.10, 0.50 or 2.00 M. 2.1.4.4 Lanthanide(III) Complex Solutions Stock solutions of CA, PDA or DPA were prepared as described in 2.1.4.C. Stock solutions of the desired lanthanide perchlorates were prepared as described in section 2.1.1. Experimental solutions of the desired lanthanide to ligand concentration ratio were made by dilution of the corresponding ligand and lanthanide perchlorate stock solutions. Aliquots of a sodium perchlorate stock solution were added to obtain the desired ionic strength. Typical experimental solutions were 10.00 or 20.00 mL with known ligand concentrations ranging from 2.00 to 10.00 mM PDA or 2.00 to 5.00 mM CA or DPA, and Ln(III) ion concentrations ranging from 1.00 to 5.00 mM with an ionic strength of 0.100, 0.500 or 2.00 M. 2.1.5 Laser Induced Fluorescence Solutions Stock solutions of CA, PDA or DPA were prepared as stated in section 2.1.4.C and the europium perchlorate stock solution was prepared as described in 2.1.1. Experimental solutions of the desired Eu(III):ligand concentration ratio were made by dilution of these ligand and europium perchlorate stock solutions. Aliquots of a sodium perchlorate stock solution were added to create the desired ionic strength. Typical experimental solutions were prepared in 25.00 or 50.00 mL volumetric flasks 17

with concentrations ranging from 1.00 to 2.00 mM Eu(III) ion, 0 to 10.00 mM PDA or 0 to 5.00 mM CA or DPA with an ionic strength of 2.00 M. The pH was adjusted with 2.00 M sodium hydroxide and 2.00 M perchloric acid solutions. Some samples were prepared with constant Eu(III) ion and ligand concentrations while the pH varied between 2.00 and 6.00. Other samples were prepared with constant Eu(III) ion concentration and pH between 5.00 and 6.00 with varying ligand concentrations. A reference solution of 0.100 M europium perchlorate solution was prepared at pH 3.50 and 2.00 M ionic strength. 2.1.6 Hypersensitivity Spectroscopy Solutions Stock solutions of CA, PDA or DPA were prepared as stated in section 2.1.4.C and the holmium and neodymium perchlorate stock solutions were prepared as stated in section 2.1.1. Experimental solutions of the desired lanthanide to ligand concentration ratios were made by dilution of these ligand and lanthanide perchlorate stock solutions to which aliquots of a sodium perchlorate stock solution were added to obtain the desired ionic strengths. Typical experimental solutions were prepared in 25.00 mL volumetric flasks with concentrations ranging from 0 to 5.00 mM Ho(III) ion, 0 to 5.00 mM Nd(III) ion, 0 to 10.00 mM PDA and 0 to 5.00 mM CA or DPA with an ionic strength of 2.00 M. The pH was adjusted using 2.00 M sodium hydroxide and 2.00 M perchloric acid solutions. Some samples were prepared with constant Ln(III) ion and ligand concentrations while the pH was varied between 2.00 and 6.00. Other samples were prepared with constant Ln(III) ion concentration and a pH between 5.00 and 6.00 while the ligand concentrations were varied. 2.1.7 Nuclear Magnetic Resonance Solutions 2.1.7.1 Solutions for Adjusting pH A 0.50 M stock solution of perchloric acid was prepared by dilution of concentrated perchloric acid with deuterium oxide (Cambridge Isotope Laboratory). The perchloric acid introduced a proton impurity into the deuterium solvent that would appear as a peak normally associated with water in 1H NMR spectra. The molecules of this study had peaks ranging from 1.0 to 4.5 ppm with no significant overlap with the water 18

peak which typically ranged from 5 to 5.5 ppm in 1H NMR spectra. Solutions were stored in a desiccator when not in use to minimize additional proton contamination from atmospheric water since an excessively large water proton peak could cause a loss of sensitivity for other 1H NMR peaks A 0.50 M sodium deuteroxide solution was prepared by dilution of a 40% sodium deuteroxide solution (Aldrich Chemical) with deuterium oxide. Solutions were stored in a desiccator when not in use to reduce contamination from atmospheric carbon dioxide and water vapor. 2.1.7.2 Ligands Solutions Solutions of CA, PDA or DPA were prepared in 10.00 mL volumetric flasks by dissolution of a sample of known mass in deuterium oxide. The sodium salt of 3(trimethylsilyl)-1-propanesolfonic acid, DSS, (98%, Aldrich Chemical) was added to each 1H NMR sample as a 1H peak position reference. The typical concentrations of the ligands ranged from 4.00 to 5.00 mM for DPA, 10.00 to 15.00 mM for PDA or 5.000 mM for CA. Two samples of each ligand were prepared for pH titration experiments. Perchloric acid was added to the first sample to create the 0.50 M ionic strength and the solution was titrated with 0.50 M sodium deuteroxide. Sodium deuteroxide was added to the second sample to create the desired 0.50 M ionic strength and the solution was titrated with 0.50 M perchloric acid. This approach was used to minimize any NMR sensitivity problems due to the dilution that occurs during the course of the titration. 2.1.7.3 Lanthanide(III) Complex Solutions Lanthanum, lutetium and europium perchlorate stock solutions were prepared as described in section 2.1.1. Solutions of CA, PDA or DPA were prepared in 100.0 mL volumetric flasks by dissolution of a sample of known mass in deuterium oxide. Aliquots of the lanthanum or lutetium perchlorate stock solutions were added to obtain the desired Ln(III) ion concentration. The sodium salt of DSS was added to each 1H NMR sample to provide a 1H peak position reference. Typical concentrations of the ligands ranged from 4.00 to 5.00 mM for DPA, 10.00 to 15.00 mM for PDA or 5.00 mM for CA. Two different concentrations of the Ln(III) ions were used for each ligand so that the mole ratio of metal to ligand ratio was 1:1 or 1:2 for the particular ligand solution. Aliquots of 19

perchloric acid were added so that the total ionic strength of the solution was 0.50 M. The solution was titrated with 0.50 M sodium deuteroxide. 2.1.8 Crystals DPN and PDA were dissolved into de-ionized water. The solutions were covered and were stored at room temperature to allow slow evaporation of the solvent. After several weeks, crystals were collected from each solution. The crystals were examined under a microscope so that single crystals could be extracted and mounted for diffractometer measurements. 2.1.9 Calorimetry Solutions 2.1.9.1 Standard Solutions A 0.1000 M hydrochloric acid solution was used as the titrant in all standardization titrations. A 0.1000 M sodium hydroxide solution was boiled for 30 minutes to remove carbon dioxide contamination before calorimetric titration. A 0.1 M solution of tris(hydroxymethyl)aminomethane (reagent grade, ICN), THAM, was prepared in a 250.0 mL volumetric flask by dissolution of a sample of known mass. An aliquot of hydrochloric acid was added so that 10% of the THAM was protonated. Scrubbed nitrogen gas (section 2.1.4.A) was bubbled through the THAM solution for 30 minutes to remove carbon dioxide contamination before titration. 2.1.9.2 Ligand Solutions Solutions of CA, PDA or DPA were prepared in a 100.0 or 250.0 mL volumetric flask by dissolution of a sample of known mass. Aliquots of a sodium perchlorate stock solution were added so that the total ionic strength was 0.50 M. Sodium hydroxide was added until completely neutralization of the ligand to measure the heat of protonation. The solutions used in the measurement of the heat of complexation were buffered with 4morpholineethanesulfonic acid, MES, (Aldrich Chemical) which does not form complexes with Ln(III) ions and does not contribute to the ionic strength of a solution. Fresh buffer stock solutions were made weekly because MES is biologically active. The MES concentration of each solution was 0.100 M. A 0.100 M sodium hydroxide solution was used to adjust the pH to between 5.30 and 5.50. Typical solutions were prepared

20

with known concentrations ranging of 15.00 mM for PDA or 10.00 to 20.00 mM for CA or DPA. 2.1.9.3 Lanthanide(III) Perchlorate Solutions Lanthanide(III) perchlorate stock solutions were prepared as described in 2.1.1. The Ln(III) ion solutions were prepared by dilution from the corresponding stock solution. Aliquots of sodium perchlorate stock solution were added to obtain a total ionic strength of 0.50 M. The MES buffer concentration of each sample was 0.100 M and the pH was adjusted to between 5.30 and 5.50 with 0.10 M sodium hydroxide solution. Solutions were typically 100.0 mL with a known concentration of the Ln(III) ion ranging from 1.00 mM to 3.00 mM. 2.2 Equipment 2.2.1 Mass Determination A Fisher Scientific model XA-100 analytical balance was used for all weight determinations under 100 g. A Fisher Scientific model XL-3000 analytical balance was used for all weight determination above 100 g. The mass of samples was determined to at least 4 significant figures. 2.2.2 Potentiometric Measurements All pH measurements of bulk solutions were performed using a Fisher Scientific model 950 pH/ion meter equipped with an external temperature sensing probe. The electrode was either a Corning micro combination Ag-AgCl reference glass electrode or a Corning semi-micro combination Ag-AgCl reference glass electrode. The potassium chloride electrolyte solution in the electrodes was replaced with 1.00 M sodium chloride for the micro electrode and saturated sodium chloride for the semi-micro electrode. The potassium chloride electrolyte solution was replaced to prevent potassium perchlorate precipitation in the junction when the electrode is used in high molarity sodium perchlorate solutions. All measurements were made at room temperature, 23 ± 1°C. 2.2.3 Potentiometric Titrations The potentiometric titration system was designed and built in the laboratory and is an updated model of the instrument described by Crooks74 and Redko75. The experiments were performed under nitrogen gas that had been bubbled through solutions of 4.0 M 21

sodium hydroxide to remove carbon dioxide and sodium perchlorate of the same ionic strength as the experiments to saturate the gas with water vapor. Each titration was performed with constant stirring using a Corning model PC 420 stirrer/hot plate and constant temperature, 25.0 ± 0.1°C, regulated by a NESLAB model RTE 100 thermostated bath. A Metrohm Dosimat 665 burette was used to titrate additions to the experimental solution. An Orion Ross™ epoxy combination electrode filled with a solution of saturated sodium chloride was connected to a Keithley model 195A digital meter to monitor the pH of the experiment solution. The system was interfaced to a computer that operated software prepared in the laboratory to control burette additions and to record the mV reading of the electrode. The software has a measurement cycle that is designed to ensure that the sample had adequate time to reach equilibrium. The readings, measured after a delay time that ranged between 30 and 120 s, were accepted if they agreed within the user selected “electrode stability” parameter, typically 0.1 mV. After two consecutive readings of the electrode that agreed within the “electrode stability” parameter, the data point was recorded and the system made the next titration addition. If consecutive readings failed to agree within the “electrode stability” parameter by the tenth electrode reading, the tenth electrode reading was recorded and the system made the next titrant addition. After the addition was made, the system waited a mixing time, typically 30 s, before restarting the measurement cycle. Several experiments were performed with a significantly longer delay time. The results from these experiments were compared to experiments with shorter delay times to ensure that the shorter times were adequate for the system to reach equilibrium. This approach ensured that data from the shorter delay time experiments were not due to very slow kinetics or from readings that failed to meet the electrode stability criteria. 2.2.4 Laser Induced Fluorescence 2.2.4.1 Fluorescence Spectra Fluorescence experiments were performed on solutions in a 1 cm quartz cell using a Nd-YAG laser (Quanta Ray DCR 2A, Spectra Physics) coupled to a pumped dye laser (Quanta Ray PDL2, Spectra Physics). The laser dye was a 1:1 solution of Rhodamine 22

590 tetrafluoroborate (Exciton) and Rhodamine 610 perchlorate (Exciton) in methanol (HPLC grade, Fisher). The concentration of the amplifier was 0.100 mM for both dyes while that of the oscillator was 0.016 mM for both dyes. The dye was pumped by the second harmonic output of the Nd-YAG laser at 532 nm which gave a laser power of approximately 15 mJ at 580 nm.

The excitation monochromator (Jarell-Ash) was

controlled with a computer interfaced stepper motor that allowed scanning of the dye laser emission throughout the entire range of the emission wavelength of the dye. The fluorescence was detected perpendicular to the incident beam using a Hamamatsu R928 photomultiplier tube. The system was operated by software that had been written by Dr. L. J. van de Burgt of the FSU laser laboratory which controlled the stepping motor of the laser and recorded the peak position and intensity.

The stability of the laser was

measured at various times during the experiment using a calibrated Scientech model 380101 volume calorimeter attached to a Laser Precision model RT-7620 energy ratiometer. 2.2.4.2 Fluorescence Lifetime Measurements Lifetime measurements were performed at a fixed wave number corresponding to the maxima of the fluorescence peak intensity of each spectrum. Fluorescence decay curves were collected using a LeCroy 9410 Dual 150 MHz oscilloscope. The data was transferred to a computer using software supplied with the oscilloscope. The software had been modified by Dr. van de Burgt. 2.2.5 Hypersensitivity Spectroscopy Ultraviolet-visible spectrophotometric measurements of Nd(III) and Ho(III) ions were made in 10 cm glass cells using a dual beam Cary 14 UV-visible spectrophotometer that has been upgraded by On Line Instrument Systems (OLIS).

The system was

interfaced with a computer to allow automated acquisition and storage of data. The light source was a tungsten lamp and the signal was collected by a photomultiplier tube. 2.2.6 NMR All 1H NMR titration experiments were performed with a Varian INOVA 500 MHz NMR spectrometer using a 5 mm quad nucleus detection probe. All 1H decoupled 13

C NMR spectra and homonuclear decoupled 1H NMR spectra were measured using a 23

Bruker AC 300 MHz NMR spectrometer using a 5 mm dedicated 1H/13C probe. The NMR tubes were not spun during any experiments. 2.2.7 Crystallography Crystallography was performed on a Bruker SMART APEX diffractometer at 25.0°C using a CCD detector. 2.2.8 Calorimetry The calorimeter, designed and built in the laboratory, is an updated model of the calorimeters described by Orebaugh76, Ensor77 and Caceci78. The water bath was kept at 25.000ºC by competition between a cooling coil connected to a Fisher Scientific model 9500 Isotemp refrigerated circulating water bath and a heating element controlled by a Tronac PTC-49 temperature controller which maintained the bath to ±0.001°C79. The calorimetry cup was a gold plated glass beaker inside a copper housing. The instrument operated in a semi-adiabatic mode using a Melcor model FC0.6-32-06L Peltier heating/cooling device in contact with the exterior of the gold plated beaker to stabilize the temperature during the course of a titration. After a measurement was made, the Peltier heating/cooling device was used to reestablish the original operating temperature of the cup through minor heating or cooling to remove the thermal effects of the reaction. Calibration was performed after every fourth or fifth measurement during the titration by a Tronac model 87-249 heating element that was connected to a Power Designs model 2005 precision power source.

Temperature readings were made with a YSI model

SP033-108 thermistor connected to a Keithley model 181 nanovoltmeter. The Peltier heating/cooling device, heating element and thermister were housed inside the calorimetry cup. A motor, mounted to the top of the apparatus was connected to a long, bladed glass rod, stirred the solution in the calorimetry cup.

The instrument was

interfaced to a computer and managed by software written by Dr. M. Caceci76 that controlled the Radiometer Copenhagen ABU-80 Autoburette, calibration heating element and Peltier heating/cooling device and recorded the nanovoltmeter reading. The average temperatures were measured by the software and typically had a standard deviation less than 20 x 10-6°C after several minutes of data collection. The program ran multiple

24

titrations for each experiment and reported both the individual values and average value of the heat per addition. 2.3 Procedures 2.3.1 Potentiometric Measurements Electrodes were calibrated with pH 7.00 and 4.00 buffers for measurements in acidic media and pH 7.00 and 10.00 buffers for measurements in basic media. The efficiency of the electrode calibration was maintained above 0.95. Electrodes were allowed to equilibrate for 1 day prior to use after the sodium chloride solutions were added as suggested by Corning. Electrode readings were taken after the measurement had stabilized as shown by the indication on the meter. 2.3.2 Potentiometric Titrations The sample holder was washed several times with de-ionized water and dried thoroughly before each experiment. The solutions were placed in the thermal jacketed sample holder that was connected to a temperature bath and were allowed several minutes to reach thermal equilibrium. The stirrer was set at the same setting for all experiments.

A 0.1 mV value was selected as the maximum difference between

consecutive readings for all experiments. 2.3.2.1 Standardization The electrode was calibrated by the titration of a standardized strong acid with a standardized strong base to provide a mV to pcH conversion equation. Aliquots of a standardized perchlorate acid solution and of a sodium perchlorate stock solution were added to de-ionized water to create 10.00 to 20.00 mL solutions with known concentrations. These solutions were typically 0.00500 to 0.02000 M perchloric acid with an ionic strength of 0.100, 0.500 or 2.00 M. This solution was titrated with the standardized base of the same ionic strength with no less than 10 additions to complete the titration using typical addition sizes ranging from 0.050 to 0.200 mL.

The

concentrations of the acid and the burette addition size were chosen so that approximately half the data points in the titration curve were in the acidic and half in the basic regions. The mixing time after titrant addition was 30 seconds. The delay between the readings

25

was 30 s. Standardization titrations were conducted once for every 1 to 3 experimental titrations. 2.3.2.2 Acid Dissociation Constant Determination Aliquots of ligand and sodium perchlorate stock solutions were combined to create 10.00 to 20.00 mL of a solution with known concentration.

The ligand

concentration of these solutions ranged from 3.00 to 10.00 mM for PDA or 3.00 to 5.00 mM for CA and DPN. The experiments were performed at ionic strengths of 0.100, 0.500 and 2.00 M. The burette addition size, typically 0.030 to 0.100 mL, was chosen so that 80 additions would cover a pH range from 2.0 to 11.5. The mixing time after titrant addition was 30 seconds. The delay between the readings was 30 or 120 s. 2.3.2.3 Stability Constant Determination Aliquots of ligand, lanthanide(III) perchlorate and sodium perchlorate stock solutions were combined to create 10.00 to 20.00 mL of a solution with known concentration. The concentrations of these solution ranged from 3.00 to 10.00 mM for PDA, 3.00 to 5.00 mM for CA or DPA and 2.00 to 5.00 mM for Ln(III) ions. The experiments were performed at ionic strength of 0.100, 0.500 and 2.00 M. The burette addition size was chosen such that 80 additions would cover a pH range from 2.0 to 6.0 for most experiments and was typically 0.006 to 0.050 mL. This pH range was selected to avoid lanthanide hydrolysis and any data points with pH greater than 6.0 were discarded. When the concentrations of CA and DPA exceeded twice the concentration of the metal ion, lanthanide hydrolysis did not occur and so a few titrations of CA and DPA complexes were run with a pH range of 2.0 to 11.5. Other titration parameters were the same as those used in the determination of acid dissociation constants as described in 2.3.2.B. 2.3.3 Laser Induced Fluorescence 2.3.3.1 Fluorescence Experiments The 1 cm quartz fluorescence cell was rinsed three times with the sample before being filled with the sample solution. The room light was reduced during the experiment to minimize background scattered light. The laser was run for 10 minutes before use in experimental measurements to allow for operational equilibrium. The laser power was 26

determined by directing the path of laser beam into a volume calorimeter80 and equilibrium was defined as a constant energy reading over 30 s. The energy was checked periodically during use and was found to remain constant. Excitation spectra were collected as the average of 50 pulses per data point over a range of 17210 to 17320 cm-1 in 1 cm-1 increments. A europium standard was run with each experiment to calibrate the wave number of the laser emission using the position of the peak of the fully hydrated europium ion. 2.3.3.2 Fluorescence Lifetime Experiments Samples were measured in a similar fashion to the fluorescence experiments except that the dye laser was tuned to a fixed wave number corresponding to the maxima of the peaks from the fluorescence spectra. The fluorescence decay curve was the average of 1000 fluorescent decay measurements. 2.3.4 Hypersensitivity Spectroscopy The Cary 14 was turned on at least 30 minutes prior to use to allow operational equilibration as determined from comparison of the intensity of a standard neodymium solution over time at selected wavelengths. Experiments were performed in a set of 10 cm glass cells made in the departmental glass shop. The cells were labeled to ensure similar orientation in all experiments. The reference cell was filled with de-ionized water and placed in the reference compartment of the Cary 14. The same cell was used as the reference throughout all experiments. The sample cell was emptied and rinsed three times with the sample solution before being completely filled with this solution and placed in the Cary 14 sample compartment. The spectra were measured over the range of 425 to 475 nm for holmium and 555 to 605 nm for neodymium in 1 nm increments. Each data point is the averaged value of 50 readings. Each experimental set included one solution of the completely hydrated, uncomplexed Ln(III) ion to ensure standardization of peak position and oscillator strength between different measurements. 2.3.5 NMR 2.3.5.1 pH Titrations The sample was placed in a Kontes 5 mm glass NMR tube. The pH was adjusted by small additions of the acid or base solution and was measured by a calibrated micro 27

electrode inserted directly into the NMR tube. The samples were placed in the Varian INOVA 500 MHz NMR spectrometer sample compartment that was operated at 25.0 ± 0.1°C. The 2H signal was used to optimize the magnet homogeneity and was used as a lock signal by the spectrometer during 1H measurements. The spectra were measured using the standard 2 pulse experiment protocol of the Varian software with the first pulse width set to 0 µs and the delay times set to 0 s. The second pulse width was set to the experimentally determined 90° pulse width, 22.2 µs. The acquisition time was 2.048 s. The data was collected as 16, 32 or 64 transients to produce spectra with a spectral width of 4000 Hz and a digital resolution of 0.244 Hz. All of the decouplers were turned off. The chemical shift scale was set at 0 ppm using the reference peak of DSS. The acquisition time was considered an adequate relaxation time since this experiment was measuring the change in chemical shift of each peak and not the integration of the individual peaks. 2.3.5.2 Coalescence Experiments Each NMR peak of DPA broadened and lost mutliplicity between pH 8 and 9. The peaks from DSS showed no change in this pH region. The broadening and loss of multiplicity in the DPA peaks showed the characteristics of a coalescence peak for an equilibrium involving kinetics that were occurring on a similar time scale as the NMR measurements.

Experiments were performed similarly to those of the pH titration

experiments in section 2.3.5.A except that the temperature was varied between 25.0 and 5.0°C in 5.0°C increments in an attempt to resolve the coalesced peaks. The instrument and sample were allowed 1 hour to equilibrate thermally after each temperature adjustment as suggested by Dr. T. E. Gedris of the FSU NMR laboratory. 2.3.5.3 Nuclear Overhauser Effect Experiments The nuclear Overhauser effect (nOe) spectra were collected in the same manner as in the pH titration experiments in section 2.3.5.A except the standard cyclic nOe experiment protocol of the Varian software was used.

These experiments were

performed on solutions of PDA and DPA at pH 2.00 and 11.00.

An additional

experiment was run on a solution of DPA at pH 9.00. Each measurement was preceded by 32 scans to allow the thermal excitation caused by the radio frequency excitation 28

pulses to reach a steady state before the collection of experimental data. The spectra were collected with an acquisition time of 2.048 s, a spectral width of 4000 Hz and a digital resolution of 0.244 Hz. Each excitation was performed by 70 pulses of the time interval, τ, of 0.10 ms as recommended by Varian. The frequency of the excitation pulses was set to the selected peak position with the correct multiplicity pattern set by the user. All decouplers were off. 2.3.5.4 13C 1H Decoupled Spectra These measurements were performed by Dr. Gedris on the Bruker AC 300 MHz NMR spectrometer. The 2H signal was used to optimize the magnet homogeneity and as a lock signal by the spectrometer during measurements. The spectra were collected with a pulse width of 4.9 µs, a spectral range of 17800 Hz and a digital precision of 1.090 Hz. Several thousand scans were collected to produce a spectrum. The 1H signals were suppressed in order to decouple the 1H and

13

C signals.

This decoupling removed

multiplicity from the 13C peaks, causing the signals to have greater intensity and allowing easier comparison of the peak positions. The experiments were performed at room temperature, 23 ± 1°C. 2.3.5.5 1H Homonuclear Decoupled Spectra The measurements were performed by Dr. Gedris on the Bruker AC 300 MHz NMR spectrometer at room temperature, 23 ± 1°C. The 1H spectra were collected with a pulse width of 5.0 µs, a spectral range of 3200 Hz and a digital precision of 0.391 Hz. A 1

H spectrum was collected while an individual 1H peak was suppressed.

For each

molecule, a series of spectra were collected with the sequential suppression of each individual 1H peak. In some cases, a 1H spectrum was collected while multiple 1H peaks were suppressed. 2.3.6 Crystallography The crystallographic experiments were performed by Dr. Ronald J. Clark of the FSU Chemistry Department. A crystal of the sample was mounted on a nylon loop and centered in the diffractometer. The detector distance was 5 cm. The X-ray diffraction pattern was collected at a specific incident angel over 10 s as a data set called a frame. Each sample was measured with 1850 frames. Measurements were made at an incident 29

angle up to 28°. The integration of the frames was performed using the program SAINT (Bruker, 1999). No absorption correction was used. The structure was solved by direct methods and refined by the program SHELXTL (Bruker, 2000). The atoms other than hydrogen were refined anisotropically and the hydrogen atoms were assigned by least squares refinement. 2.3.7 Calorimetry The water bath temperature and heating element were monitored and adjusted for several weeks using a thermometer to ensure the temperature was stable at 25.000 ± 0.001ºC. All titrations were performed at this temperature. The burette was filled with the titrant and the system was flushed thoroughly to remove all air bubbles. A small air bubble was left in the tip of the burette to prevent diffusion of the titrant into the sample. The sample was placed in the calorimetry cup and, after assembly, the calorimetry cup and burette tubing were submerged in the water bath for at least 10 hours of stirring to allow thermal equilibration. The thermister was connected to a nanovoltmeter by a Wheatstone bridge. The Wheatstone bridge was used to adjust the zero point of the thermister to the temperature of the cup. A titration of water into water was performed with 25 or 50 additions of 0.100 mL to check the zero point setting of the Wheatstone bridge. After each titration, the Wheatstone bridge was adjusted to set the zero point to as close to 0 mJ per addition as possible. A sample of the sodium hydroxide solution was placed in the calorimetry cup before assembling the apparatus to allow it to come to thermal equilibrium after placement in the bath. Standardization titrations were performed after both 10 hours and 2 days of thermal equilibration time. These results of the titrations agreed within error, indicating that 10 hours was adequate time for thermal equilibrium of the calorimetry cup apparatus to occur. 2.3.5.1 Heat of Dilution The calorimetry cup was filled with 50.00 mL of a solution with the same ionic strength, pH and MES concentration as the titrant solution. The system was immersed and allowed to come to thermal equilibrium for at least 10 hours before a titration. The solution was titrated by 25 or 50 additions of 0.200 mL of the titrant. 30

2.3.7.2 Standardization Titrations of 0.1000 ± 0.0005 M standard sodium hydroxide solution by 0.1000 ± 0.0005 M standard hydrochloric acid solution were performed to calibrate the instrument by measuring the heat of neutralization. The calorimetry cup was filled with 50.00 mL of the standard sodium hydroxide solution and placed in the water bath for at least 10 hours with stirring to reach thermal equilibrium. The solution was titrated by 25 or 50 additions of 0.100 or 0.050 mL of the standard hydrochloric acid solution. The average value obtained from the standardization titrations was -52.4 ± 2.1 kJ/mole and was within two standard deviations of the value reported by Grenthe81 of -55.84 ± 0.01 kJ/mole. An additional standardization was performed using THAM. A 50.00 mL aliquot of the 0.1 M THAM solution was placed in the calorimetry cup and was titrated by 0.1000 ± 0.0005 M hydrochloric acid. The change in enthalpy was determined to be -45.6 ± 1.0 kJ/mol for the proton addition and was within two standard deviations of values found in the literature, -47.53 kJ/mol82, -47.48 kJ/mol83 and –47.36 kJ/mol84. 2.3.5.3 Heat of Protonation A 50.00 mL aliquot of a solution of the neutralized ligand, as discussed in section 2.1.9.B, was added to the calorimetry cup and allowed to thermally equilibrate for at least 10 hours. The system was titrated with 0.1000 ± 0.0005 M perchloric acid at the same ionic strength as the ligand solution. 2.3.5.4 Heat of Complexation The calorimetry cup was filled with 50.00 mL of the Ln(III) ion solution. The burette was filled with the ligand solution with the same ionic strength, pH and MES concentration. The calorimetry cup was immersed and allowed to come to thermal equilibrium for at least 10 hours. The solution was titrated with 25 or 50 additions of 0.100 mL of the ligand solution in studies measuring the formation of the ML species or both the ML and ML2 species.

31

CHAPTER 3 DATA TREAMENT

3.1 Potentiometric Titrations

3.1.1 Electrode Calibration The electrode used for acid dissociation constant and stability constant determinations was calibrated by titration of a standardized perchloric acid solution by a standardized sodium hydroxide titrant.

The concentration of the hydrogen ion was

calculated for each titrant addition from the concentrations and volumes of the acid and base solutions, the total volume of the solution and the water ionization constant for the specific ionic media85.

The electrode readings plotted versus –log (hydrogen ion

concentration), pcH, from an acid-base standardization titration was fitted by Excel 98 (Microsoft) using the equation: y = mx + b

Eq. 3.1

where y was the mV reading of the electrode and x was the calculated pcH value. The data points between pcH 4.5 and 9.5 were excluded from the linear fit of calibration titrations as is the standard practice86. Data in this pcH range is excluded because small errors in the concentrations and volumes of reagents could have large effects on the electrode reading.

All of the linear fits of acid-base calibration had correlation

coefficients better than 0.9900.

32

The linear fit from calibration titrations was used to calculate the pcH from the mV readings in titrations of the ligands and the metal ligand complexes. The slope determined from each acid-base standardization titration at the same ionic strength was nearly constant, having a standard deviation of less than 0.1 %. Therefore, the average value of the slope was used in the mV to pcH conversion for a given ionic strength. Significant fluctuations were seen in the intercept value of the linear fit, ranging from 6.0 to 8.3. The intercept values measured on a particular day at the same ionic strength tended to agree well so the pcH was calculated using the average of these daily values. 3.1.2 Acid Dissociation Constant Determination The stepwise acid dissociation equilibrium of DPA can be described by the equations: L-m + H+ = HL1-m

Eq. 3.2

HL1-m + H+ = H2L2-m

Eq. 3.3

H2Lm + H+ = H3L3-m

Eq. 3.4

m

+

4-m

H3L + H = H4L

Eq. 3.5

where L is ligand and the m = 3 for DPA. The corresponding acid dissociation constants are defined as:

Ka1 =

[HL] [H][L]

Eq. 3.6

Ka 2 =

[H 2 L] [H][HL]

Eq. 3.7

Ka 3 =

[H 3 L] [H][H 2 L]

Eq. 3.8

Ka 4 =

[H 4 L] [H][H3 L]

Eq. 3.9

where Ka is the concentration acid dissociation constant, the species concentrations are in molarity, the species charges have been omitted. CA and PDA have 3 acidic sites so the stepwise acid dissociation equilibrium can be described by Eq. 3.2, 3.3 and 3.4 with m = 2 and using the corresponding acid dissociation constants, Eq. 3.6, 3.7 and 3.8. These

33

constants were calculated from the potentiometric titration of the ligand with a standardized base using PSEQUAD. 3.1.3 Lanthanide(III) Complex Stability Constant Determination

The formation of first and second Ln(III) complexes are described by the equations: M3+ + Lm-3 = [ML]3+m-3

Eq. 3.10

M3+ + 2 Lm-3 = [ML2]2m-3

Eq. 3.11

where M is the Ln(III) ion, L is the ligand and m = 0 for DPA or m = 1 for PDA and CA. The corresponding stability constants are defined as: β1 = K 1 =

K2 =

[ML] [M][L]

[ML2 ] [ML][L]

β 2 = K 1K 2 =

[ML 2 ] [M][L]2

Eq. 3.12 Eq. 3.13 Eq. 3.14

where K1 and K2 are the concentration equilibrium constants for the stepwise formation of the complexes, β1 and β2 are the stability constant of the ML and ML2 species, the species concentrations are in units of molarity and the charges of the species have been omitted. These constants were calculated from the potentiometric data obtained from the titration of the ligand complex solution with a standardized base using the PSEQUAD software. 3.2 PSEQUAD

Zékány and Nagypál87 developed the Fortran based computer program PSEQUAD for the evaluation of potentiometric and/or spectrophotometric equilibrium data using analytical derivatives. This program is designed to use a Newton-Raphson procedure to calculate the values of the mass balance and a Gauss-Newton method to refine equilibrium constants and molar absorptivities. The mass balance is then fitted to the experimental titration curve using the selected chemical model. The non-interactive program can fit multiple data sets simultaneously and can give statistical information such as the partial, total and multiple correlation coefficients for the fitted parameters as 34

well as a statistical error on the determined values. PSEQUAD has a statistical parameter that can give an indication of how the selected chemical model of the system fitted the experimental data. 3.3 Laser Induced Fluorescence 3.3.1 Fluorescence Lifetime Measurements

The intensity of the fluorescence emission of the 7F0→5D0 excitation of Eu(III) has been shown to be inversely proportional to the number of water molecules in the inner hydration sphere of the ion59,62. The dipole moment of an oxygen-hydrogen bond can oscillate with a similar energy as the excited state of the Eu(III) ion and so water and other molecules that contain an oxygen-hydrogen bond can provide a possible quenching mechanism of the Eu(III) fluorescence.

It has been shown that the lifetime of the

fluorescence is proportional to the number of water molecules in the inner hydration sphere of the ion62,88. The lifetime of the fluorescence decay of the various Eu(III) species was fitted using Sigma Plot 4.0 (Jandel Scientific) with the mono-exponential decay equation: y = yoe-kx

Eq. 3.15

where y was the fluorescence intensity, x was time in s, yo was the initial intensity and k was the fluorescence decay rate constant in s-1. All fluorescence decay curves had a correlation coefficient fit better than 0.9990. The number of water molecules in the inner coordination sphere can be calculated using the equation89: nH2O = 1.07k – 0.62

Eq. 3.16

where nH2O is the number of water molecules in the inner hydration sphere of the Eu(III) ion. The number of inner sphere water molecules was reported with a relatively high uncertainty, ± 0.5, due to the potential of a small degree of quenching of the Eu(III) excited state by oxygen-hydrogen bonds in the outer coordination spheres89. The number of coordination sites of the ion being occupied by the ligand was calculated as the difference between 9 and nH2O, assuming the fully hydrated Eu(III) ion has 9 water molecules in its primary coordination sphere90. 3.3.2 Fluorescence Spectra

35

Each peak in the fluorescence spectrum represents a different Eu(III) ion environment.

The fluorescence emission shifts to a lower wavenumber as water

molecules in the inner coordination sphere of the Eu(III) ion are removed due to the formation of a coordination complex. The calculated ligand coordination number can be calculated by the equation91: CNL = 0.237∆υ + 0.628

Eq. 3.17

where CNL is the total number of coordination bonds between the Eu(III) ion and ligand or ligands and ∆υ is the change in the fluorescence peak position relative to the 17276 cm-1 peak of the full hydrated Eu(III) ion. The coordination number of a ligand can be calculated from the total coordination number if the speciation of the system is known. The coordination numbers determined by this work have been reported with a relatively large error, ± 0.5, to account for quenching effects by water molecules in outer hydration spheres7. 3.4 Hypersensitivity Spectroscopy

The oscillator strength of integrated intensity, P, is defined by the equation92: P = 4.3198E- 09∫ ε (σ )dσ i

Eq. 3.18

where ε is the molar absorptivity at the energy σ(cm-1). The integral can be calculated by90:

∫ε

i

(σ )dσ =

S [M]c

Eq. 3.19

where S is the area of the hypersensitive transition spectrum in units of absorbance·cm-1, [M] is the total concentration of the Ln(III) ion in molarity and c is the path length of the cell in cm. This substitution of Eq. 3.19 for the integral reduces Eq 3.18 to:

P=

(4.3198E − 09)S [M]c

Eq. 3.20

The value of S was determined by integration of the area of the hypersensitive transition peak by a trapezoidal approximation method (GRAMS/32®).

36

A plot of P versus the ligand to metal concentration ratio, Figure 4.1 and 4.2, can be resolved into linear regions for systems in which the complexes are formed between the ligands and a single metal ion and the ligands have a sufficiently high affinity for the metal ion90. The first region can be defined by [L] : [M] < 1, limiting complexation to the formation of the ML species. The second region can be defined by the formation of the ML2 species, occurring when the values of the [L] : [M] ratio is between 1 and 2. The ML3 and ML4 species, if present, should have linear ranges occurring when the [L] : [M] ratio is between 2 and 3 and 3 and 4, respectively. A linear fit, Eq. 3.1, of P as a function of the [L] : [M] ratio was performed where x = [L] : [M] ratio and y = P. The linear fit of the first region was performed on [L] : [M] values between 0 and 0.7 to limit complex formation to the ML species to avoid the influence of the ML2 species on the value of P. The intercept of this linear fit is PM, the oscillator strength of the Ln(III) aqua ion. Another linear fit was performed for the region where the [L] : [M] ratio was between 1.3 and 1.7, a range where the formation of the ML2 species should be the major species formed. The value of the oscillator strength of the ML species, PML, was calculated from the intersection of the linear regressions of ML and ML2 regions. The oscillator strength of additional species can be calculated using this method with the linear regressions of the appropriate regions. At some [L] : [M] ratio, P approaches a constant value which indicates that the metal ion is not adding an additional ligand as the [L] : [M] ratio continues to increase. This constant value can be treated as the oscillator strength of the MLn species and n is the highest [L] : [M] integer ratio value where changes in P are still observed. 3.5 NMR 3.5.1 Acid Dissociation Constant Determination

The acid dissociation constants defined by the concentration of the species are discussed in section 3.1. Acidic equilibria are typically much faster than the NMR time scale, so the 1H NMR spectra would be expected to have peaks that are representative of the weighted average of the concentrations of the acid and conjugate base forms of the molecule. Acidic protons of a molecule rapidly exchange with the deuterium ions of deuterium oxide so these proton environments should not have peaks that appear in 1H 37

NMR spectra.

The concentration acid dissociation constants in this study were

determined from the chemical shift change of the non-exchangeable protons (protons covalently bonded to a carbon atom) caused by the acid/base equilibria of the molecule. NMR techniques can not differentiate equilibria involving the carboxylic acid groups in the 2 or the 6 positions of PDA and DPA because of the symmetry of these molecules. Therefore, the values of the acid dissociation constants for the 2,6-carboxylic acid groups for each ligand were reported as a single, average acid dissociation constant. Because these experiments were performed in deuterium solvents, the deuterium ion was substituted for the hydrogen ion in calculations of pcH. The value of pcD was calculated using the equation93: pcD = pHr + 0.40

Eq 3.21

where pcD is the –log (deuterium ion concentration) and pHr is the meter reading from a calibrated electrode. This equation accounts for the activity coefficient difference between the different isotopes of the hydrogen ion and so the values of pKa calculated with pcD and pcH should be approximately the same. The acid dissociation constants were determined by analyzing the change in the chemical shift of nonexchangable protons on the molecule versus pcD. The change in the chemical shift was fitted using the Hendersen-Hasselbach equation94:

pcD= pKax + log

[Hx-1L] [Hx L]

Eq. 3.22

where pKa is the negative log of the concentration acid dissociation constant, [HxL] and [Hx-1L] are the concentrations of the acidic and basic forms of the molecule in units of molarity and the species charges are not shown. Plots of the 1H NMR chemical shift versus pcD for PDA and DPA are shown in Figures 4.21 and 4.22. The molecules are undergoing acid equilibrium in the pcD regions of the plots where the chemical shift is changing. The [Hn-1L] : [HnL] ratio was calculated from the change in the chemical shift between the plateau regions using the equation:

[H n -1 L] ⎡ δ A − δ e ⎤ = [H n L] ⎢⎣ δ e − δ B ⎥⎦ 38

Eq. 3.23

where δA is the value of the constant chemical shift of HnL at lower pcD than the inflection region, δB is the value of the constant chemical shift of Hn-1L at a higher pcD than the inflection region and δe is the chemical shift of the experimental point in the inflection region between the plateaus of δA and δB. This reduces Eq. 3.22 to:

⎡ δA − δe ⎤ pKax = pcD - log ⎢ ⎣δe − δB ⎥⎦

Eq. 3.24

The pKa values were calculated for each data point between the selected plateau areas for each peak and then were averaged to obtain the peak pKa value. For peaks that showed a change of chemical shift over the same range, the peak pKa values were averaged to obtain the reported concentration acid dissociation constant. 3.5.2 Nuclear Overhauser Effect

The nuclear Overhauser effect (nOe) is a through-space coupling of the nuclear dipoles of protons on a molecule. This dipole coupling can provide a mechanism for the relaxation of an excited nuclide.

The coupling intensity can be described by the

equation95: ⎛ ρ* ⎞ 6 1 ∝ ⎜ ⎟r ηi (s) ⎜⎝ τc ⎟⎠

Eq. 3.25

where ηi(s) is the peak enhancement for the ith off-resonance peak caused by nOe when the resonance peak has been completely saturated, ρ* is a relaxation factor due to the dipole interaction, τc is the molecular correlation time and r is the distance between the two nuclei. The values of ρ* and τc are assumed to be constant because the variations of these paramaters are typically insignificant for different atoms on the same molecule95. The value of η for a peak was calculated from the integrated area of the off-resonance and resonance peaks, shown by95:

η=

I A = I0 A0

39

Eq. 3.26

where I is the intensity of the peak, A is the area of the peak and the subscript denotes values from the resonance peak. This leads to the relationship:

Ao ∝ r6 A

Eq. 3.27

which allows a comparison of the ratio of the resonance and off-resonance peaks to the distances between the proton environments those peaks represent. 3.6 Calorimetry 3.6.1 Change of Enthalpy

The speciation was controlled by using the ligand solution as the titrant, allowing the step-wise formation of the ML complex followed by the ML2 shown in Eq. 3.10 and 3.11. The heat change produced by the formation of each complex species can be used to calculate the change in enthalpy from the equation:

∆H n =

- qn c

Eq. 3.28

where ∆Hn is the change in enthalpy of the reaction in kJ/mole, qn is the heat per addition in kJ/L, c is the concentration of the titrant in mole/L and the subscript n refers to the formation of the first (n=1) or second (n=2) complex as given by Eq. 3.10 and 3.11. The ∆Hn value of each complex was calculated as the average from each titrant addition corresponding to the [L] : [M] ratio for the desired species. Each titration was repeated two or more times and the values were reported as the average of those experiments. The heat of dilution was determined for each titrant and used as a background correction factor for experiments performed with the same titrant and in the same pH and ionic media. At the pH of these experiments, the predominant species of DPA and CA is the monoprotonated form of the ligand. Upon complexation, this proton is displaced. The change in enthalpy of protonation for the fully deprotonated ligand species was experimentally determined and used to account for the change of heat caused by the proton displacement. The error values of ∆Hn were propagated using the statistical error for the experimental heats and the errors associated with the heat of dilution and the heat of protonation. 40

3.6.2 Change of Entropy

The value of the change in the Gibbs free energy was calculated by the equation: ∆Gn = -RT ln (Kn)

Eq. 3.31

where ∆Gn is the change in the Gibbs free energy of the reaction in kJ/mole, Kn is the equilibrium constant, the value of n is the number of ligands coordinating to the metal ion in the complex species as given by Eq. 3.10 and 3.11, R is the gas constant in units of J/mole·K and T is the temperature in Kelvin. The change in entropy can be calculated by the equation: ∆Gn = ∆Hn – T∆Sn

Eq. 3.32

where ∆Sn is the change of the entropy of the reaction in J/mole K. The error of ∆Sn was propagated from the errors associated with the values of ∆Gn and ∆Hn.

41

CHAPTER 4 RESULTS

4.1 Acid Dissociation Constants The concentration acid dissociation constants of PDA, DPA and CA were determined by potentiometric titration at 25.0 ± 0.1°C and 0.10, 0.50 and 2.00 M sodium perchlorate media and are listed in Tables 4.1, 4.2 and 4.3, respectively. The values were calculated using PSEQUAD as explained in section 3.2 by fitting the pH profile of 20 to 40 titrations of each ligand. All ligands were fit with a triprotic chemical model (Eq. 3.3, 3.4 and 3.5). The third deprotonation of DPA can be neglected under the conditions of this work because the proton is completely dissociated in the initial solution because of its highly acidic nature (pKa < 1).

The PSEQUAD fits produced using a triprotic

chemical model have high correlation coefficients for each ligand system, indicating that the model is appropriate for fitting the data. There was a large relative error associated with the pKa3 values of PDA and CA and the pKa2 value of DPA. This error can be attributed to the pKa values being obtained from solutions in which the pH are outside of the linear range of a glass electrode1. The other pKa values have relative errors of less than 2%. The acid dissociation constants of PDA and DPA were also measured by 1H NMR (Tables 4.1 and 4.2). The values are from multiple titrations and are calculated from over 100 data points at 25.0 ± 0.1°C in 0.50 M sodium perchlorate media. The pKa values of 42

the 2, 6-carboxylic acid groups of PDA and DPA were reported as an average value as discussed in section 3.5.A. The values have relative errors less than 9% and are within one standard deviation of the pKa values determined by potentiometry. 4.2 Lanthanide(III) Stability Constants The lanthanide(III) stability constants of DPA and CA, Table 4.4 and 4.5, were determined by potentiometric titration at 25.0 ± 0.1°C. The Ln(III)-DPA system was studied in 0.50 and 2.00 M sodium perchlorate media while the Ln(III)-CA system was studied in 0.10, 0.50 and 2.00 M sodium perchlorate media. The values were calculated using PSEQUAD, as explained in section 3.2, by fitting the pH profile of between 10 and 30 titrations of each lanthanide(III) ligand system. The pH profiles of the lanthanide(III) complexes were fit with a chemical model that assumed all protons were displaced upon coordination using the concentration acid dissociation constants determined by this study. The model allows for the stepwise formation of ML and ML2 species as shown by Eq. 3.10 and 3.11.

The concentration of the ligand was never greater than twice the

lanthanide(III) ion concentration so a ML3 species was not included in the fit. DPA was not expected to form a ML3 complex with lanthanide(III) ions due to the number of coordination sites the ligand can occupy and the steric bulk of the ligand. CA can potentially form a ML3 species with lanthanide(III) ions but the stability constant of this complex was not determined since the results are to be compared with IDA which does not form a ML3 complex2. All fits produced high correlation coefficients that indicate the selected model was appropriate for fitting the data. The lanthanide(III) stability constants of PDA could not be determined by potentiometry as the changes in acidity profile of the ligand and the complex were negligible. 4.3 Calorimetry The change in enthalpy was calculated from experimental heats using Eq. 3.28. The change of Gibbs free energy was calculated from the corresponding equilibrium constant of the reaction using Eq. 3.31. The change of entropy was calculated from ∆H and ∆G values for the reaction using Eq. 3.32.

43

The complexation titrations were performed at 25.000 ± 0.001°C in pH 5.3 to 5.5, 0.100 M MES and 0.50 M sodium perchlorate. The DPA and CA are predominantly the monoprotonated species at this pH.

The change in enthalpy and entropy for the

deprotonation of the ligands were determined by titration at a pH equal to the highest pKa value with a known acid and used to correct for the deprotonation which occurs upon complexation. The titrations of each ligand were performed under the same temperature and ionic media conditions as complexation experiments. The values of ∆H and ∆S were calculated as the average of three titrations and were determined to be –8.36 ± 0.64 kJ mol-1 and 157.4 ± 2.2 J mol-1 K-1 for DPA and –23.7 ± 1.1 kJ mol-1 and 127.3 ± 3.8 J mol1

K-1 for CA. The heat of complexation of PDA could never be distinguished from the

heat of dilution. The ∆G, ∆H and ∆S values of complexation CA and DPA formation are shown in Table 4.6 and Table 4.7. 4.4 Hypersensitivity Spectroscopy Representative spectra of the 4I9/2 → 2G7/2, 4G5/2 transition of Nd(III) and of the 5

I8 → 5F1, 5G6 transition of Ho(III) are shown in Figures A.1 and A.2. Upon complex

formation, the shape of the peak changes and the area of the absorbance peak increases as a result of the hypersensitivity of the metal ion. The use of 10 cm path length cells allows the measurement of mM concentrations of the hypersensitive region of the Nd(III) and Ho(III) spectra with high reproducibility. The area of the peak can be used to calculate the oscillator strength, P, as defined in Eq. 3.17. 4.4.1 PDA Little change is observed in the area and shape of the Nd(III) (Figure A.3) or Ho(III) (Figure A.4) hypersensitive spectra upon increasing the concentration of PDA. As the [L] : [M] ratio changes from 0 to 0.8 at pH ~5.5, the oscillator strength of Nd(III) (Table A.1) changes by less than 5% and the oscillator strength of Ho(III) (Table A.2) increases by less than 9%. Similar change was seen when the PDA was present in two and three fold excess, indicating that PDA formed only a relatively weak complex with Nd(III) and Ho(III) ions. 4.4.2 DPA and CA

44

A significant change occurs in the shape and area of the peak of the Nd(III) spectrum upon the addition of DPA (Figure A.5). The oscillator strength has a large increase (~30%) when the [L] : [M] ratio increases to 1.0, Table A.3. Similar changes are seen in the spectra of Ho(III) as DPA was added, Figure A.6 and Table A.4. The changes in the peak shape and the increase of the oscillator strength are indications of strong complex formation. Significant changes in the shape and area of the peak of the Nd(III) and Ho(III) hypersensitive spectra (Figure A.7 and A.8) are observed as the CA concentration increases. The oscillator strength shows a large increase (~30%) as the [L] : [M] ratio approaches 1.0, listed in Tables A.5 and A.6, and indicates strong complex formation. The degree of complexation of DPA with Nd(III) or Ho(III) shows a dependence on pH when varied between 2.1 and 5.2 as can be seen in the representative Ho(III) spectra shown for conditions favoring ML species (Figure A.9) and ML2 species (Figure A.10. The change of the oscillator strength between the values at pH 3 and pH 5 is larger for the ML2 species (∆P = 1.8) versus the ML species (∆P = 0.8), indicating that formation of the second complex is more sensitive to pH. The oscillator strengths are nearly constant for both the ML and ML2 species between pH 4.5 and 5.5, indicating that the different pH values had only a minimal effect in the experiments used to determine complex oscillator strength values. The complexes of CA show similar trends for Nd(III) and Ho(III), exhibiting a minimal pH dependence over a range from 2.3 and 5.5 as shown in a representative set of Nd(III) spectra, Figure A.11.

The oscillator strength values, Table A.9, show no

significant change over this pH range. The ML2 species of CA complexes with Nd(III) and Ho(III) are more sensitive to pH as shown in the Nd(III) spectra, Figure A.12. The experiments used to determine the complex oscillator strength values are not expected to show a significant pH effect for either species under the conditions of this work. The initial ligand solutions for the hypersensitive spectroscopy experiments were prepared at pH ~5.0. The ligands are predominatly the monoprotonated species at this pH. Upon addition of the metal ion, the pH showed a sharp decrease of several units for DPA and CA but did not measurably change for PDA. This indicated that DPA and CA 45

are completely deprotonated when forming complexes with lanthanide(III) ion under these conditions. The lack of pH change for PDA samples indicates that either only a minor amount of complexation occurs or that any PDA complex that forms does not dissociate the proton of the HPDA-1 species. 4.4.3 Oscillator Strengths The experimental oscillator strengths of the Nd(III) complexes plotted against [L] : [M] for both DPA and CA are shown in Figure 4.1. The change of P indicates the stepwise formation of ML and ML2 complexes of DPA with Nd(III) as seen in the separate linear regions of the plot discussed in section 3.4. The constant P seen after [L] : [M] exceeds 2.50 indicates no higher order complexation is occurring. The change of P of the CA complexes as the [L] : [M] ratio increases indicates stepwise complexation of CA, forming ML, ML2 and ML3 species (Figure 4.1). All of the linear fits of have a high correlation coefficient, Table 4.8, except the [Nd(DPA)2] region. The experimental oscillator strengths of Ho(III) are plotted against [L] : [M], Figure 4.2, for both DPA and CA. The change of P in the plot indicates the stepwise formation of ML and ML2 complexes of DPA with Ho(III) and CA with Ho(III). The oscillator strength becomes constant after the [L] : [M] ratio exceeds 2.00 for both DPA and CA systems, indicating no further complexation is occurring.

Linear fits of each

species region are shown in Table 4.9 and have high correlation coefficients. Ho(III)-CA spectra and oscillator strengths that have been published from the work are shown in Appendix E. Because the oscillator strength values of the PDA complex did not change significantly, PDA was not analyzed. The oscillator strengths of the M, ML and ML2 species, Table 4.10, were calculated as discussed in section 3.4. The values of the oscillator strengths determined for the free metal ions from both DPA and CA experiments show good agreement. The PML3 value for Nd(III) complexes of CA were not determined due to the limited solubility of the ligand and the extinction coefficient of the Nd(III) ion which restricted the [L] : [M] range that could be studied. 4.5 Eu(III) Laser Fluorescence 4.5.1 Excitation Spectra 46

The number of peaks corresponds to the number of distinguishable environments of the Eu(III) ion in the 7F0 → 5D0 selective excitation spectrum3. The intensities of different sets of spectra cannot be directly compared because different amplification settings were used during the collection of different data sets to adjust to the much weaker intensities of the highly hydrated Eu(III)3. The relative intensities of different data sets are not important since this work focused on the determination of the peak positions and the measurement of the fluorescent decay lifetime of the peak.

The

excitation spectrum of europium perchlorate had one very weak peak centered at 17276 cm-1, Figure B.1, which is assumed to be the [Eu(H2O)9]3+ species. The experiments were performed at 23 ± 1°C in 2.00 M sodium perchlorate media. 4.5.2 PDA Representative spectra of the 7F0 → 5D0 selective excitation of Eu(III) in solutions of varying PDA concentration are shown in Figure B.2 with arbitrary units of intensity. A single, very weak peak was seen as the ratio [PDA] : [Eu(III)] was varied between 0.1 and 5.0 (Table B.1) at pH ~5.5. The peak, 17273 cm-1, corresponds to CNL = 1.3 ± 0.5, calculated by Eq. 3.17 as discussed in section 3.3.2. This coordination number value indicates that the probable species is [Eu(HxPDA)(H2O)8]1+x with PDA coordinating as a monodentate ligand and x equal to 0 or 1, Table 4.11. The weak intensity of the peak at 17273 cm-1 prohibited lifetime measurements with PDA. 4.5.3 DPA Representative spectra of the 7F0 → 5D0 selective excitation of Eu(III) during a titration with DPA are shown in Figure B.3 with arbitrary units of intensity. The peaks and concentrations are listed in Table B.2. When [DPA] < [Eu(III)], a single peak is observed in the spectra at 17260 cm-1. This peak has a shoulder at 17252 cm-1 that could not be isolated and maintained a constant intensity of approximately 10% relative to the intensity of the 17260 cm-1 peak. The peaks of Eu(III) excitation spectra are sharp and can be fit as Lorentzian lineshapes so the shoulder must be considered as an additional Eu(III) environment4. The peak and the shoulder have CNL = 4.4 ± 0.5 and 6.6 ± 0.5, Table 4.11. Because the shoulder could not be resolved from the peak using different pH and concentration ratios and was always present in a fixed ratio relative to the peak, the 47

species that the shoulder and peak represent must be related. Therefore, the number of water molecules in the inner hydration sphere of 3.9 ± 0.5, nH2O calculated by Eq. 3.16, for the peak and shoulder is reported as an average of both species.

The peak

corresponds to the ML complex where DPA is coordinating in a tetradentate manner, corresponding to a [Eu(DPA)(H2O)4] species (Table 4.11). The shoulder is probably a different geometrical form(s) of the tetradentate coordinating DPA species, probably involving

some

geometric

isomerization

of

the

ligand

corresponding

to

a

[Eu(DPA)(H2O)3] species (Table 4.11). DPA can undergo ring isomerization similar to cyclohexane and has several variations such as boat and chair forms.

Molecular

modeling performed by Dr. Kenneth L. Nash at Argonne National Laboratory with the Alchemy program indicates that both the boat and chair forms should displace 4 to 5 water molecules upon coordination. The seemingly higher coordination number of the shoulder represents the intermediate as the molecule converted between boat and chair forms since the modeling indicates that the intermediate form could displace 1 to 2 additional water molecules. Another peak is observed at 17239 cm-1 when the DPA concentration exceeds that of the Eu(III) ion. This peak has CNL = 9.2 ± 0.5 and nH2O = 1.5 ± 0.5 and corresponds to the ML2 species where two DPA molecules coordinate the Eu(III) ion in a tetradentate fashion, [Eu(DPA)2(H2O)]3- (Table 4.11). 4.5.4 CA Representative spectra of the 7F0 → 5D0 selective excitation of Eu(III) during a titration by CA are shown in Figure B.4 with arbitrary units of intensity with the total concentrations of each reagent are listed in Table B.3. An initial peak forms at 17266 cm

-1

in the spectra when [CA] < [Eu(III)]. The values of CNL = 3.2 ± 0.5 and nH2O =

4.5 ± 0.5 indicate that the predominant species is [Eu(CA)(H2O)4,5]+ (Table 4.11). A second peak appears at 17252 cm-1 when the CA concentration exceeds the concentration of Eu(III) ion. The values of CNL = 6.6 ± 0.5 and nH2O = 2.9 ± 0.5 of this peak are consistent with tridentate complexation by two CA molecules to the metal ion, corresponding to a [Eu(CA)2(H2O)3]- species (Table 4.11). 48

A third peak was observed at 17233 cm-1 when the CA concentration was several times in excess of the metal ion. The peak was not observed at an intensity that would allow the measurement of the fluorescence decay lifetime. The peak has a CNL = 10.8 ± 0.5, corresponding to the [Eu(CA)3]3- species (Table 4.11). Eu(III)-CA flourescence spectra, waters of hydration and calculated coordination numbers that have been published from the work are shown in Appendix E. 4.6 NMR Typical 1H NMR spectra of PDA and DPA are shown in Figures C.1 and C.2, respectively. 4.6.1 Acid Dissociation Constant Determination Plots of the change in chemical shift of nonexchangable protons versus pD are shown in Figures 4.3 and 4.4 for PDA and DPA. The acid dissociation constants were determined from the change of the chemical shift of the peak as pD was varied using Eq. 3.20 as discussed in section 3.5.A. The calculated acid dissociation constants are listed in Tables 4.1 and 4.2 and have been discussed in section 4.1. 4.6.2 1H NMR Peak Identification of PDA The peaks in the spectrum shown in Figure C.1 can be attributed to the different proton environments of PDA. The nOe experimental results are shown in Table 4.12 with the different proton position assignments shown in Figure 4.5. The proton bonded to the carbon in the alpha position from the amine appears as a doublet of doublets with a chemical shift between 4.1 and 3.1, depending on pD, and shown as Ha in Figure 4.23. This peak appears at chemical shifts typical of protons in proximity to amine and carboxyl functional groups and with the expected multiplicity. The integration value of the peak located at 1.7 to 1.2, depending on pD, corresponds to 2 protons. The homonuclear decoupling of this peak affects the splitting pattern of all other peaks. The decoupling of the proton alpha to the amine results in significant change in the multiplicity of the 1.7 ppm peak and only a small change in the 2.4 ppm peak. This indicates that this peak could be assigned to the protons bonded to the carbon in the beta position from the amine, shown as Hb in Figure 4.5. The nOe results in Table 4.12 supported this assignment of the proton. 49

A symmetric doublet of doublets appears in the spectra at 2.4 and 2.1 ppm in acidic systems. These peaks merge as the acidity decreases, forming a single peak at 1.9 ppm as pD approaches 10.0. Both doublets show strong coupling with the peak at 1.7 ppm while the peak at 2.4 ppm shows coupling to Ha proton. This indicates that this peak system represents the protons bonded to the carbon in the para position from the amine. The proton environment is split by the geometry of the molecule with one of the protons having a significantly higher interaction with the carboxylic acids when it is in the axial position, shown as Hc in Figure 4.5, and corresponds to the peak at 2.4 ppm in acidic systems. The interaction of the Hd proton is weaker with the carboxylic acid groups and so a lower chemical shift, 2.1 ppm, is observed. The peaks merge in less acidic systems because the carboxy groups orient equatorially because of the charge repsulion, limiting the carboxy-Hc interaction. 4.6.3 1H NMR Peak Identification of DPA The peaks in the spectrum shown in Figure C.2 can be attributed to the different proton environments of DPA shown in Figure 4.6. The nOe experimental results are shown in Table 4.13. The proton bonded to the carbon alpha to the amine, Ha in Figure 4.6, appears as the peak between 4.4 to 3.5 ppm, depending on pD. The peak from 4.2 to 3.2 ppm, depending on pD, is due to the proton of the N-acetate group shown as Hf in Figure 4.6. Both peaks appear in predictable regions for protons in proximity to amine and carboxyl functional groups and have the expected splitting patterns and integration values. The integration value of the peak between 2.0 to 1.4 ppm, depending on pH, corresponds to 2 protons. The homonuclear decoupling of this peak affects all proton peaks except the Hf proton peak but only shows a change in multiplicity when the proton alpha to the amine is decoupled. This indicates that the peak corresponds to the protons bonded to the carbon in the beta position relative to the amine, shown as Hb in Figure 4.6. The nOe indicates that this is the correct assignment of the proton positions on the structure but the nOe results are not as simple as might be expected. The complication of the nOe spectra is due to multiple geometries of the ring structure as well as the addition of different splitting caused by protons on the varying sides of the ring. 50

The multiplet peaks between 2.3 to 1.85 ppm and 1.8 to 1.7 ppm, depending on pD, each integrate to the equivalent of 1 proton. The homonuclear decoupling of these peaks induces changes in the Hb proton peak. The peak at 2.3 ppm also affects the peak from the protons at Ha in acidic systems, indicating that these protons are bonded to the carbon in the para position relative to the amine. The proton Hc corresponds to the peak at 2.3 ppm, having an increased interaction with the carboxylic acid groups. The peak at 1.8 ppm is least affected by changes in pH and can be assigned to Hd, seen in Figure 4.6. The nOe results in acidic systems in Table 4.13 are consistent with these proton assignments. 4.6.4 NMR of Lanthanide(III) Complexes A series of 1H NMR spectra of PDA were taken for solutions of Eu(III) at 25.0 ± 0.1°C and 0.5 M ionic strength. The pHs, 3.4, 1.9 and 0.6, were chosen as conditions that should show a range of complex formation from some Eu(III)-PDA complexation at pH 3.4 to very little interaction at pH 0.6. The spectra should show a large shift and broadening of the peaks in the NMR spectrum upon the complexation due to the paramagnetic Eu(III) ion. Only a small change is observed in the chemical shifts of the peaks that corresponds to acidity effects upon the ligand. Only a small broadening is observed, indicating that either only a minimal interaction occurs between the Eu(III) ion and PDA or that the exchange rate of the complex is faster than the NMR time scale with the ligand predominantly residing in the uncomplexed form. Spectra of the La(III) and Lu(III) complexes of DPA were studied using 1H NMR. The peaks in the spectrum of the La(III) complexes are very broad (Figure C.3), indicating that the ligand is involved in a chemical equilibrium with kinetics on the NMR time scale. The broadness of the peaks makes comparisons between the [La(III)DPA] spectrum and the DPA spectrum difficult. The peaks in the spectra of the Lu(III)-DPA complexes remain sharp and maintain their multiplicity. Both the chemical shift and the multiplicity of the peaks change upon the formation of the [Lu(III)DPA] complexes (Figure C.4). Many additional peaks are observed in the spectrum of a solution with a molar ratio of DPA twice that of Lu(III) (Figure C.5). Some of the peaks are similar to the ones from the [Lu(III)DPA] and DPA solutions but a series of additional peaks is also 51

observed. The peaks that are similar to DPA indicate that the uncomplexed ligand is present. The additional peaks indicate that the DPA ligands do not complex Lu(III) symmetrically. One of the DPA molecules in the ML2 complex is in a geometrically similar form to the DPA molecule in the ML species while the other DPA molecule exists in some other chemical or geometrical form. The La(III) and Lu(III) complexes of CA were studied using 1H decoupled

13

C

NMR. The spectra of CA consist of peaks at 118.3, 153.9, 175.9 and 176.5 ppm. These peaks, based on typical functional group chemical shifts of 13C NMR, can be assigned to the C3, C2, C1 and C1A carbon as shown on the structure in Figure 4.7. The carboxyl carbon atoms are highly insensitive in 13C NMR experiments and require a significantly higher number of data acquisitions5. The position of the carboxylic acid chemical shift does not change significantly when bonded to an ethylene or aromatic carbon and so will not give additional information on the resonance form of the CA. Therefore, the spectra were not collected over adequate times to determine the chemical shift of the carboxylic acids. The spectrum of La(III)-CA has peaks at 117.4, 155.4, 171.2 and 176.2 ppm, having a significant change in the C1 or C1A carbon atoms. The spectrum of Lu(III)-CA has peaks at 120.0, 147.4, 169.1 and 185.6 ppm. Most of the peaks did not show significant change upon formation of the complex. Because the aromatic and alkene region significantly overlap in

13

C NMR, it is not possible to determine if one of the

resonance forms of CA was preferred upon coordination to the lanthanide(III) ions. The two carbon atoms ortho to the amine show a significantly change in their chemical shift. The chemical shifts of the other peaks show an increased separation of the positions in the complexed forms when compared to the free ligand. This can be explained by nonsymmertical coordination where the metal ion favors coordination with one carboxylate group instead of equally coordinating to both groups. This effect would increase in Lu(III) complexes when compared to La(III) because of the smaller ion size, as is observed. Because of the low solubility of CA and the relative insensitivity of 13C NMR, exceedingly long periods were required for data collection required and so further experiments were not performed. 4.7 Crystallography 52

The crystal structures of PDA and DPA were determined with a high correlation of fit. The structure of PDA is shown in Figure D.1 with structure parameters shown in Tables D.1, D.2, D.3, D.4 and D.5. The structure of DPA is shown in Figure D.2 with structure parameters shown in Tables D.6, D.7, D.8, D.9 and D.10.

53

Table 4.1. The concentration acid dissociation constants for PDA, T = 25.0°C, NaClO4 media. Ionic Strength (M)

Method

log Ka1

log Ka2

log Ka3

0.10

potentiometry

10.39 ± 0.05

3.46 ± 0.04

1.7 ± 0.3

potentiometry

10.51 ± 0.03

4.17 ± 0.02

1.9 ± 0.3

NMR

10.52 ± 0.06

potentiometry

10.84 ± 0.07

0.50 2.00

54

2.74 ± 0.24 4.09 ± 0.05

1.8 ± 0.3

Table 4.2. The concentration acid dissociation constants for DPA, T = 25.0°C, NaClO4 media. Ionic Strength (M)

Method

log Ka1

log Ka2

log Ka3

0.10

potentiometry

9.42 ± 0.05

2.98 ± 0.05

1.4 ± 0.3

potentiometry

9.69 ± 0.03

3.02 ± 0.04

1.1 ± 0.3

NMR

9.78 ± 0.18

potentiometry

9.55 ± 0.01

0.50 2.00

55

2.07 ± 0.08 3.14 ± 0.02

1.1 ± 0.3

Table 4.3. The concentration acid dissociation constants* for CA, T = 25.0°C, NaClO4 media. Ionic Strength (M)

log Ka1

log Ka2

log Ka3

0.10

10.75 ± 0.04

3.02 ± 0.06

1.3 ± 0.3

0.50

10.81 ± 0.05

3.47 ± 0.06

1.0 ± 0.3

2.00

10.45 ± 0.02

3.27 ± 0.03

1.0 ± 0.3

* Determined by potentiometry.

56

Table 4.4. Stability constants of select lanthanide(III) ions with DPA, T = 25.0° C NaClO4 media. *

La Nd Eu Ho Lu

Ionic Strength (M)

log β1

log β2

0.50

9.73 ± 0.20

17.98 ± 0.22

2.00

9.38 ± 0.25

16.72 ± 0.26

0.50

10.58 ± 0.60

20.59 ± 0.40

2.00

10.18 ± 0.56

19.90 ± 0.50

0.50

10.87 ± 0.19

19.56 ± 0.19

2.00

10.57 ± 0.16

19.09 ± 0.16

0.50

9.54 ± 0.04

15.15 ± 0.05

2.00

10.70 ± 0.13

16.77 ± 0.16

0.50

10.18 ± 0.08

16.15 ± 0.10

2.00

12.09 ± 0.36

21.62 ± 0.40

* Determined by potentiometry.

57

Table 4.5. Stability constants of select lanthanide(III) ions with CA, T = 25.0° C, NaClO4 media. *

La

Nd

Eu

Ho

Lu

Ionic Strength (M)

log β1

log β2

0.10

7.37 ± 0.19

14.74 ± 0.22

0.50

6.92 ± 0.12

14.24 ± 0.13

2.00

7.36 ± 0.13

13.63 ± 0.83

0.10

7.50 ± 0.17

14.66 ± 0.43

0.50

6.83 ± 0.34

14.75 ± 0.14

2.00

7.71 ± 0.21

15.38 ± 0.20

0.10

7.67 ± 0.18

14.75 ± 0.54

0.50

6.47 ± 0.64

14.81 ± 0.11

2.00

7.49 ± 0.13

14.53 ± 0.25

0.10

7.82 ± 0.07

15.23 ± 0.12

0.50

7.19 ± 0.27

15.28 ± 0.15

2.00

7.20 ± 0.21

14.88 ± 0.21

0.10

9.20 ± 0.18

16.28 ± 0.71

0.50

7.18 ± 0.15

14.97 ± 0.07

2.00

7.82 ± 0.11

15.23 ± 0.14

* Determined by potentiometry.

58

Table 4.6. Thermodynamic values for the DPA complexation with Ho(III) or Nd(III), pH = 5.30 to 5.50, [MES] = 0.100 M, T = 25.000°C, I = 0.50 M NaClO4. Ln(III) Ho(III) Nd(III)

n

∆Gn (kJ mole-1)*

∆Hn (kJ mole-1)

∆Sn (J mole-1 K-1)

1

-54.46 ± 0.10

11.5 ± 2.9

221 ± 10

2

-32.02 ± 0.12

0.6 ± 3.1

109 ± 10

1

-60.4 ± 1.5

11.8 ± 3.0

242 ± 11

2

-57.14 ± 0.99

1.0 ± 3.3

203 ± 11

* Calculated from the equilibrium constant.

59

Table 4.7. Thermodynamic values for the CA complexation with Ho(III) or Nd(III), pH = 5.30 to 5.50, [MES] = 0.100 M, T = 25.000°C, I = 0.50 M NaClO4. Ln(III) Ho(III) Nd(III)

n

∆Gn (kJ mole-1)*

∆Hn (kJ mole-1)

∆Sn (J mole-1 K-1)

1

-41.04 ± 0.67

20.1 ± 3.8

205 ± 13

2

-46.18 ± 0.37

16.6 ± 3.9

211 ± 13

1

-38.99 ± 0.84

14.7 ± 3.8

180 ± 13

2

-45.21 ± 0.35

15.9 ± 3.9

165 ± 13

* Calculated from the equilibrium constant.

60

26

[Nd(CA)3] 24 22

P(x106)

20

[Nd(CA)2]

18 16 14

[NdCA]

12

[Nd(DPA)2]

10

[NdDPA] 0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

[L] : [M]

Figure 4.1. The oscillator strength of Nd(III) complexes plotted versus the ligand to metal concentration ratio, I = 2.00 M NaClO4, pH = 4.50 - 6.00, T = 23°C.

61

Table 4.8. Values from the linear regressions of Figure 4.1 for the change of the oscillator strength of Nd(III) when complexed with DPA and CA, I = 2.00 M NaClO4, pH = 4.50 - 6.00, T = 23°C. complex

slope

intercept

correlation coefficient

NdCA

4.76 ± 0.12

9.53 ± 0.07

0.984

Nd(CA)2

3.59 ± 0.15

10.86 ± 0.29

0.976

Nd(CA)3

0.83 ± 0.12

20.61 ± 0.53

0.980

NdDPA

3.11 ± 0.12

9.47 ± 0.66

0.951

Nd(DPA)2

0.56 ± 0.13

12.36 ± 0.20

0.588

62

24

[Ho(CA)2]

22 20

P(x106)

18 16 14 12

[HoCA]

[Ho(DPA)2]

10 8

[HoDPA]

6 0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

[L] : [M]

Figure 4.2. The oscillator strength of Ho(III) complexes plotted versus the ligand to metal concentration ratio, I = 2.00 M NaClO4, pH = 4.50 - 6.00, T = 23°C.

63

Table 4.9. Values from the linear regressions of Figure 4.2 for the change of the oscillator strength of Ho(III) when complexed with DPA and CA, I = 2.00 M NaClO4, pH = 4.50 - 6.00, T = 23°C. complex

slope

intercept

correlation coefficient

HoCA

7.79 ± 0.10

5.92 ± 0.05

0.996

Ho(CA)2

8.24 ± 0.25

5.56 ± 0.36

0.994

HoDPA

4.64 ± 0.22

6.20 ± 0.12

0.954

Ho(DPA)2

4.20 ± 0.41

6.65 ± 0.60

0.904

64

Table 4.10. The oscillator strengths of Nd(III) and Ho(III) and their complexes with DPA and CA*, I = 2.00 M NaClO4, T= 23°C. system

PM / 106

PML / 106

PML2 / 106

Nd(III) DPA

9.47 ± 0.66

12.98 ± 0.12

14.51 ± 0.04

Ho(III) DPA

6.20 ± 0.12

10.94 ± 0.39

15.26 ± 0.12

Nd(III) CA

9.53 ± 0.07

14.46 ± 0.18

22.89 ± 0.42

Ho(III) CA

5.95 ± 0.05

12.15 ± 0.21

23.26 ± 0.05

* The subscript of P indicates the metal ion species.

65

Table 4.11. Summary of the peaks observed by fluorescence peaks of Eu(III) for the various PDA, DPA and CA, I = 0.50 and 2.00 M NaClO4, T = 23°C. system

peak (cm-1)

CNL

nH2O

probable species

Eu

17276

0

9.0 ± 0.5

[Eu(H20)9]3+

PDA

17273

1.3 ± 0.5

17260

4.4 ± 0.5

17252

6.6 ± 0.5

17239

9.2 ± 0.5

1.6 ± 0.5

[Eu(DPA)2(H20)]3-

17266

3.2 ± 0.5

4.5 ± 0.5

[Eu(CA)(H20)4]+

17252

6.6 ± 0.5

2.9 ± 0.5

[Eu(CA)2(H20)3]-

17233

10.8 ± 0.5

DPA

CA

66

[Eu(PDA)(H20)8]+ 3.9 ± 0.5

[Eu(DPA)(H20)4] [Eu(DPA)(H20)3]

[Eu(CA)3]3-

chemical shift (ppm)

4.00 3.75 3.50 3.25 3.00 2.75 2.50 2.25 2.00 1.75 1.50 1.25

Ha Hc Hd Hb

0

1

2

3

4

5

6

7 pD

8

9

10 11 12 13

Figure 4.3. The chemical shift of the peaks in the 1H NMR spectrum of PDA versus –log of concentration of the deuterium ion, I = 0.50 M NaClO4 and T = 25.0°C. The peak legend refers to proton assignments given in Figure 4.5.

67

4.5

chemical shift (ppm)

4.0

Ha Hf Hc Hb Hd

3.5

3.0

2.5

2.0

1.5

0

1

2

3

4

5

6

7 8 pD

9

10 11 12 13 14

Figure 4.4. The chemical shift of various peaks in the 1H NMR spectrum of DPA versus –log of concentration of the deuterium ion, I = 0.50 M NaClO4 and T = 25.0°C. The peak legend refers to proton assignments given in Figure 4.6.

68

Table 4.12. The nuclear Overhauser effect of PDA* at different pH, I = 0.50 M NaClO4 and T = 25.0°C. pH 0.30 peak (ppm)

4.20

2.40

2.10

1.70

IP

3.35

0

2.73

7.79

IP

0.32

34.0

0

0

IP

41.61

5.57

20.06

12.14

IP

pH 10.05 peak (ppm)

3.00

1.85

1.45

1.15

IP

3.93

2.99

spt

4.54

IP

13.50

20.05

13.44

42.60

IP

spt

0.85

36.61

spt

IP

* Values are reported as the ratio of the absolute value of the irradiated peak area divided by the enhanced peak area. The symbol “IP” indicates the peak was the irradiation peak of the nOe experiment. The symbol “spt” indicates selective population transfer occurred at the peak during the nOe experiment.

69

H

H

H H O

D

OD

OD

O

D N

O

N OD D D

Ha

H H

OD

Hc O

OD

Ha

Hb

Hd

O

Hb

Ha

Hd

D

Hb

Figure 4.5. PDA proton atom designations used with 1H NMR.

70

Hc

Hb

D N

Hb

Hb

Ha OD O

Hb Hb

Table 4.13. The observed nuclear Overhauser effect of DPA* at different pH, I = 0.50 NaClO4 and T = 25.0°C. pH 0.86 peak (ppm)

4.3

4.2

2.3

2.0

1.7

IP

54.2

59.2

0

51.9

58.9

IP

0

55.2

0

67.0

0

IP

80.6

56.1

50.4

58.9

79.6

IP

61.0

76.1

0

65.9

74.5

IP

4.0

3.8

2.2

2.0

1.7

IP

56.3

59.2

45.3

45.0

65.0

IP

0

0

0

68.8

19.0

IP

77.4

60.2

67.8

44.3

IP

75.2

59.3

58.8

0

76.2

IP

spt

62.7

spt

IP

pH 3.20 peak (ppm)

56.7 pH 14.10 peak (ppm)

3.5

3.2

1.85

1.75

1.45/1.35

IP

53.7

59.6

0

54.4

63.7

IP

0

0

0

69.3

0

IP

0

83.8(broad)

0

0

0

IP

82.3(broad)

0

0

81.6

63.8

IP(1.45)

0

54.9

81.9

65.2

IP(1.45)

58.0

0

78.6

67.9

IP(1.35)

* Values are reported as the ratio of the absolute value of the irradiated peak area divided by the enhanced peak area. The symbol “IP” indicates the peak was the irradiation peak of the nOe experiment. The symbol “spt” indicates selective population transfer occurred at the peak during the nOe experiment.

71

H H H H O

N D

H H H O OD

OD

OD OD

O Ha

D N

O

DO

Hc OD

O

Ha Hf

Hf

OD

Hb

Hd

O

Hb

Ha

Hd

Hf O OD

Hf

Hb

O

O

Figure 4.6. DPA proton atom designations used with 1H NMR.

72

Hc

Hb

D N

Hb

Hb

Ha OD

Hb Hb

OD H C2 O

C1 OD

C3

N

H C2 C1 A

D

Figure 4.7. CA carbon atom designation used with 13C NMR.

73

O OD

CHAPTER 5 DISCUSSION

5.1 Acid Dissociation Constants The pKa2 value determined for PDA, 3.46 ± 0.04, disagrees with the values reported in the literature (Table 5.1). The value reported in the literature was fit with a diprotic acid model. The potentiometric data from this study was analyzed using both a diprotic model and a triprotic model of PDA to examine the effect on the second acid dissociation constant value. The first and second pKa values of the triprotic model agree with values determined using the diprotic model within 2 standard deviations. The H3PDA+1 molecule is a minor species under the conditions of the potentiometric study and so the inclusion of this proton results in only a minor change in the data fitting. The literature pKa2 value is similar to the value determined by 1H NMR titrations, 2.74 ± 0.24, which is the average of the second and third acid dissociation constant values as discussed in section 4.1. PDA is an analog of iminodiacetic acid (IDA) which has three reported pKa values (Table 5.2). The similarity between the acid dissociation constants of nitrilotriacetic acid (NTA) and its analog DPA, discussed in the following paragraph, would suggest that IDA and PDA should also have similar acidic behavior and would support the use of a triprotic model for PDA. The difference of pKa2 values of PDA and IDA is caused by the additional electron donating aliphatic groups of the piperidine ring100. 74

The pKa1 and pKa3 values, 9.42 ± 0.05 and 1.4 ± 0.3, respectively, determined for DPA are within 2 standard deviations of the values reported in the literature (Table 5.1). The pKa2 value of DPA determined in this study was slightly higher than the value reported in the literature, 2.98 ± 0.05 versus 2.71. The difference may be due to the ionic media of the two studies, potassium nitrate versus sodium perchlorate ionic of this work. The acid dissociation constant values determined for CA, 10.75 ± 0.04, 3.02 ± 0.06 and 1.3 ± 0.3 in 0.1 M sodium perchlorate, are within 2 standard deviations of the previously determined values (Table 5.1). A direct comparison is difficult due to the different ionic media and temperatures in the different studies. The pKa values of dipicolinic acid (Table 5.2) are significantly different from those of CA despite the structural similarity. The pKa1 value of CA more closely resembles the value listed for 4hydroxybenzene (Table 5.1), indicating that the acidity of the CA amine is strongly affected by the phenolic resonance form. 5.2 Lanthanide(III) Stability Constants 5.2.1 PDA The stability constants of PDA and lanthanide(III) ions could not be determined from potentiometry as discussed previously in section 4.2. The values reported in the literature (Tables 5.3 and 5.4) indicate that a moderately strong complex forms with lanthanide(III) ions; these values are similar to the values reported for IDA, listed in Table 5.5. The discrepancies of the values from this work and those in the literature are discussed in section 5.8.1. 5.2.2 DPA The previously determined lanthanide(III) stability constants of DPA are listed in Tables 5.3 and 5.4. The stability constants values of DPA have a similar trend across the lanthanide(III) series but a direct comparison of the values can not be made due to the different ionic strength and media. The first stability constants determined in this study are within 2 standard deviations of the reported values except for the values determined with La(III) at 0.5 M ionic strength. The second stability constants determined by this work tend to be slightly higher than the values from the literature. The nitrate ionic media of the literature study could compete with the ligand, lowering the value compared 75

to the value determined in a non-complexing ionic media. The stability constants of DPA are lower than the values reported for the coordination analog NTA (Table 5.5). The stability constants of DPA and NTA values tend to have a smaller difference as the ionic radius of the lanthanide(III) ion decreases. The lower stability constant values of DPA are a result of the piperidine ring structure which can undergo an inversion of the coordination site due to interchange of boat confirmations similar to cyclohexane. This geometric transition of the piperidine is referred to as ring flip as discussed in section 1.4. The ring flip would invert the amine and cause a lengthening of the Ln-N bond, reducing the stability of the complex. 5.2.3 CA The previously reported stability constant of CA with lanthanum(III) (Table 5.6) is significantly lower than the value reported in this study, (Tables 5.3 and 5.4). The very low ionic strength and the lack of experimental details of the previous study make further comparison difficult. For the reported set of values, the first and second stability constant values of CA (log β1 ranging from 6.92 ± 0.12 to 7.18 ± 0.15 and log β2 14.24 ± 0.13 to 14.97 ± 0.07 across the lanthanide(III) series) are an order of magnitude lower than the values for dipicolinic acid, listed in Table 5.5. The non-aromatic piperidine ring structure of CA causes the amine to change the coordination site geometry as the nitrogen atom orbitals alternate from planar “sp2” character to nonplanar “sp3” character, discussed in section 1.4. The non-aromatic amine would not be expected to be planar and would alternate being above and below the plane of the ring.

The non-aromatic amine

rearrangement would also affect the position of the carboxylic acid groups, decreasing their planarity. The alternating nitrogen position should introduce a longer solution Ln-N coordination distance, while the alternating position of the coordination groups should displace additional water molecules around the inner hydration sphere of the lanthanide(III) ion upon coordination and slightly decrease the affinity of the ligand for the coordinated metal compared to dipicolinic acid66,101. This agrees with results from the laser fluorescence of Eu(III) that indicate that the first complex of CA displaces more water molecules from the Eu(III) inner hydration sphere than the dipicolinic acid complex. CA also shows a decreased lanthanide(III) ion size dependency relative to the 76

variation of dipicolinic acid stability constants across the lanthanide series. The smaller size dependency may be explained by the longer Ln-N coordination bond, the alternating coordination group geometry and the slightly increased bite size of the CA coordination site due to the non-aromatic resonance form. The decreased stability constants can also be explained by a reorientation of the proton and coordination sites of the imino atom of the non-aromatic CA molecule, called nitrogen inversion. The nitrogen inversion inverts the tetrahedral orbital configuration of the imino atom and breaks the coordination bond until inversion occurs again or the ligand reorients to allow the lanthanide(III) ion access to the lone pair of electrons on the nitrogen atom. The nitrogen inversion would weaken the Ln-N coordination bond and decrease the stability of the complex compared to dipicolinic acid. 5.3 Calorimetry 5.3.1 Ligand Protonation 5.3.1.1 DPA The values for the change in entropy and enthalpy for the first protonation of the free ligand for DPA, IDA and NTA at 25.000 ± 0.001°C and 0.5 M sodium perchlorate media are listed in Table 5.7. Comparison of IDA and NTA indicates that the change of enthalpy of protonation decreases with increasing size of the molecule.

1

H NMR

experiments indicate that the DPA molecule adapts different confirmations as the pH is varied as seen in the behavior of the peak corresponding to the protons in the para position, discussed in section 4.6.3. The proton is localized on the amine but can also be associated with one or more of the carboxy groups, either directly or through one or more water molecules. The added proton would change the hydration sphere around the amine site. This increase in the organization of the DPA molecule would result in a net increase of entropy as water molecules are released. The decrease of the ∆H observed in the series IDA, NTA and DPA may be related to a decreased polarity of the amine site caused by the presence of additional electron donating alkyl groups. 5.3.1.2 CA The ∆H and ∆S values for the first protonation, listed in Table 5.7, of CA cannot be directly compared to dipicolinic acid due to the different functionality associated with 77

the first protonation of the molecules. The aromatic amine of dipicolinic acid is much more acidic than the phenol or the aliphatic amine of the CA tautomers as can be observed in the pKa1 values of dipicolinic acid and CA, 4.51 and 10.75 ± 0.04, respectively. The first acid dissociation constant of CA is more similar to the aliphatic amine of NTA or the phenol of hydroxybenzene than to that of dipicolinic acid (Table 5.1). The ∆H and ∆S of protonation values, -23.7 kJ mole-1 and 127.3 J mole-1 K-1, are also more similar to the values reported for NTA or hydroxybenzene (Table 5.5). Both the potentiometry and calorimetry studies indicate that the first acid protonation constant corresponds to either the phenol or aliphatic structure of the CA tautomers. 5.3.2 Change of Enthalpy 5.3.2.1 PDA The heat of complexation of PDA is below the limit of detection of the calorimeter used in these experiments. The low heat of complexation is due to the weak lanthanide(III)-PDA complex, observed in the inability to measure the stability constants by potentiometry. 5.3.2.2 DPA The first DPA complexes of Ho(III) or Nd(III) have an endothermic change of enthalpy and a large, positive change in entropy (11.5 ± 2.9 kJ mole-1 and 221 ± 10 J mole-1 K-1 for Ho(III) and 11.8 ± 3.0 kJ mole-1 and 242 ± 11 J mole-1 K-1 for Nd(III)), indicating that the complex formation is an entropy driven process. The second DPA complex with Ho(III) or Nd(III) forms with a small, endothermic ∆H and a large, positive ∆S (0.6 ± 3.1 kJ mole-1 and 109 ± 10 J mole-1 K-1 for Ho(III) and 1.0 ± 3.3 kJ mole-1 and 203 ± 11 J mole-1 K-1 for Nd(III)). The complexes of NTA form with an exothermic ∆H except in the case of the Ho-NTA that has a slightly endothermic enthalpy (Table 5.6). Previous work has shown a trend relating the residual enthalpy, δ∆H, of aminopolycarboxylate complexation of Nd(III) ions and the sum of the pKa values of the amine functional groups of the ligand, Figure 5.114. The δ∆H is a value corresponding to the amine-lanthanide(III) interaction and is calculated from ∆H. The values of δ∆H is the calculated by: δ∆H = ∆H – n 6.5 kJ mole-1 78

Eq. 5.1

complexation where n is the number of coordinating carboxylate groups and each carboxylate coordination is assumed to have a similar enthalpy change as an acetate coordination.

The δ∆H value increases as the amine or amines of an

aminopolycarboxylate have decreased acid character. The values of dipicolinic acid, IDA and N,N’-ethylenediaminediacetic acid (EDDA) show significantly deviations from the δ∆H-ΣpKa(N) trend. The dipicolinic acid deviation has been explained by a shorter bond distance between the lanthanide(III) ion and the nitrogen atom due to the more polar nature of the aromatic amine4.

The IDA and EDDA deviations are due to longer

lanthanide(III) amine bond distances4. The DPA molecule does not have an imino site but can undergo a ring flip as discussed in section 5.2.2. The ring flip would cause the coordination site to fluctuate as the carboxy groups convert between axial and equatorial positions. Then ring flip could also induce a nitrogen inversion as the N-acetate group reorients to coordinate to the lanthanide(III) ion. If the carboxylate coordinations of DPA behave similarly to acetate coordination, then the lower δ∆H values of the Nd-DPA complex suggests that the Nd-N bond distance of the DPA complex is longer than the typical bond distance of the selected aminopolycarboxylate ligands.

The apparent

decreased Nd-N interaction of DPA can be related to the ring structure that restricts the geometry of the coordination site. The aminopolycarboxylate ligands included in Figure 5.1 are straight chain aliphatic molecules except α-picolinic acid, dipicolinic acid and trans-1,2-diaminocyclohexanetetraacetic acid (DCTA). DCTA does not have a cyclic structure involved in the molecular geometry of the acetate groups. Dipicolinc acid and α-picolinic acid also have ring structures directly attached to the carboxy groups but cannot be directly compared to DPA due to the aromatic nature of both the ring structure and the amine of these ligands. The carboxylic acid groups of DPA may have longer coordination bonds and weaker interactions with the lanthanide(III) ions than acetate due to restricted molecule geometry of the piperidine ring. If the carboxy coordinations are weaker, the method used to calculate δ∆H would not represent the Ln-N interaction and meaningful comparisons to other aminopolycarboxylate ligands cannot be made. 5.3.2.3 CA

79

The ∆H values determined in this study for Ho(III) and Nd(III) complexation with CA, 20.1 ± 3.8 J mole-1 K-1 for Ho(III) and 14.7 ± 3.8 J mole-1 K-1 for Nd(III), do not agree with the literature values listed in Table 5.6. The values are difficult to compare due to the different ionic strengths, 0.001 M versus 0.50 M. The change of enthalpy of CA also does not agree with the values reported for dipicolinic acid even though a tautomer of CA and dipicolinic acid have an identical coordination site. The dipicolinate complex formation reactions are exothermic while the CA complex formation reactions are endothermic. The ∆H of CA would be expected to be lower than dipicolinic acid due to the increased aliphatic nature of the amine which typically are less exothermic complexes formation reactions with lanthanide(III) ions. The increased ∆H of aromatic amine coordination has been explained by a decreased distance between the lanthanide(III) ion and the amine4 but there is not a significant difference in the Gd-N distance reported in the crystal structures of CA102 and dipicolinic acid103. The average Gd-O bond of dipiconlinic acid is slightly shorter than the Gd-O bond distance of CA, 2.395 Å6 versus 2.447 Å5, and is caused by increased interligand repulsion of the [Gd(CA)3]3- species versus the [Gd(Hdipic)(dipic)]0 species.

The solid state bond

distances may not represent the aqueous lanthanide(III) coordination distances as the complexes would be expected to fluctuate more in an aqueous environment. Assuming the carboxylic acid coordinations of dipicolinic acid and CA are similar, the low δ∆H of CA, Figure 5.1, must indicate an increased Ln-N bond distance. Unlike dipicolinic acid, CA has both a phenolic aromatic and a keto-diene resonance form2,3. The imino group in the non-aromatic keto form of CA can undergo nitrogen inversion and can cause elongation of the Ln-N bond, discussed in section 5.2.3. The carboxylic acid groups of CA and dipicolinic acid may not coordinate similarly. The hybridization of the amine orbitals changes in these resonance forms from “sp2” in the aromatic form to “sp3” in the diene form. The geometry of the N bonds would change from a planar confirmation (“sp2” = trigonal planar) to a non-planar confirmation (“sp3” = tetrahedral) as discussed in 5.2.3. The tetrahedral imino alters the geometry of the carboxylic acid groups on the CA molecule and could cause them to reorient out of the plane of the molecule and change the interaction of the carboxy groups with the lanthanide(III) ion. The reoriented carboxy 80

coordinations may not be similar to acetate and so the δ∆H value would not represent a Ln-N interaction. 5.3.3 Change of Entropy 5.3.3.1 PDA The heat of complexation of PDA is below the limit of detection of the calorimeter used in these experiments. The low heat of complexation is due to the weak lanthanide(III)-PDA complex, observed in the inability to measure the stability constants by potentiometry. 5.3.3.2 DPA NTA and DPA have similar ∆S values for the formation of the first Ho(III) complexes while DPA has a larger change of entropy that NTA upon formation of the first Nd(III) species (Table 4.6 and 5.5).

The ∆S value of the first complex between

Nd(III) and a series of aminopolycarboxylate ligands has been shown to be related to the number of inner sphere water molecules displaced during by complexation, Figure 5.214. The value for DPA fits the general trend of increased ∆S with increased number of carboxylate

bonds

but

is

slightly

higher

than

other

tricarboxylic

acid

aminopolycarboxylate ligands. The slightly larger ∆S values of the first complex of DPA versus NTA indicates that DPA displaces a larger number of water molecule than NTA upon coordination to Nd(III). This agrees with results from fluorescence experiments that noted increased water displacement from the Eu(III) inner hydration sphere upon coordination of DPA compared to NTA. The additional water molecules are displaced by fluctuations of the coordination site caused by the piperidine ring flip.

The ring flip causes reorientation of the

carboxylate and amine coordinations. The ∆S value of the second complex formation of DPA and NTA agree well for Ho(III). The second complex of Nd-DPA has a significantly higher ∆S value than the corresponding NTA complex, indicating that additional water molecules are displaced from the Nd(III) inner hydration sphere upon DPA complex formation. The increased dehydration of Nd(III) during the second complex formation is caused by the piperidine ring flip as discussed for the first complex formation. The significant difference between 81

the values of Nd(III) and Ho(III) are probably related the inter-ligand charge repulsion which would be greater for the smaller Ho(III) ion. The repulsion would increase the HoDPA coordination distances and so displace fewer inner sphere water molecules than the Nd(III) ion complex. 5.3.3.3 CA The La-CA ∆S value, Table 4.7, does not agree with the value previously reported (Table 5.6), but comparison is difficult due to the different ionic strengths. The ∆S values of complex formation are larger for CA than those of dipicolinic acid but follow the same trend of increased entropy as the lanthanide(III) radii decrease30. The ∆S value for the Nd-CA complex formation is higher than typical dicarboxylic acid aminpolycarboxylate ligands, shown in Figure 5.2. This indicates that CA displaces an additional water upon complex formation when compared to dipicolinic acid.

This

agrees with results from laser fluorescence which indicate that CA displaces additional water molecules from the Eu(III) inner coordination sphere compared to the number discplaced by complexation by dipicolinic acid. The additional waters are displaced from the inner hydration sphere of the lanthanide(III) ions because CA has varying carboxy and amine coordination geometry due to the keto tautomer and possible nitrogen inversion, discussed in section 5.2.3. The formation of the second complexes of Nd(III) and Ho(III) have ∆S values similar to the formation of the first complex and are much higher than the values observed with dipicolinic acid, Table 5.8. The fluorescence results indicate that a similar number of water molecules are displaced from the Eu(III) inner hydration sphere by the second complexation of both the CA and dipicolinic acid, discussed in section 5.5. The higher ∆S values for the second CA complex formation can be explained by additional reorganization caused by the phenol-keto functional group.

Upon coordination, the

deprotonated phenol would reorganize water molecules in the secondary lanthanide(III) hydration sphere as well as the primary phenol hydration sphere to stabilize the charge of the functional group. These additional displaced water molecules can contribute to both the first and second ∆S values for CA coordination. 5.4 Hypersensitivity Spectroscopy 82

5.4.1 PDA The change in the oscillator strength caused by complexation of PDA was too weak to allow the determination of a PML value. 5.4.2 DPA The values of the oscillator strengths determined in this study are listed in Table 4.10. Previously measured oscillator strengths of IDA, NTA, dipicolinic acid and CA are listed in Table 5.9. The oscillator strength of the Nd(III) and Ho(III) aquo ions are considered to have values of 9.68 x 10-6 and 6.12 x 10-6, respectively53. The oscillator strengths of the aquo ions determined in this study are within 2 or 3 standard deviations of these values. The oscillator strength values determined for the first complexes of DPA are within 1 standard deviation of the values reported for the first complex of NTA, molecules with a similar coordination site. The PML2 of DPA are higher than those reported for NTA and are similar to the value seen for IDA. The reported oscillator strength values of NTA show similar values for the first and second complex, indicating that the formation of the ML2 species significantly decreases the amount of covalent interaction between the ligands and the lanthanide(III) ion. The plots of P versus [L] : [M], Figures 4.1 and 4.2, have 3 distinct regions. The oscillator strength increases linearly with the concentration of the species, showing the formation of ML and ML2 complexes. After the ML2 species forms, the oscillator strength becomes constant, indicating that either any ML3 species does not form in significant concentrations or that the covalancy term of the ML3 does not affect the hypersensitive transition as is observed with NTA53. Since an ML3 complex is not expected to form with the tetradentate DPA ligand, the lack of this species seem reasonable and confirms the coordination model used in stability constant determination. 5.4.3 CA The PML values of CA determined in this work do not agree with the values found in the literature. The oscillator strength values of dipicolinic acid and IDA are much higher in the Devlin et al. and Stephens et al. studies compared to the values reported by Fellows et al. as listed in Table 5.9. It is not readily apparent why the dipicolinic acid 83

and IDA values of these studies have such large differences since similar values were determined for the aquo ions. Since the Devlin et al. and Stephens et al. studies were performed by the same scientific group but reported significantly different oscillator strength values for IDA and dipicolinic acid, use of these values for comparison seems unreliable. Most of the CA oscillator strengths agreed, within 1 standard deviation, with the values reported Fellows et al. for dipicolinic acid, a molecule with a similar coordination sites as CA. The value determined for Ho-CA was lower than the value reported for Ho-dipicolinic acid, 12.15 ± 0.21 versus 14.26. The value of the Hodipicolinic acid from Fellows et al. appears to be higher than would be expected as oscillator strengths of Nd(III) are higher than those of Ho(III) for oscillator strength values below 20 and may reflect an increased hypersensitivity between Ho(III) and dipicolinic acid52,53. The plots of P versus [L] : [M], Figures 4.1 and 4.2, have different profiles for Nd(III) and Ho(III). Figure 4.1 shows 3 regions with positive slope, indicating the formation of 3 different complexes with Nd(III) which are most probably ML, ML2 and ML3 species. Figure 4.2 has 2 regions with positive slope and one region of constant values, indicating that Ho(III) forms two complexes with CA and the third species either forms in low concentrations or that the covalancy term of that species does not affect the hypersensitive transition as discussed with DPA in section 5.4.1. The different ML3 profiles indicates that the size of the lanthanide(III) ion may prevent formation of the third CA complex due to increased steric crowding and inter-ligand charge repulsion. 5.5 Eu(III) Laser Fluorescence 5.5.1 PDA The Eu(III) fluorescence peaks, coordination number of the ligand and the number of water molecules in the inner hydration sphere of the Eu(III) ion for the various species of this study and other ligands of interest are listed in Table 5.10. The number of water molecules was calculated assuming that the inner coordination sphere of the aquo Eu(III) ion is occupied by 9 water molecules89,104. A single peak was observed at 17273 cm-1 for PDA complexation throughout all studies including varying pH from 2 to 6 and from ligand to metal concentration ratios of 0.1 to 5. The peak position of the PDA 84

species is the same as the value reported for the first acetate complexation (Table 5.8). This indicates that PDA is forming only a weak complex through a single carboxylic acid which displaces a single water from the inner coordination sphere of the Eu(III) ion. The number of water molecules in the inner hydration sphere of the Eu(III) ion could not be determined because the PDA peak was too weak under all conditions. 5.5.2 DPA The Eu(III) fluorescence spectra of the previously reported NTA complex was performed on solid material and so comparisons to solution systems have to be made with the consideration that the aqueous system has a less rigid coordination system in which the coordinated ligands can still rapidly exchange. The ML species of DPA exists in two different environments (Table 5.10). The first Eu-DPA environment has a similar peak position as NTA (Table 5.10). The number of water molecules in the inner hydration sphere of Eu(III) of DPA is slightly lower than the value reported for NTA and the ligand coordination number of DPA is slightly higher than the value reported for NTA. The difference in number of water molecules and ligand coordination number indicates that the DPA complex occupies a larger area in the inner coordination sphere of Eu(III) than the NTA complex due the piperidine ring flip. The ring flip would cause the carboxy and amine coordination groups to reorganize around the metal ion.

This reorientation

corresponds to the hump that was always observed on the peak of the first complex that represents the conversion from one ring form to another. 5.5.3 CA The peak positions of the CA complexes with Eu(III), listed in Table 4.11, agree the values reported previously within 1 or 2 cm-1 (Table 5.8). The calculated ligand coordination number is within 1 standard deviation of the values reported in the literature. The peak positions values of dipicolinic acid from previous studies generally vary within 1 or 2 cm-1. The ligand coordination numbers are similar to the values reported for CA. The first dipicolinic acid complex has a Eu(III) hydration number of 3.2, indicating that only 3 waters are displaced. The number of water molecules in the inner sphere of the Eu(III) ion are slightly lower than those predicted in the literature which assumed the tridentate CA molecule would displace 3 water molecules. The CA molecule displaced 4 85

to 5 water molecules upon forming the first complex due to the keto tautomer. As the molecule switches between the keto and phenol tautomers, the coordination sites would change geometric positions and additional water molecules are displaced. The second CA complex has an average of 3 water molecules displaced per ligand. The decrease in the number of displaced water molecules indicates that the CA molecules of the second complex have a slightly longer coordination bonds due to inter-ligand charge repulsion, causing each ligand to occupy a smaller area in the inner coordination sphere of the Eu(III) ion. 5.6 NMR 5.6.1 PDA The 1H NMR of Eu-PDA system was performed at varying pH and did not exhibit the shift in the peaks expected upon the complexation of the paramagnetic Eu(III) ion. Previous work has studied the lanthanide(III) complexes of dipicolinic acid105 and CA106 by NMR and PMR. Diamagnetic lanthanide(III) ion shift the position of the 1H NMR peaks of the complexed ligands107. This shift was not observed in the Eu-PDA system. Additionally, the PDA peaks did not show significant 1H NMR peak broadening that would have been expected if the paramagnetic Eu(III) complexed the ligand17. This indicates that the Eu-PDA complex was only a minor species that was not detected by the NMR measurement. 5.6.2 DPA The 1H NMR spectra of La(III) and Lu(III) DPA complexes have very different (Appendix C). The La(III)-DPA spectrum has very broad peaks while the 1:1 and 1:2 Lu(III)-DPA spectra have sharp peaks that retain their multiplicity. The spectra of DPA and the 1:1 [Lu(III)]:[DPA] system had the same number of peaks but all peaks had different chemical shifts. The sharpness of the peaks indicates that Lu-DPA kinetics were faster than the NMR time scale. The broadness observed in the peaks of the 1:1 [La(III)]:[DPA] spectrum indicates that this complex is undergoing an equilibrium on the NMR time scale with the broad peaks representing the average of two or more systems. This suggests that DPA complex formation is sensitive to changes in the size of the lanthanide ion with larger lanthanide(III) ions having slower reaction rates. The 1:1 86

[La]:[DPA] system was studied temperatures as low as 5.0ºC but the peaks remained coalesced. The equilibrium may be the DPA piperidine ring flip with the broad peaks representing the transition from one chair form of DPA to the other chair form, involving boat and twist boat configurations of the ring. The 1H NMR peaks of 1:2 [Lu]:[DPA] systems are sharp and retain their multiplicity but have peaks that are also in the 1:1 [Lu]:[DPA] systems as well as additional peaks. The similar peaks in the 1:1 and 1:2 [Lu]:[DPA] system indicates that one of the DPA ligands in the ML2 complex is similar to the DPA in the ML species while the additional peaks in the 1:2 [Lu]:[DPA] system indicates that the second DPA ligand in the ML2 species must be in a different coordination environment. The steric bulk and inter-ligand charge repulsion of the second complexed DPA must cause this molecule to have a different orientation around the metal ion. 5.6.3 CA 13

C NMR experiments were performed on CA as well as on the Ca complexes of

La(III) and Lu(III). The 13C NMR peaks of the carbon shifted slightly. The alkene and aromatic carbon peaks are located at similar chemical shift and any effects on the tautomer specition of CA could not be discerned using this technique. 5.7 Crystallography The crystal structure of CA has been previously determined70. The molecule was observed as a monohydrate zwitterion with a protonated amine and a deprotonated carboxylic acid. The other carboxylic acid and the phenol were also protonated. The crystal structure of the Gd-CA complex102 has also been determined and shown to exist as a Na5Gd(CA)2(HCA) complex with 16 waters of hydration. Two CA molecules are completely deprotonated while one complex has a phenolic hydrogen. The authors fail to mention how the hydrogen atom position was assigned and also state that the water molecules were disordered. The phenolic proton could be reassigned as a proton of a water molecule that was hydrogen bonded to the CA, linking the Eu(CA)3 units. Several studies report coordination distances for metal ion and the amine and carboxy oxygen atoms for CA18,108 and dipicolinic acid20,109,110.

The coordination

distances of Fe(III)OH(L), when L = CA or dipicolinic acid, have values that agree 87

within 2 standard deviations20. The coordination distance of [Gd(HxL)3]-n species, where L = CA or dipicolinic acid, are slightly longer for the CA18 system than the values calculated for dipicolinic acid18. The Gd(III) coordination distances reported in [18] were calculated from data reported in [19]. Although errors are not reported for the dipicolinic acid values, the coordination distances probably agree within 2 standard deviation values. The bond lengths reported for [Ho(Hdpic)(dipic)] species21 agree with the bond lengths reported for the [Gd(dipic)3]2- species19 within 0.5%. The CA ring structure is more strained than the dipicolinate ring, observed by a ~3% lengthening of the bonds to the carbon in the para position relative to the amine. The crystal structures determined for PDA exhibited ~30% zwitterion behavior, calculated from the relative proton density as calculated by least square refinement where one carboxylic acid was founded to only account for ~70% of a proton. The DPA molecule was founded exclusively as the zwitterion with full deprotonation of a carboxylic acid group. 5.8 Conclusions 5.8.1 PDA The acid dissociation constants determined by this work indicate that PDA is triprotic instead of a diprotic acid as reported in the literature63,64. The results from this work strongly disagree with the results reported by Thompson64 for the lanthanide(III) stability constants of PDA. The PDA of this study was identified as the cis form of the compound by diffractometry. The ligand of the previous study might be suspect as it was not synthesized by the author and was characterized using elemental analysis.

All

techniques used in this work reported weak, monodentate complexation of lanthanide(III) ions by PDA. The coordination structural analog, IDA, has first stability constants of ranging between 105.31 to 106.71 for the lanthanide(III) series.

The dissimilarity in

complex strength of IDA and PDA can be explained by the PDA piperidine ring flip that reorients the coordination site, preventing chelate formation. The imino atom of the PDA molecule would also have a nitrogen inversion during the ring flip that would destabilize and lengthen any Ln-N coordination bonds. 5.8.2 DPA 88

The acid dissociation constants and lanthanide(III) stability constants show good agreement with the values previously reported for DPA. The DPA stability constants increase as the size of the lanthanide(III) radii decrease for both the first and second complexes. The complex formation is entropy driven and displaces 5 and 8 water molecules as DPA coordinates in tetradentate fashion to form first and second complexes, respectively. The larger number of water molecules displaced by the first complex and the decreased lanthanide(III) complex stability can be explained by the DPA piperidine ring flip which would reorient the coordination groups, causing the coordination groups to displace additional water molecules from the inner coordination sphere of the metal ion. Larger lanthanide(III) ions appear to have slower ring reorientation as observed in 1

H NMR experiments, indicating that smaller lanthanide(III) ions are a better fit for the

coordination site of DPA. The endothermic ∆H values also indicate that the average coordination distances may be longer than in straight chain aminopolycarboxylate ligands. The longer coordination distances would be caused by the coordination site reorientation caused by the piperidine ring flip. This ring flip reorientation kinetics may explain the improved the lanthanide(III) size selectivity of DPA compared to a straight chain molecule NTA. 5.8.3 CA The acid dissociation constants show general agreement with those found in the literature. The CA stability constants are slightly lower and have less change across the lanthanide(III) series compared to dipicolinic acid. The formation of the first complex is entropy driven and displaces 4 to 5 water molecules from the inner hydration sphere. The additional water molecules displaced during the complexation of the first ligand can be explained by the keto tautomer that removes the planar geometry of both the amine and the carboxy groups. The formation of the second complex is entropy driven and each ligand displaces 3 water molecules from the inner hydration sphere. The size of the lanthanide(III) ion is important in the formation of the ML3 species. The hypersensitive spectrum of Nd clearly indicates the formation of a ML3 complex while this species is not observed for the Ho(III) system (Figure 4.1 and 4.2). The tautomer forms of CA appear

89

to create a larger, fluctuating coordination site that reduces the lanthanide(III) size selectivity compared to dipicolinic acid. 5.8.4 Steric Hindrance of Aminopolycarboylic Ligands 5.8.4.1 Piperidine Ring Aminopolycarboxylate ligands with carboxy groups directly attached to the piperidine ring appear to decrease the complexation stability.

The ring flip of the

piperidine ring can reorient the coordination groups. In the case of PDA, the ring flip dominates the lanthanide(III) coordination chemistry and reduces the ligand to the formation of very weak complexes similar to those observed with acetate. DPA forms stable complexes but displaces additional water molecules upon formation of the first complex compared to the coordination structure analog NTA. The additional displaced water molecules can be observed both by the decreased Eu(III) waters of hydration of the complex measured by fluorescence and the increased change of entropy of complex formation measured by calorimetry.

DPA does not follow the trend of increasing

δ∆H(N) with increasing ΣpKa(N) that has been noted for straight chain aliphatic aminopolycarboxylate ligands. The deviation from of δ∆H(N) is most likely related to the piperidine ring flip that would weaken the carboxylate coordinations and the Ln-N coordination by increased coordination bond lengths, as observed in the change of enthalpy determined by calorimetry. NMR experiments indicate that the piperidine ring flip appears to be less significant for smaller lanthanide ions like Lu(III) and may be the cause of the increased size selectivity of DPA compared to NTA. 5.8.4.2 CA The tautomerization of CA causes the lanthanide(III) complexes to be less stable than with dipicolinic acid. The non-aromatic tautomer introduces non-planar alignment of both the amine and the carboxylic acid groups. The fluctuation of the coordination sites causes additional water molecules to be displaced from the inner hydration sphere as observed in both number of waters in the inner coordination sphere of Eu(III) as determined by flourescence and the change of entropy determined by calorimetry. The fluctuation of the coordination groups increases the lanthanide(III) coordination bond lengths as seen in the change of enthalpy determined by calorimetry. The size of the 90

lanthanide(III) ions appears to have more influence on the formation of ML3 complex with CA than with dipicolinic acid due to the increased steric crowding caused by fluctuation of the coordinating groups of the CA molecule. 5.9 Future Work The coordination bond lengths of PDA, DPA and CA appear to be longer due to effects caused by the ring structures. CA appears to have different coordination geometry in solution than in solid state. Experiments using EXAFS techniques could measure the coordination bond lengths and examine the CA geometry to confirm any differences from the values of the solid state materials. Synthesis of a methoxy form of chelidamic acid may eliminate the tautomerization and the resulting compounds should behave more similarly to dipicolinic acid. Although 5.0 ºC NMR experiments were not able to resolve the coalesced La-DPA peaks, low temperature studies might give insight into the kinetics of the ring flip and the influence of the ring flip on the coordination chemistry of DPA and PDA. Ring structures could be incorporated into future ligand design if the enhanced size lanthanide(III) selectivity of DPA over NTA is related to the kinetics of the DPA piperidine ring flip. Additional studies of the lanthanide(III) coordination chemistry with ligands such as piperidine-2-carboxylic acid and 2-carboxypiperidine-N-acetic acid may establish

trends

in

the

thermodynamic

aminopolycarboxylic ligands

91

values

for

structurally

hindered

Table 5.1. Acid dissociation constant values for PDA, DPA and CA at 0.1 M I. ligand PDA DPA

CA

log Ka1

log Ka2

9.92

log Ka3

ionic media

T (°C)

ref

2.87

0.1 M KCl

30

63*

10.12

2.54

0.1 M KNO3

25

64*

9.33

2.71

1.3

0.1 M KNO3

25

73*

10.88

3.11

1.4

0.1 M NaNO3

20

67*

10.85

3.18

1.9

0.1 M NaClO4

22

68*

11.4

3.47