Ligand Exchange Processes on Solvated Lithium Cations · • “Ligand Exchange Processes on...

Ligand Exchange Processes on Solvated

Lithium Cations

Den Naturwissenschaftlichen Fakultäten

der Friedrich-Alexander-Universität Erlangen-Nürnberg

zur

Erlangung des Doktorgrades

vorgelegt von

Ewa Maria Pasgreta

aus Więcbork (Polen)

Als Dissertation genehmigt von den Naturwissenschaftlichen Fakultäten der Universität Erlangen-Nürnberg

Tag der mündlichen Prüfung:

Vorsitzender der Promotionskommission:

Erstberichterstatter:

Zweitberichterstatter:

23. 3. 2007

Prof. Dr. Eberhard Bänsch

Prof. Dr. Dr. h.c. mult. Rudi van Eldik

Prof. Dr. Lutz Dahlenburg

Meiner Familie

5

Acknowledgements

I wish to thank Prof. Dr. Dr. h.c. mult. Rudi van Eldik for the supervision of

this work and many fruitful discussions. I am particularly grateful for his

advice, encouragement, his trust in me, and helping me to present my projects

internationally. I appreciate very much the good working conditions and the

support that I could always count with.

Dr. Achim Zahl is gratefully acknowledged for all, sometimes very demanding,

NMR measurements.

I would like to thank Prof. Dr. Lothar Helm (EPFL Lausanne) for introducing

me to NMRICMA 2.7 program and for vary fruitful discussions.

I am very grateful towards Dr. Ralph Puchta for theoretical calculations, for

correcting the German version of the summary and for his constant help and

encouragement.

I would like to thank Dr. Maria Wolak and Joanna Procelewska for their help,

in particular, at the beginning of my stay in Erlangen, and for their continuous

support. I also thank Dr. Ivana Ivanović-Burmazović for many interesting

discussions and Dr. Immo Weber for his help in the laboratory work.

I am also highly indebted to my colleagues from the practical courses: Dr. Kai

Peters and Dr. Hans Hanauer for their help at the beginning and Dr. Jelena

Galinkina and Klaus Pokorny for their support in the last months.

6

Many thanks to my room-mates: Christian, Erika, Alex, Fanny and Sabine as

well as to all members of the group and my friends outside the institute.

I would like to thank people involved in the European project which financed

part of my work: Kersti, Jan, Casey, Michael, Rossend, Wiktor, Klaus, Marco,

Daniel, Sami, Bruno, Gregor, and Barbara.

I am very grateful towards my family for their support and encouragement

during all these years.

Last but not least, I would like to thank Radim for everything. Dziękuję Ci

bardzo!

Die vorliegende Arbeit entstand in der Zeit von November 2001 bis Dezember

2006 am Institut für Anorganische Chemie der Friedrich-Alexander-

Universtität Erlangen-Nürnberg.

7

Publications and Conference Contributions

Publications

• Ralph Puchta, Michael Galle, Nico van Eikema Hommes, Ewa

Pasgreta, Rudi van Eldik, „Evidence for Associative Ligand Exchange

Processes on Solvated Lithium Cations“, Inorg. Chem., 2004, 43, 8227-

8229

• Ewa Pasgreta, Ralph Puchta, Michael Galle, Nico van Eikema

Hommes, Achim Zahl, Rudi van Eldik, „Ligand Exchange Processes on

Solvated Lithium Cations Part II – Complexation by Cryptands in γ-

Butyrolactone as Solvent“, J. Inclusion Phenom. Macrocyclic Chem.,

2007, 58, 81-88.

• Ewa Pasgreta, Ralph Puchta, Michael Galle, Nico van Eikema

Hommes, Achim Zahl, Rudi van Eldik, „Ligand Exchange Processes on

Solvated Lithium Cations Part III – DMSO and Water/DMSO Mixtures“,

ChemPhysChem, 2007, 8, 1315-1320.

• Ewa Pasgreta, Ralph Puchta, Achim Zahl, Rudi van Eldik, „Ligand

Exchange Processes on Solvated Lithium Cations Part IV – Acetonitrile

and Hydrogen Cyanide“, Eur. J. Inorg. Chem., 2007, 1815-1822.

• Ewa Pasgreta, Ralph Puchta, Achim Zahl, Rudi van Eldik, „Ligand

Exchange Processes on Solvated Lithium Cations Part V – Complexation

by Cryptands in Acetone as Solvent“, Eur. J. Inorg. Chem., 2007, 3067-

3076.

8

Conference Contributions

• “7Li NMR Study of the Kinetics of Lithium Ion Complexation by C222, C221

and C211 in different solvents”, 35th International Conference on Coordination

Chemistry, Heidelberg, Germany, July 2002, poster presentation.

• “7Li NMR Study of Lithium Ion Complexation by C222, C221 and C211 in

different solvents”, Graduiertenkolleg Symposium, Hirschegg-Kleinwalsertal,

Austria, March 2003, oral presentation.

• „High Pressure 7Li NMR Kinetic Studies of Li+ Complexation by Cryptands in

Acetone“, Dalton Discussion Meeting on Inorganic Reaction Mechanisms,

Athens, Greece, January 2004, poster presentation.

• “The Mechanism of Ligand Exchange on [Li(H2O)n]+ and [Li(NH3)n]+”,

Chemistry Symposium Erlangen-Rennes, Erlangen, Germany, June 2004,

poster presentation.

• “7Li NMR Studies of the Kinetics of Lithium Ion Complexation by Cryptands in

Gamma-Butyrolactone as Solvent”, The 2004 Younger European Chemists’

Conference, Torino, Italy, August 2004, poster presentation.

• „7Li NMR Studies of the Kinetics of Lithium Ion Complexation by Cryptands in

Gamma-Butyrolactone as Solvent“, Catalysis Summer School, Liverpool, Great

Britain, September 2005, poster presentation.

• “High Pressure 7Li NMR Kinetic Study of Li+ Complexation by Cryptands in

Gamma-Butyrolactone”, Dalton Discussion Meeting on Inorganic Reaction

Mechanisms, Cracow, Poland, January 2006, poster presentation.

• “Ligand Exchange Processes on Solvated Lithium Cations. Complexation by

Cryptands in γ-Butyrolactone as Solvent”, Chemistry Symposium Erlangen-

Rennes, Rennes, France, June 2006, poster presentation.

9

Abbreviations

∆H# activation enthalpy

∆S# activation entropy

∆V# activation volume

h Planck constant, 6.6261 × 10-34 J s

kb Boltzmann constant, 1.3807 × 10-23 J K-1

k rate constant

kobs observed rate constant

kon forward reaction rate constant

koff backward reaction rate constant

R gas constant, 8.3145 J K-1 mol-1

T temperature

t time

A associative reaction mechanism

D dissociative reaction mechanism

I interchange reaction mechanism

IA interchange associative reaction mechanism

ID interchange dissociative reaction mechanism

C211 4,7,13,18-Tetraoxa-1,10-diazabicyclo[8.5.5]eicosane

C221 4,7,13,16,21-Pentaoxa-1,10-diazabicyclo[8.8.5]tricosane

C222 4,7,13,16,21,24-Hexaoxa-1,10-diazabicyclo[8.8.8]haxacosane

LiOTf Lithium Trifluoromethanesulfonate

CH3CN acetonitrile

10

MeOH methanol

EC ethylene carbonate

PC propylene carbonate

DMF N, N-dimethylformamid

DMSO dimethylsulfoxide

GBL γ-butyrolactone

DFT density functional theory

B3LYP Becke’s 3 Parameter Hybrid Functional using Lee-Yang-Paar

Correlation

HF Hartree-Fock

MP Møller-Plesset Perturbation

NMR nuclear magnetic resonance

DNMR dynamic nuclear magnetic resonance

11

Table of Contents

1. INTRODUCTION ............................................................................... 13

1.1. A EUROPEAN PROJECT ...................................................................................... 14 1.2. OUTLINE OF THIS THESIS .................................................................................. 14 1.3. LITHIUM........................................................................................................... 17 1.4. SOLVATION .......................................................................................................18

1.4.1. Conventional Solvents ........................................................................................19 1.4.2. Plasticizers ......................................................................................................... 20

1.5. CRYPTANDS...................................................................................................... 22 1.6. EXCHANGE PROCESSES..................................................................................... 26 1.7. KINETIC APPLICATIONS OF NMR SPECTROSCOPY .............................................. 29 1.8. ALKALI METAL NMR SPECTROSCOPY ............................................................... 32 1.9. PROGRAMS ...................................................................................................... 34 1.10. COMPUTATIONAL METHODS ......................................................................... 36

1.10.1. Ab-initio Methods...............................................................................................37 1.10.2. Density Functional Theory ................................................................................ 38

1.11. REFERENCES.................................................................................................... 40

2. LIGAND EXCHANGE PROCESSES ON SOLVATED LITHIUM CATIONS. DMSO AND WATER/DMSO MIXTURES .............................43

2.1. GENERAL REMARK ........................................................................................... 43 2.2. ABSTRACT........................................................................................................ 43 2.3. INTRODUCTION ................................................................................................ 44 2.4. EXPERIMENTAL SECTION.................................................................................. 46

2.4.1. General Procedures............................................................................................ 46 2.4.2. 7Li NMR Spectroscopy....................................................................................... 46 2.4.3. Quantum-Chemical Calculations .......................................................................47

2.5. RESULTS AND DISCUSSION ................................................................................47 2.5.1. Coordination Number of Li+ in DMSO ..............................................................47 2.5.2. Selective Solvation of Li+ ................................................................................... 49 2.5.3. Ligand Exchange on Solvated Li+...................................................................... 52 2.5.4. Computed Mechanisms of Solvent Exchange................................................... 52

2.6. CONCLUSIONS ...................................................................................................55 2.7. REFERENCES.....................................................................................................57

3. LIGAND EXCHANGE PROCESSES ON SOLVATED LITHIUM CATIONS. ACETONITRILE AND HYDROGEN CYANIDE ....................63

3.1. GENERAL REMARK ........................................................................................... 63 3.2. ABSTRACT........................................................................................................ 63 3.3. INTRODUCTION ................................................................................................ 64 3.4. EXPERIMENTAL SECTION...................................................................................65

3.4.1. General Procedures............................................................................................ 65 3.4.2. 7Li NMR Spectroscopy....................................................................................... 66 3.4.3. Quantum-Chemical Calculations ...................................................................... 66

3.5. RESULTS AND DISCUSSION ............................................................................... 66 3.5.1. Coordination Number of Li+.............................................................................. 66 3.5.2. Selective Coordination of Li+............................................................................. 69 3.5.3. Ligand Exchange on Solvated Li+....................................................................... 71 3.5.4. Quantum-Chemical Calculations ....................................................................... 71

12

3.6. CONCLUSIONS................................................................................................... 77 3.7. REFERENCES ....................................................................................................78

4. LIGAND EXCHANGE PROCESSES ON SOLVATED LITHIUM CATIONS. COMPLEXATION BY CRYPTANDS IN ACETONE AS SOLVENT ................................................................................................. 84

4.1. GENERAL REMARK ............................................................................................84 4.2. ABSTRACT.........................................................................................................84 4.3. INTRODUCTION.................................................................................................85 4.4. EXPERIMENTAL SECTION ..................................................................................87

4.4.1. General Procedures ............................................................................................87 4.4.2. 7Li NMR Spectroscopy........................................................................................87 4.4.3. Programs............................................................................................................ 88 4.4.4. Quantum-Chemical Calculations.......................................................................89

4.5. RESULTS AND DISCUSSION ................................................................................89 4.5.1. Spectroscopic and Kinetic Observations ...........................................................89 4.5.2. Suggested Exchange Mechanism.......................................................................93 4.5.3. High Pressure 7Li NMR Measurements.............................................................99 4.5.4. Theoretical Calculations...................................................................................100

4.6. CONCLUSIONS.................................................................................................104 4.7. REFERENCES ..................................................................................................106

5. LIGAND EXCHANGE PROCESSES ON SOLVATED LITHIUM CATIONS. COMPLEXATION BY CRYPTANDS IN γ-BUTYROLACTONE AS SOLVENT ........................................................................................... 112

5.1. GENERAL REMARK ...........................................................................................112 5.2. ABSTRACT........................................................................................................112 5.3. INTRODUCTION................................................................................................113 5.4. EXPERIMENTAL SECTION .................................................................................116

5.4.1. General Procedures ...........................................................................................116 5.4.2. 7Li NMR Spectroscopy.......................................................................................116 5.4.3. Programs............................................................................................................117 5.4.4. Quantum-Chemical Calculations......................................................................117

5.5. RESULTS AND DISCUSSION .............................................................................. 118 5.5.1. Spectroscopic Observations ............................................................................. 118 5.5.2. Suggested Exchange Mechanism..................................................................... 122 5.5.3. High Pressure 7Li NMR Measurements........................................................... 126 5.5.4. Influence of Water ............................................................................................ 127 5.5.5. Theoretical Calculations................................................................................... 127

5.6. CONCLUSIONS................................................................................................. 130 5.7. REFERENCES ...................................................................................................131

6. SUMMARY ........................................................................................137

7. ZUSAMMENFASSUNG ....................................................................144

1. Introduction

13

1. Introduction

Lithium, the lightest of all metals and the least reactive, has been the object

of numerous experimental and theoretical studies. One might say that lithium

chemistry is a branch of chemistry in itself. Anomalous properties of lithium,

because of its small size, include such phenomena as the lithium bond (similar

to the hydrogen bond), formation of aggregates, and complexation by crown

ethers or cryptands.1

Lithium was first discovered in 1817 by the Swede Johann August

Arfvedson and pure lithium metal was first prepared the following year by

Davy and Brande by electrolysis of lithium oxide. Lithium’s name originates

from Greek word “lithos” meaning stone and it is the lightest of all solids (0.53

g/cm3). The industrial process by which lithium is prepared today is

electrolysis of lithium chloride or lithium carbonate. Today, the amount of

lithium produced yearly is about 10.000 tons.2

Coordination chemistry of lithium and other alkali metals was viewed until

the 1960’s as a rather barren area, made so since the singly charged and

relatively large M+ ions were held to have rather poor coordinating ability. The

majority of the known metal salt complexes were hydrates. Two developments

have transformed this research area over past the 40 years. The first began

with the Nobel prize-winning discovery by Pedersen, Lehn and Cram of

macrocyclic ligands like “crown” ethers, cryptands, and spherands. Such

ligands wrap around the metal ion, encapsulating it to greater or lesser degree,

and are highly selective regarding which metal cation can be accommodated.

This feature has led to important application in cation separation, ion-specific

measuring devices, and phase-transfer catalysis, and as mimics of naturally

occurring macrocyclic antibiotics. The second major development was brought

about by extensive investigation of non-aqueous alkali metal chemistry

prompted largely by the importance of lithium reagents in organic synthesis.3

1. Introduction

14

1.1. A European Project

The research work presented in this thesis has been developed within the

framework of the European Research Training Network on Solvation

Dynamics and Ionic Mobility in Conventional Solvents and

Plasicizers – by Computational Chemistry and Experiment. The

overall goal of this Network (which involved the collaboration between six

European research groups) was to seek firm knowledge about the mechanisms

of the dynamical processes present, and their relative importance. The primary

scientific aims were:

• to discover unifying and differentiating features in ligand and solvent

exchange processes around metal ions in conventional and polymer

solvents;

• to study the connection between solvation dynamics and ionic

diffuson/conduction mechanism in these systems;

• to help in the search for new applied systems by providing a more

detailed understanding of the basic phenomena occurring in model

electrolytes

In this thesis these aims are approached from the microscopic point of

view, making use of both computer calculations and experimental techniques.

The way solvent molecules interact with the ion, and how this interaction

affects the physico-chemical properties of the system are studied.

1.2. Outline of this Thesis

The contributions of the present work to the network objectives can be

divided into three main topics which constitute the backbone of this thesis:

Solvation: how is the first coordination sphere around Li+ constituted;

how does the nature of the solvent influence solvation;

1. Introduction

15

Solvent exchange: the exchange process between the first and second

coordination spheres of the Li cation is studied; the interplay between

diffusion and the exchange mechanism is addressed.

Lithium complexation by cryptands and ligand exchange in

conventional solvents and plasticizers: how is Li+ complexed by

different cryptands; how do solvents of different properties influence the

process of Li+ exchange between complexed and solvated sites.

The Li+ cation remained the centre of our interest: its solvation shell and

the way it is solvated. Both experimental and theoretical methods were

employed. Solvents of different properties and their influence on the solvation

shell were investigated. Questions concerning selective solvation around Li+

and the solvation number of the Li+ cation were studied as a function of the

solvent properties. The Gutmann Donor Number was mostly considered as a

factor describing the donating abilities of the solvent. However, we were not

only interested in thermodynamic phenomena. Kinetic measurements were

also of primary importance. Once the solvation shell of the Li cation was

known, we wanted to investigate the exchange of certain solvent molecules

around Li+. As Li+ is a very labile cation, solvent exchange was found to be very

fast on the NMR time scale and this process was therefore also investigated by

computational methods.

Solvent exchange is the simplest ligand exchange process, where, in fact,

there is no net change in the composition of the first solvation shell. However,

this process helps to understand the exchange of Li+ between different

environments, i.e. between solvent and another ligand.

Macrocyclic ligands called cryptands were found to bind the alkali and

alkaline earth metal ions very efficiently. They have been extensively explored

with respect to their structure and binding properties. Although there is a

wealth of information about thermodynamic and structural studies on the

complexation of metal ions by macrocyclic ligands in different solvents, kinetic

and mechanistic studies have received less attention and hence are still poorly

understood. For a better understanding of the pronounced selectivities of

1. Introduction

16

macrocyclic ligands towards alkali and alkaline earth cations, it is necessary to

investigate their mechanisms of formation and dissociation. In this

contribution, we were particularly interested in the mechanistic aspects of the

Li+ exchange process between cryptands and solvent. In the case of many

solvents the exchange was very fast or, on the contrary, in the case of small

cavity ligands, there was almost no exchange observed. Therefore, two

solvents, viz. acetone and γ-butyrolactone, were chosen for more detailed

mechanistic investigations. They possess different properties and applications,

such that the obtained results could be of importance in different application

fields.

The experimental method used throughout was 7Li NMR. During the past

few decades it has been shown that multinuclear NMR is a very powerful

technique for the study of alkali complexes with macrocyclic ligands,

particularly in non-aqueous solutions. 7Li is active in NMR and possesses

features that make it convenient for this kind of investigation. Particular

attention was given to high pressure 7Li NMR, which is the first example

reported for this kind of research. High pressure NMR is a powerful tool

employed to elucidate details of the reaction mechanism. The determination of

activation volumes from the pressure dependence of exchange rates for

solvated metal ions is a useful method for the diagnosis of the intimate

mechanism.

The outline of this thesis is as follows: Part I is dedicated to introduce the

main topics studied and to give a brief introduction of the methods used. Parts

II, III, IV and V contain the results and discussion of specific studies on

Solvation of the Li Cation in DMSO and DMSO/water mixure, Solvation of

the Li Cation in CH3CN and CH3CN/water mixure, Ligand Exchange in

Conventional Solvent and Ligand Exchange in Plasticizer, respectively.

Finally, concluding remarks are made in Parts VI and VII (in German).

1. Introduction

17

1.3. Lithium

Lithium is the lightest of all metals. It has the highest specific heat of any

solid element and therefore is very useful in heat transfer applications. Lithium

is also a leading contender as battery anode material due to its high

electrochemical potential. The lithium ion is exceptionally small and has,

therefore, an exceptionally high charge to radius ratio. As a result, its

properties are considerably different from similar ions such as sodium and

potassium. Contrarily, its hydrated radius is much larger than similar ions and

therefore exhibits different solution properties such as very low vapor

pressure, freezing point, and other colligative properties. The large solvation

shell around the lithium ion also causes it to have lower mobility in solution.

The aforementioned properties of the lithium ion contribute to its unique

behavior and effects in a variety of environments. The chloride salt of lithium

is one of the most hygroscopic materials known and is therefore used in many

air conditioning and drying systems. In general, the lithium ion has been well

documented to have profound biological effects of varying intensity on many

life forms. Numerous publications confirmed its influence on enzyme activity,

metabolism, respiration, and active transport. A prominent effect of lithium on

the nervous system can be found in the field of medicine.4, 5 Here it is used to

treat severe mental disorders such as manic-depression by controlling mania

and stabilizing mood swings.6-16 Although its therapeutic value has been

debated, there is no doubt that a number of manic patients have benefited

from lithium treatment. The lithium ion seems to counteract the manic

symptoms in a specific way, however, little is understood about the mechanism

of its therapeutic action. In conclusion, it is evident that lithium can have some

profound effects on a variety of chemical and biological systems. Although it

has been intensively studied, little is known about the details of the action of

lithium including required concentrations and mechanisms by which changes

occur.

1. Introduction

18

1.4. Solvation

Solute-solvents studies involving electrolytes have been a subject of interest

to solution chemists for quite sometime. Such studies on the transport

properties of electrolytes in aqueous and non-aqueous solvents are of interest

in various technologies like high energy density batteries, photo-

electrochemical cells, electrodeposition, wet electrolytic capacitors and in

electro-organic synthesis.

Ion solvation is a phenomenon of fundamental interest in many aspects of

chemistry because solvated ions are almost omnipresent on Earth. Hydrated

ions occur in aqueous solution in many chemical and biological systems. Ions

solvated in organic solvents or mixtures of water and organic solvents are also

very common. Selective solvation occurs when the solvate prefers one of the

solvents over the other. This phenomenon can be used to separate different

ions from each other, a process known as solvent separation. The exchange of

solvent molecules around ions in solution is fundamental to the understanding

of the reactivity of ions in solution. Solvated ions also play a key role in

electrochemical applications, where for instance the conductivity of

electrolytes depends on the ion-solvent interactions.17

A literature survey suggests that the use of mixed solvents involving high

permittivity, low viscosity, and a wide liquid range is practicable for use in high

energy density batteries, especially lithium batteries. The advantage of using

mixed solvents over their pure counterparts in lithium batteries is to

manipulate the suitable properties of these solvents for fulfilling the above

conditions.

It is known that aprotic solvents show high permittivities, low viscosities,

and strong temperature dependences. Hence, in order to develop suitable

mixed solvents fulfilling the conditions mentioned above, mixtures of solvents

must be prepared to balance the drawbacks of pure solvents. The

conductivities of lithium perchlorate in different solvent systems are explained

1. Introduction

19

in terms of solution viscosity and the donor-acceptor properties of the

solvents, which mainly govern the ion-solvent interaction.

1.4.1. Conventional Solvents

Solvents and solutes can be broadly classified into polar and non-polar.

There are a number of factors describing solvent properties. Polarity can be

measured as the dielectric constant or the dipole moment of a compound.

Polar solvents can be further subdivided into polar protic solvents (e.g. water

or methanol) and polar aprotic solvents (e.g. acetone, acetonitrile,

dimethylformamide or dimthylsulfoxide).

Table 1.1. Gutmann Donor Number values for common organic solvents.

Solvent DN, kcal mol-1

1,2-Dichloroethane (reference solvent) 0.0

Nitromethane 2.7

Acetonitrile 14.1

Propylene carbonate 15.1

Ethylene carbonate 16.4

Acetone 17.0

γ-Butyrolactone 18.0

Diethyl ether 19.2

Tetrahydrofuran 20.0

N,N-Dimethylformamide 26.6

Dimethyl sulfoxide 29.8

Methanol 30.0

Water 33.0

Triethylamine 61.0

1. Introduction

20

Donor Number (DN) is a qualitative measure of Lewis basicity. It is defined

as the negative enthalpy value for 1:1 adduct formation between a Lewis base

and the standard Lewis acid SbCl5, in dilute solution in the non-coordinating

solvent 1,2-dichloroethane with a zero DN. The Donor Number describes the

ability of a solvent to solvate cations and Lewis acids. The method was

developed by V. Gutmann in 1976.18 Values of DN for common organic solvents

are summarized in Table 1.1.

Solvents undergo various weak chemical interactions with the solute. The

most usual interactions are the relatively weak van der Waals interactions

(induced dipole interactions), the stronger dipole-dipole interactions, and even

stronger hydrogen bonds.

The role of a solvent on organometallic substitution reactions cannot be

overestimated; even a small modification of the solvent structure may cause a

large change in the rate and in the pattern of the processes in solution.18

1.4.2. Plasticizers

Battery technology has achieved spectacular progress in recent years. A

most successful product is the rechargeable Lithium Ion Battery (LIB), which

has reached an established commercial status with a production rate of several

millions of units per month. Currently, the lithium salt electrolyte is not held in

an organic solvent like in the past models, but in a solid polymer gel electrolyte

such as polyacrylonitrile, polyvinylidine fluoride, polyethylene oxide, etc.

There are many advantages of this design, for instance, the solid polymer

electrolyte is not flammable as the organic solvent that the Li-ion cell uses.

Thus, these batteries are less hazardous if mistreated.19

The vast majority of electrolytes are the electrolytic solution-types that

consist of salts (also called “electrolyte solutes”) dissolved in the solvents (also

called plasticizers), either water (aqueous) or organic molecules (non-

aqueous), and remain in the liquid state within the service-temperature range.

The most used solvents are ethylene carbonate, propylene carbonate and γ-

butyrolactone. (Figure 1.1)

1. Introduction

21

Figure 1.1. Structural formulas of ethylene carbonate (a), propylene carbonate (b) and γ-butyrolactone (c).

Experimentally it has been found that a mixture of two or more plasticizers

is more convenient, as it allows optimization of the balance between different

features (such as dielectric constant, viscosity, ionic diffusion, salt dissociation,

and chemical stability) and thus to enhance the battery performance and

cyclability.

The overall goal of the project was to study the solvation dynamics and

ionic mobility in conventional solvents and to seek firm knowledge about the

mechanisms of the dynamical processes present, and their relative importance.

The importance of this mechanism can be illustrated with an example from the

polymer field. High molecular-weight poly(ethylene oxide), PEO, with

dissolved ions is a polymeric system with low ionic conductivity at room

temperature. On the other hand, PEO with dissolved ions and a second, low-

molecular weight organic solvent added is an example of a polymeric system

with high ionic conductivity at room temperature. The addition of even modest

amount of such a small molecule to the pure polymer electrolyte drastically

increases the conductivity. The low-molecular weight organic solvent is called

plasticizer.

O O

O

O O

O

O

O

(a) (b) (c)

O O

O

O O

O

O

O

O O

O

O O

O

O

O

(a) (b) (c)

1. Introduction

22

1.5. Cryptands

Whereas coordination chemistry was almost exclusively concerned with the

complexes formed between organic or inorganic ligands and transition metal

ions, only scattered examples of alkali cation (AC) complexes were

documented. The situation changed entirely with the discovery (around the

middle 1960’s in short sequence) of three classes of organic ligands capable of

forming strong and selective complexes with alkali, as well as alkaline-earth,

cations. First, natural macrocyclic antibiotics (in particular valinomycin) were

found to bind and to make membranes permeable to AC’s (namely K+), acting

as ionophores. Then, it was discovered that macrocyclic polyethers, the crown

ethers, formed complexes with AC’s. Thirdly, macrobicyclic receptors, the

cryptands, were designed and shown to form even much more stable AC

cryptate complexes by inclusion into the three-dimensional molecular cavity.

Numerous other ligands were then developed in particular the spherands.

Examples of mycrocyclic ligands are illustrated in Figure 1.2. A rich

coordination chemistry of AC’s has thus been established, displaying a whole

range of complexes of varied structural, thermodynamic and kinetic

properties. Strong and selective complexation of AC’s was achieved, yielding

for instance stability constants higher than 105 l mol-1 in aqueous solution for

the Li+, Na+ and K+ cryptates of the optimal macrobicyclic receptors. Very

stable and selective alkaline-earth complexes of macrocyclic and

macropolycyclic ligands were also obtained.20

1. Introduction

23

Figure 1.2. Structural formulas of crown ethers (a), cryptands (b), and spherands (c).

The chelate effect of acyclic ligands was extended to macrocyclic and

cryptate effects defined by special features (such as substrate inclusion, high

stability, high selectivity, slow exchange rates, shielding from the environment)

presented by macrocyclic and macropolycyclic ligands and their metal ion

complexes.

The field of macrocyclic and macropolycyclic chemical compounds has

greatly advanced over the last years mainly with the goal of developing

complexing-agents, or molecular receptors, capable of binding very efficiently

and very selectively a given substrate. Thus, by taking into account specific

molecular interactions and the relationship between geometrical structure and

binding sites, the chemistry of molecular recognition has been developed, this

being a major part of supramolecular chemistry.21

1. Introduction

24

Compared to the usual macrocycles of the crown type which have a two

dimensional framework, the cryptands are three dimensional in structure

containing an intramolecular cavity (or “crypt”). Cryptands form inclusion

complexes, the cryptates, in which the substrate is located inside the cavity.

Several reviews and monographs which contain important information on

the structures and applications of macrocyclic complexes with alkali metal ions

are available.22-29

Many natural substances have a macrocyclic structure whether or not they

exhibit complexing properties, e.g., cyclic antibiotics, cyclodextrins. Since the

macrocyclic structure is so widespread in nature, one can conclude that it

offers over the linear arrangement certain “useful” properties with respect to

the functions performed. The fundamental observation has been incorporated

in the chemistry of molecular recognition.

The alkali metal cations are spherical species bearing a positive charge. The

recognition of these species involves the selection of a sphere from a collection

of spheres of different radii.

In 1975 Lehn and Sauvage30 reported stability contants and selectivities for

complexation of alkali metal and alkaline earth cations by cryptands. In these

multidentate ligands, two nitrogen atoms were linked by three oxygen atom

containing bridges. Diameters of the alkali metal cations and the cryptand

cavities are recorded in Table 1.2. The cavity size of the cryptand C211 is

appropriate for strong complexation of Li+, but too small to accommodate Na+

or any of the other alkali metal cations. Stability constants for complexation of

Li+, Na+, and K+ by cryptands C222, C221, and C211 are summarized in Table

1.3.29

1. Introduction

25

Table 1.2. Diameters of alkali metal cations and cryptand cavities.

Cation Diameter (Å) Cryptand Diameter (Å)

Li+ 1.48 C211 1.6

Na+ 2.04 C221 2.2

K+ 2.76 C222 2.8

Table 1.3. Stability constants for formation of alkali metal ion complexes with cryptands in water at 25 °C.

log K

Cryptand Li+ Na+ K+

C211 5.5 3.2 < 2

C221 2.5 5.4 4.0

C222 1.0 4.0 5.5

A cation is characterized by its ionic radius, charge and charge density,

polarizability, solvation energy, and also the ionization potential of the

corresponding metal. Cation complexation is governed by coordination

chemistry. The selection of the cation is ensured by a cavity of adequate size,

and the bond to the macrocycle arises through the presence of polar

interaction sites lining the cavity.

The first macrobicyclic diamines were synthesized by Simmons and Park.31

These ligands display a particular form of isomerism since the pair at the

bridgehead nitrogen may be directed either inwards or outwards with respect

to the cavity thus giving rise to three forms: exo-exo or out,out; exo-endo or

out,in; endo-endo or in,in. The predominant form depends on the size of the

bicyclic system.

Inclusion of a Mn+ metal cation in the intramolecular cavity of the three

possible forms of cryptands should give rise, in principle, to the following three

topographies: exo-exo, exo-endo and endo-endo. The endo-endo conformation

1. Introduction

26

is greatly favored because in this case both amine groups participate in

stabilizing the complex.

In this study we emphasize the analogy in structure between ion-polymer

complexes and ion-macrocyclic ligand complexes. It is suggested that the two

types of systems are complementary, in that studies of polymer-salt systems

can give information about interactions in mycrocyclic ligand-salt systems, and

vice versa.32 Crown ethers have been found to improve the electrical

conductivity of solid polymer electrolytes that might be used in lithium

batteries.33 However, there is another way in that the above case makes much

closer contact with the practical world of lithium batteries. Poly(ethylene

oxide), PEO, has been a prototypical polymer used in studies of lithium battery

solid electrolytes. The same –CH2CH2O– moiety that appears in the crown

ethers and cryptands is the building block of PEO. Thus it is hardly surprising

that PEO has a high affinity for alkali metal cations just as crown ethers do.34

1.6. Exchange Processes

The exchange of solvent molecules around a metal ion is the simplest

ligand substitution process and is fundamental to the understanding of the

reactivity of the metal ions in solution. The mechanism of ligand exchange is

often discussed in terms of the scheme developed by Merbach et al.,35 which

emphasizes a structural definition of the different exchange mechanisms, i.e.

the mechanism is defined by the positions of the incoming and leaving ligands

relative to the metal ion, as can be seen in Figure 1.3.

1. Introduction

27

Figure 1.3. Schematic volume profiles for different solvent exchange mechanisms: kP is the observed rate constant at a given pressure P; kP is decelerated or accelerated on increasing pressure for a D or A mechanism, respectively.36

The prevailing exchange mechanism can experimentally be deduced from

the activation parameters ∆S# and ∆V#. Large negative activation entropies or

activation volumes are interpreted as corresponding to an associative

mechanism (A) and large positive values as dissociative (D), whereas small

values are interpreted as concerted mechanisms (Ia, I, or Id).

The pressure dependence of the exchange rate constant leads to the

activation volume, ∆V#, which has become the major tool for the experimental

determination of the exchange mechanisms.37, 38 The volume of activation,

∆V#, is defined as the difference between the partial molar volumes of the

1. Introduction

28

transition state for the reaction and the reactants and is related to the pressure

variation of the rate constant at a constant temperature T by Equation (1.1).

RT

V

dP

kd

T

#ln Δ−=⎟

⎠⎞

⎜⎝⎛

(1.1)

An approximate solution of the above differential Equation (1.1) is Equation

(1.2).

PRTV

kkP

P

00

=

= ⎟⎟⎠

⎞⎜⎜⎝

⎛ Δ−=

#

lnln (1.2)

Solvent and ligand exchange and substitution reactions in liquid solutions

are basic chemical processes that in many cases control the rate and efficiency

of chemical reactions. This research topic is related to various technological

applications both in the polymer electrolyte field and for basic chemical

processes. Solvent dynamics is one of the phenomena which enable ionic

conductivity in a polymeric salt.

Exchange of a solvent molecule between the first and the second

coordination sphere is the simplest reaction that can occur on a metal ion in

aqueous and non-aqueous solution. This reaction is fundamental in

understanding not only the reactivity of metal ions in chemical and biological

systems but also the interaction between the metal ion and the solvent

molecules. The displacement of a solvent molecule from the first coordination

shell represents an important step in complex-formation reactions of metal

cations and in many redox processes.39 The measured rates of solvent

exchange vary extensively with the nature of the metal ion and, to a lesser

extend, with that of the solvent. Metal ions of main group elements have filled

electron shells, and they differ mainly in electric charge and ionic radius. The

number of solvent molecules in the first shell around the ion, commonly called

the coordination number, CN, ranges from 4 up to 10.40 Alkali group ions are

very labile, due to the low surface charge density and the absence of ligand

field stabilization effects.36 Solvent exchange at Li+ in water is extremely fast,

making it difficult to study directly. An exchange rate constant of

1. Introduction

29

approximately 109 s-1 at 25 °C was predicted on the basis of complex-formation

data.36 (Figure 1.4)

Figure 1.4. Mean lifetimes, τH2O, of a particular water molecule in the first coordination shell of a given metal ion and the corresponding water exchange rate kH2O at 298 K. (taken from the recent review36)

1.7. Kinetic Applications of NMR Spectroscopy

The rates of chemical reactions vary over many orders of magnitude. Some

reactions are so slow that they practically do not occur, even though the free

energy would decrease considerably. On the other hand, many chemical

reactions are so fast that until recently they have been called “immeasurably

fast”. During the past 50 years, however, a wealth of techniques have been

developed to study reaction times much shorter than one second and well

outside the range of conventional analytical methods.41

Figure 1.5 shows various experimental methods and their time scales.

1. Introduction

30

Figure 1.5. Various experimental methods and their time scales.

Spectroscopic techniques such as optical spectrophotometry or NMR

spectroscopy have often been used to monitor concentration changes as a

function of time, thereby allowing the kinetic parameters to be determined.

Fast chemical exchange may play a direct role in the optical (or NMR)

absorption process itself, changing the line shape of the observed spectra. This

kinetic modification of line shapes occurs at complete equilibrium, whereas all

other fast reaction techniques require finite deviations from equilibrium.

Kinetic information is available from the resulting line broadening. The older

CW (continuous wave) NMR technique has been superseded by FT (Fourier

transform) NMR.41

Nuclear magnetic resonance (NMR) is one of the most widely used methods

for studying rapid intra- and intermolecular equilibrium processes of

inorganic, organic, and biochemical systems involving the transfer of

1. Introduction

31

components of ions and molecules or the complete entities between different

chemical environments in the liquid state. The only requirement of the method

is that the chemical equilibration process must involve the exchange of nuclear

spins between different magnetic environments, under which circumstances

first order rate constant in the range from 5×10-3 s-1 to 106 s-1 are accessible

using commercially available NMR spectrometers. In the vast majority of

applications, the chemical system studied is at thermal equilibrium and the

chemical kinetic parameters are determined from the modification of the NMR

spectrum induced by the nuclear spin exchange characterizing the chemical

equilibration processes.42

The derivation of the chemical parameters from NMR data requires

knowledge of the factors determining spectral line shapes and the sources of

spectral modifications induced by nuclear spin exchange accompanying

chemical equilibration processes.

Among other spectroscopic techniques, high-pressure NMR especially has

found remarkable applications in the investigation of reaction mechanisms. A

high-pressure NMR probe designed for a 400 MHz narrow bore NMR

spectrometer was developed in our laboratory43 and is presented in Figure

1.6(a) The aim of the proposed design was to allow high resolution, high-

pressure NMR experiments in a narrow bore magnet system without any

modification of the magnet. The probe should be easy and safe to handle. More

details are reported elsewhere.43 The NMR tubes used for high pressure

measurements are presented in Figure 1.6(b)

1. Introduction

32

Figure 1.6. Narrow bore NMR probeheads (a) and NMR high pressure tubes (b).

1.8. Alkali Metal NMR Spectroscopy

The magnitudes of chemical shifts for alkali metal resonances increase

markedly from lithium to caesium, and are a further order of magnitude larger

for thalium. However the sensitivity of the alkali metal chemical shifts to the

environment, specifically the solvation variation, increases in the opposite

direction. Chemical shifts depend both on the nucleus and the nature of the

solvent. They also, for a given alkali metal cation, depend on the nature of the

counterion and the concentration of the salt.44

The properties of 7Li nucleus are quite favorable for NMR studies. (see

Table 1.4) The resonance lines of Li+ ion in solutions are exceptionally narrow

and chemical shift can be measured with considerable accuracy.45

Lithium consists of two stable isotopes: 6Li and 7Li. The magnetogyric ratio,

γ, of 7Li is comparatively high. At a magnetic field strength of 9.4 T (1H: 400

MHz), the correlate Larmor frequency of 7Li is 155.45 MHz. Because of its high

1. Introduction

33

γ number and its high natural abundance, 7Li is a sensitive and easy to detect

nucleus. However, many NMR studies on organolithium compounds are now

frequently carried out by observing the 6Li nucleus, preferably in 6Li isotope-

enriched samples.46

Table 1.4. Nuclear properties of key isotopes of lithium.46, 47

Isotope 6Li 7Li

Atomic mass 6.01512 7.01600

Natural abundance (%) 7.42 92.58

Nuclear spin 1 3/2

υo at 9.4 T (1H = 40 MHz) 58.862 155.454

Nuclear magnetic moment μ +0.82203 +3.25636

Quadrupole magnetic

moment Q (10-28 m2) -8 × 10-4 -4.5 × 10-2

Receptivity Dc (13C = 1.00) 3.58 1540

Width factor (7Li = 1.0) 2.0 × 10-3 1.00

Selective solvation of alkali metal ions in mixed solvents can be monitored

from the extent to which the chemical shift of the alkali metal nucleus varies

with solvent composition.

The linewidth measurements may also give some idea of solvation in mixed

solvents. The dependence of the resonance linewidths on solvent composition

is investigated in the mixtures containing one good and one poor solvent. The

difference in pattern is striking and is attributed to relatively long-lived cation-

solvent complexes formed by the strongly coordinating solvents. Such a

difference in behavior is reasonable. It is consistent with solvating abilities of

the solvents, and with their respective Gutmann donor numbers.

1. Introduction

34

The exchange of ions between different environments is usually rapid with

respect to the NMR time scale, resulting in only one resonance signal at an

average frequency determined by the magnetic shielding and lifetime of the

nucleus in each of the sites. Alteration of parameters such as concentration,

counterions, and solvent produces changes in the relative proportion and type

of environment which may be reflected by alteration of the NMR spectra on

chemical shift and/or line shape, and/or line width of the observed

resonance.45

1.9. Programs

The line widths of the NMR signals can be obtained by fitting Lorentzian

functions to the experimental spectra using the NMRICMA 2.7 program.48

NMRICMA is a program written for MATLAB to evaluate 1D NMR and EPR

spectra. It reads experimental NMR spectra treated with 1D WIN-NMR from

Bruker. NMRICMA can fit Lorentzian resonance lines to experimental spectra,

fit NMR multiplets (formed by Lorentzian lines) to experimental spectra, fit

exchange broadened NMR spectra using Kubo-Sack formalism and simulate

exchange broadened NMR spectra.

Complete line-shape analysis based on the Kubo-Sack formalism using

modified Bloch equations49 was performed with NMRICMA 2.7 to extract rate

constants from experimental spectra. In order to elucidate the exchange rate

constant between two species, the exchange matrix must be created. In case of

a two site exchange, a 2 × 2 exchange matrix was created.

⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢

⎣

⎡

−

−

=

ABB

AAB

B

A

ABAB

kp

pk

p

p

kk

D

1. Introduction

35

where: kAB refers to the rate constant between species A and B,

pA and pA represent the populations of the NMR signals referring

to the species A and B, respectively.

The exchange matrix is created and saved as the matex file in Matlab:

function m = matex2(p,k) %---------------------------------------------------------------------------- % funtion subroutine for exchange %---------------------------------------------------------------------------- % % alpha - version; started on january 1994 % © lothar helm, icma université lausanne % % rate constant(s) = k % populations = p % % global xtitle matexfile si td sw sf limits nlimits; nsites = length(p); m = zeros(nsites); m(1,1) = -k(1); m(1,2) = +k(1); m(2,1) = p(1)/p(2)*k(1); m(2,2) = -m(2,1);

The adjustable parameters are: resonance frequency, intensity

(population), line width and baseline. The temperature and pressure

dependent 7Li line widths and chemical shifts employed in the line shape

analysis were extrapolated from low temperatures where no exchange-induced

modification occurred.

The fitting method used throughout was the Lavenberg-Marquard method.

The progress of fitting can be followed by observing the MATLAB window. The

iteration stops when the maximum number of iterations is reached. The

spectrum plot window of NMRICMA shows the result of the fit: experimental

and fitted spectra on an overlay plot, difference between fitted and

experimental spectrum, and fitting results are given as well as some other

1. Introduction

36

useful information. One example of the output spectrum can be seen in Figure

1.7.

Figure 1.7. The output 7Li NMR spectrum from the NMRICMA fitting program.

1.10. Computational Methods

The theoretical calculations employed in this work were mostly based on

quantum chemical methods. They were based either on the wave function or

on the electron density of the investigated molecule to calculate its structure

and properties. The significance of these two methods can be recognized from

two Nobel Prize Awards given in 1998 to J. A. Pople for the development of the

computational methods in quantum chemistry and to W. Kohn for the

development of the density functional theory.

1. Introduction

37

The quantum chemical methods based on the wave function come from the

Schrödinger equation and describe all relevant electronic, time dependent and

relativistic effects. For the purpose of many chemical problems one can often

neglect the time dependent and relativistic effects to obtain in this way the

time-independent, non-relativistic Schrödinger equation. The fundamental

assumption (known as the Born-Oppenheimer approximation) is that, in the

adiabatic limit, we can decouple the nuclear and the electronic degrees of

freedom; in this way the time-independent electronic Schrödinger equation is

obtained.

1.10.1. Ab-initio Methods

In order to solve the electronic Schrödinger equation, an additional

approximation is introduced: Hartree-Fock (HF) approximation.

The wave function which describes both the single electron spatial

distribution and its spin is called spin orbital. The fundamental idea of the HF-

theory is that the N-electrons wave function (N is the number of electrons) is

displaced by more single electron wave functions. To a first approximation we

can consider that the electrons do not interact. The total Hamiltonian will be

the sum of N individual Hamiltonians. The total wave function is the product

of N spin orbitals. Since electrons in fact interact with each other, in the

Hartree-Fock approximation, the Coulomb interaction among electrons is

averaged and the many electrons wave function (and energies) is obtained in

an iterative way. In this model one electron is influenced by the presence of

Quantum Chemical Calculations

electron density based methodswave function based methods

density functional theory (DFT)

e.g. B3LYP, BLYP, PBE, BP86…

Ab-initio methods

e.g. HF, MP2, MP4…

semi empirical methods

Quantum Chemical Calculations

electron density based methodswave function based methods

density functional theory (DFT)

e.g. B3LYP, BLYP, PBE, BP86…

Ab-initio methods

e.g. HF, MP2, MP4…

semi empirical methods

1. Introduction

38

other electrons. For practical purposes, the unknown spatial molecular orbitals

are often transformed into one set of known spatial functions, viz. the basis set.

In practice the Hartree-Fock equation is solved by introducing a finite set of K

basis functions, which lead to a set of N occupied spin orbitals and K-N virtual

ones.

Since there is no correlation in the electron movement in the HF-

approximation, this must be considered by means of some other methods. A

different procedure to find the correlation energy, which is not variational but

size consistent at each level, is the Perturbation Theory (PT). In 1934 Møller

and Plesset proposed a perturbation treatment of atoms and molecules in

which the unperturbated wave function is the Hartree-Fock function; this is

the so called Møller-Plesset (MP) perturbation theory (general: many body

perturbation theory). It is based on Rayleigh-Schrödinger perturbation theory

where the total Hamiltonian of the system is divided into two parts: a zero-

order part which has known eigenvalues and eigenfunctions, and a

perturbation operator. Second (MP2) and fourth (MP4) order Møller-Plesset

calculations (where we refer to the order of the energies) are standard levels

used for small systems and are implemented in many computational chemistry

codes.

1.10.2. Density Functional Theory

Hohenberg and Kohn in 1964 suggested that the many-electron wave

function was a too complicated entity to deal with as the fundamental variable

in a variational approach. Firstly, it cannot adequately be described without ~

1023 parameters, and secondly it has the complication of possessing a phase as

well as a magnitude. They chose instead to use the electron density as their

fundamental variable.

The advent of modern DFT methods for molecular systems has

revolutionised computational chemistry, especially for the transition metals.50

In a nutshell, DFT is based on the fact that all molecular electronic properties

can be calculated if the electron density is known. The only blemish is that the

exact form of all the components of the functional that gives the energy from

1. Introduction

39

the electron density is unknown. There are, however, a number of approximate

functionals available. These can be divided into three classes:

1. Functionals based on the local density approximation (LDA) only depend

on the value of the electron density at any given point in space.

2. Functionals that use not only the value of the electron density, but also

its gradient (how fast it changes in space) are based on the generalised

gradient approximation (GGA).

3. Hybrid density functional techniques combine GGA functionals with a

parameterised proportion of the exchange energy calculated by Hartree–Fock

(HF) theory. They are thus a cross between DFT and conventional HF ab initio

theory. However, hybrid DFT methods are the most accurate of the three

classes and, therefore, enjoy great popularity. The best-known hybrid level of

theory is Becke's three-parameter technique with the Lee–Yang–Parr

correlation functional (B3LYP).51, 52

The two major advantages of DFT calculations are that they give good to

excellent results for metal systems, including transition metals, and that they

scale relatively well with the size of the system to be calculated, so that large

complexes and clusters can be treated. However, DFT shows some serious

limitations for studying water-exchange reactions. Hydrogen-bonding and

proton-transfer reactions are generally not treated well by DFT.53 Compared

with MP2, even hybrid DFT with an adequate basis set also favors lower

coordination numbers.54

1. Introduction

40

1.11. References

1. Sapse, A.-M.; Schleyer, P. V. R.; Editors Lithium Chemistry: A

Theoretical and Experimental Overview; John Wiley & Sons, Inc.: New

York, 1995; p 595.

2. Deberitz, J.; Boche, G. Chem. unserer Zeit 2003, 37, 258.

3. Snaith, R.; Wright, D. S. In Lithium Chemistry; Sapse, A.-M.; Schleyer,

P. V. R., Eds.; John Wiley & Sons, Inc.: New York, 1995; pp 227.

4. Rosenthal, N. E.; Goodwin, F. K. Annu. Rev. Med. 1982, 33, 555.

5. Wood, A. J.; Goodwin, G. M. Psychol. Med. 1987, 17, 579.

6. Soares, J. C.; Gershon, S. Neuropsychopharmacol. 1998, 19, 167.

7. Bhagwagar, Z.; Goodwin, G. M. Clin. Neurosci. Res. 2002, 2, 222.

8. Fountoulakis, K. N.; Vieta, E.; Sanchez-Moreno, J.; Kaprinis, S. G.;

Goikolea, J. M.; Kaprinis, G. S. J. Affect. Disorders 2005, 86, 1.

9. Shaw, M. The role of lithium clinics in the treatment of bipolar disorder;

South West Yorkshire Mental Health NHS Trust: England: United

Kingdom, 2004; pp 42.

10. Salzman, C. Harvard review of psychiatry 2003, 11, 230.

11. Cowen, P. J. Br. Med. Bull. 1996, 52, 539.

12. Goodwin Guy, M.; Young Allan, H. J. Psychopharmacol. 2003, 17, 3.

13. Thase, M. E.; Sachs, G. S. Biol. Psychiat. 2000, 48, 558.

14. Fetter Jeffrey, C.; Askland Kathleen, D. Am. J. Psychiatry 2005, 162,

1546; author reply 1547.

15. Gijsman Harm, J.; Geddes John, R.; Rendell Jennifer, M.; Nolen

Willem, A.; Goodwin Guy, M. Am. J. Psychiatry 2004, 161, 1537.

16. Geddes John, R.; Burgess, S.; Hawton, K.; Jamison, K.; Goodwin Guy,

M. Am. J. Psychiatry 2004, 161, 217.

17. Spångberg, D., Ph.D. Thesis, Uppsala University, Uppsala, 2003.

18. Gutmann, V. Electrochim. Acta 1976, 21, 661.

19. Masia, M., Ph.D. Thesis, Universitat Politècnica de Catalunya,

Barcelona, 2005.

1. Introduction

41

20. Lehn, J. M. In Perspect. Coord. Chem.; Williams, A. F.; Floriani, C.;

Merbach, A. E., Eds.; VCH Verlagsgesellschaft/Verlag Helvetica Chimica

Acta: Weinheim/Basel, 1992; pp 447.

21. Dietrich, B.; Viout, P.; Lehn, J. M. Macrocyclic Chemistry: Aspects of

Organic and Inorganic Supra Molecular Chemistry; 1993; p 384.

22. Hubberstey, P. Coord. Chem. Rev. 1985, 66, 1.

23. Izatt, R. M.; Bradshaw, J. S.; Nielsen, S. A.; Lamb, J. D.; Christensen, J.

J.; Sen, D. Chem. Rev. 1985, 85, 271.

24. Hubberstey, P. Coord. Chem. Rev. 1988, 85, 1.

25. Bajaj, A. V.; Poonia, N. S. Coord. Chem. Rev. 1988, 87, 55.

26. Olsher, U.; Izatt, R. M.; Bradshaw, J. S.; Dalley, N. K. Chem. Rev. 1991,

91, 137.

27. Izatt, R. M.; Pawlak, K.; Bradshaw, J. S.; Bruening, R. L. Chem. Rev.

1991, 91, 1721.

28. Izatt, R. M.; Pawlak, K.; Bradshaw, J. S.; Bruening, R. L. Chemical

Reviews 1995, 95, 2529.

29. Bartsch, R. A.; Ramesh, V.; Bach, R. O.; Shono, T.; Kimura, K. In

Lithium Chemistry; Sapse, A.-M.; Schleyer, P. V. R., Eds.; John Wiley &

Sons, Inc.: New York, 1995; pp 393.

30. Lehn, J. M.; Sauvage, J. P. J. Am. Chem. Soc. 1975, 97, 6700.

31. Park, C. H.; Simmons, H. E. J. Am. Chem. Soc. 1968, 90, 2429.

32. Lorimer, J. W. Pure Appl. Chem. 1993, 65, 1499.

33. Kaplan, M. L.; Reitman, E. A.; Cava, R. J. Polymer 1989, 30, 504.

34. Eyring, E. M.; Petrucci, S. In Cation Binding Macrocycles; Inoue, Y,;

Gokel, G. W., Eds.; Marcel Dekker, Inc.: New York, 1990, pp 179.

35. Helm, L.; Merbach, A. E. Coord. Chem. Rev. 1999, 187, 151.

36. Helm, L.; Merbach, A. E. Chem. Rev. 2005, 105, 1923.

37. Van Eldik, R. Studies in Inorganic Chemistry 1986, 7, 1.

38. van Eldik, R.; Klarner, F.-G. High Pressure Chemistry: Synthetic,

Mechanistic, and Supercritical Applications; 2003; p 458.

39. Richens, D. T. The Chemistry of Aqua Ions; Wiley&Sons: Chichester,

1997.

1. Introduction

42

40. Radnai, T. J. Mol. Liq. 1995, 65/66, 229.

41. Strehlow, H.; Editor Rapid Reactions in Solution; VCH: Weinheim,

1992; p 341.

42. Lincoln, S. F. Prog. React. Kinet. 1977, 9, 91.

43. Zahl, A.; Igel, P.; Weller, M.; van Eldik, R. Rev. Sci. Instrum. 2004, 75,

3152.

44. Burgess, J. Metal Ions in Solution; Ellis Horwood: Chichester, 1978; p

464.

45. Cahen, Y. M.; Handy, P. R.; Roach, E. T.; Popov, A. I. J. Phys. Chem.

1975, 79, 80.

46. Bauer, W. In Lithium Chemistry; Sapse, A.-M.; Schleyer, P. V. R., Eds.;

John Wiley & Sons, Inc.: New York, 1995; pp 125.

47. Emsley, J. The Oxford Book of the Elements; Oxford University Press:

Oxford, 2001; p 448.

48. Helm, L.; Borel, A. NMRICMA 2.7, University of Lausanne, 2000.

49. Pople, J. A.; Schneider, W. G.; Bernstein, H. J. High-Resolution Nuclear

Magnetic Resonance; McGrow-Hill Book Co.: New York, 1959; p 513.

50. Koch, W.; Holthausen, M. C. A Chemist's Guide to Density Functional

Theory, 2nd Edition; John Wiley & Sons, Inc.: Chichester, UK, 2001; p

528.

51. Becke, A. D. J. Chem. Phys. 1993, 98, 5648.

52. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter Mater.

Phys. 1988, 37, 785.

53. Pavese, M.; Chawla, S.; Lu, D.; Lobaugh, J.; Voth, G. A. J. Chem. Phys.

1997, 107, 7428.

54. Diaz, N.; Suarez, D.; Merz, K. M., Jr. Chem. Phys. Lett. 2000, 326, 288.

2. DMSO and Water/DMSO Mixtures

43

2. Ligand Exchange Processes on Solvated

Lithium Cations. DMSO and Water/DMSO

Mixtures

2.1. General remark

The work presented in this chapter was conducted by several authors within a

joint project and has been published as:

Pasgreta, E.; Puchta, R.; Galle, M.; van Eikema-Hommes N.; Zahl A.; van Eldik

R. ChemPhysChem 2007, 8, 1315-1320.

2.2. Abstract

Solutions of LiClO4 in solvent mixtures consisting of dimethylsulfoxide

(DMSO) and water, or DMSO and γ-butyrolactone, were studied by 7Li NMR

spectroscopy. Chemical shifts indicate that the Li+ ion is coordinated by four

DMSO molecules. In the binary solvent mixture of water and DMSO, no

selective solvation is detected, thus indicating that on increasing the water

content of the solvent mixture, DMSO is gradually displaced by water in the

coordination sphere of Li+. The ligand-exchange mechanism of Li+ ions

solvated by DMSO and water/DMSO mixtures was studied using DFT

calculations. Ligand exchange on [Li(DMSO)4]+ was found to follow a limiting

associative (A) mechanism. The displacement of coordinated H2O by DMSO in

[Li(H2O)4]+ follows an associative interchange mechanism. The suggested

mechanisms are discussed in reference to available experimental and

theoretical data.


44

2.3. Introduction

Lithium salts are widely used in industrial applications, synthesis, and

medicine, for example, in psychiatry.1 Therefore, detailed understanding of

fundamental reactions of solvated lithium cations, for example ligand-

exchange reactions in different solvents, is essential.

Knowledge of elementary reactions, such as solvent exchange, in different

solvents is an important prerequisite for selecting, optimizing and tailoring the

ideal solvent and solvent mixture. Whereas experimental methods investigate

reactions under real conditions in a whole ensemble of molecules, quantum-

chemical calculations can focus on a single molecule in the presence of a small

number of solvent molecules, for example, those of the first solvation shell.2-7

This complementary approach allows further mechanistic insight to be gained.

Dimethylsulfoxide (DMSO) is a common solvent in chemical laboratories,8-

11 even though industry tries to avoid it. It is widely used in laboratories due to

its ability to dissolve a wide range of chemical compounds (from ionic

inorganic salts to purely covalent compounds), because it is an aprotic but

highly polar polyfunctional molecule with both a hydrophobic and a

hydrophilic site. Moreover, it has the highest dipole moment (4.3 Debye12) and

highest dielectric constant (48 at 20 °C) of any dipolar aprotic liquid.

Furthermore, DMSO is also used as a valuable reactant for example in a

number of oxidation reactions,13, 14 like the Swern oxidation of alcohols to

aldehydes and ketones using activated DMSO,15, 16 or in a Pummerer-type

reaction17, 18 to obtain the Tröger’s bases.19

The use of DMSO in human and veterinary medicine is based on its ability

to penetrate deeply through skin and other membranes without damaging

them, while carrying other compounds deep into a biological system.20-24 On

the other hand, this effect causes the biggest toxicological concerns.25-28

DMSO possesses very strong donating abilities. This property has also been

confirmed by various studies of stability constants of alkali-metal complexes.29

It is well-known that the solvating ability of the solvent, as expressed by the


45

Guttmann donor number (DN)30 plays a fundamental role in different

complexation reactions. DMSO is a solvent of high solvating ability (DN =

29.8) and it can strongly compete with ligands such as crown ethers or

cryptands for alkali-metal ion coordination.29

A variety of theoretical and experimental studies on liquid DMSO,31-41

water/DMSO mixtures,42-47 DMSO/nonaqueous electrolyte solutions32, 48-51 and

solutions of biomolecules in DMSO52-54 have been performed in recent years to

understand the physical properties of DMSO solutions. Solutions of alkali-

metal ions, mostly Li+, in DMSO have also been investigated experimentally55-

60 and theoretically,32, 61-69 as well as in water/DMSO mixtures70-73 and in other

solvent combinations.74-76 One of the first reports on the utilization of 7Li NMR

in the investigation of LiClO4 in various solvents (including DMSO) was

published by Maciel et al..77 In DMSO, no evidence for contact ion pairing was

observed,73, 78 as might be expected, for instance, in THF (DN = 7.4).79 It was

also shown that Li+ interacts preferentially with DMSO electrostatically

through the oxygen donor of the S=O group.76

The special interest in Li+ ions in DMSO and DMSO mixtures with other

solvents originates from the use of Li+ ions in batteries with high energy

densities.80-83 Such studies on the transport properties of electrolytes in

aqueous and nonaqueous solvents are of interest in various technologies, such

as high-energy-density batteries, photo-electrochemical cells,

electrodeposition, wet electrolytic capacitors and in electro-organic

synthesis.82 Electrochemical technologies search for electrolyte solutions with

optimized properties for the special tasks which they must fulfill in devices and

processes.83 A literature survey suggests that the use of mixed solvents

involving high permittivity, low viscosity and wide liquid range is practicable

for use in high-energy-density batteries, especially lithium batteries. The

advantage of using mixed solvents over using their pure counterparts in

lithium batteries is to be able to manipulate the suitable properties of these

solvents for fulfilling the above conditions.


46

While many studies explored the behaviour of Li+ in DMSO and

water/DMSO mixtures, to our knowledge, no attempts have been made to

investigate the ligand-exchange mechanism of solvated Li+ in DMSO and

water/DMSO mixtures. Herein, we extend our combined quantum-chemical

and experimental investigations of the ligand-exchange mechanism around

Li+7 and focus on DMSO and water/DMSO mixtures.

2.4. Experimental Section

2.4.1. General Procedures

All transformations were carried out in an Ar or N2 atmosphere using

standard Schlenk techniques, in an Ar atmosphere in a drybox, or under

vacuum. Lithium perchlorate (Aldrich, battery grade) was vacuum-dried at 120

°C for 48 h and then stored under dry nitrogen. DMSO (Acros, extra dry, with

molecular sieves, water < 50 ppm) was used as received. γ-Butyrolactone

(Aldrich) was dried over anhydrous CaSO4, then fractionally distilled under

vacuum and stored under nitrogen over Linde 3Å molecular sieves. Solutions

of LiClO4 were prepared under dry nitrogen. For 7Li NMR studies, these

solutions were sealed in Ar atmosphere in 5-mm NMR tubes.

2.4.2. 7Li NMR Spectroscopy

Fourier-transform 7Li NMR spectra were recorded at 155 MHz on a Bruker

Avance DRX 400WB spectrometer equipped with a superconducting BC-

94/89 magnet system. Spinning 5-mm tubes were used with a 1-mm outer

diameter melting-point capillary inserted coaxially in the spinning tube and

filled with an external reference solution (usually 1 M LiClO4 in DMF). The

experiments were conducted at room temperature and at ambient pressure.

For a typical kinetic run, 0.8 mL of solution was transferred under an Ar to a

Wilmad screw-cup NMR tube equipped with a poly(tetrafluoroethylene)

septum. All samples were prepared in the same fashion.


47

2.4.3. Quantum-Chemical Calculations

All structures were fully optimized using the B3LYP hybrid density

functional84-86 and the LANL2DZp basis set, that is, with pseudo-potentials on

sulfur and the valence basis set augmented with polarization functions,87-91 and

characterized as minima or transition-state structures by computation of

vibrational frequencies (for minima, all frequencies are positive, NImag = 0;

for transition structures, exactly one imaginary frequency is present, NImag =

1). Structures were subsequently refined at B3LYP/6-311+G**,84-86, 92 that is,

with diffuse functions on non-hydrogen atoms, in order to minimise the basis

set superposition error (BSSE).93 Zero-point energies and contributions to

enthalpy and entropy were computed for a temperature of 298 K at

B3LYP/LANL2DZp or B3LYP/6-311+G**, as indicated. The influence of the

bulk solvent was evaluated via single-point calculations using the CPCM

(conductor-like polarizable continuum model)94, 95 formalism, that is,

B3LYP(CPCM)/6-311+G**//B3LYP/6-311+G** and water as solvent. Structure

optimizations at this level, probed for selected system, were found to yield

essentially the same results. The approximation of solvent effects by the

polarizable continuum model (PCM)96 or the isodensity molecular surface

polarizable continuum model (IPCM)97 unfortunately failed for some systems,

[see 98, 99] but where successful, they agreed with CPCM results. The Gaussian

03 suite of programs was used throughout.100

2.5. Results and Discussion

2.5.1. Coordination Number of Li+ in DMSO

To determine the coordination number of Li+ in DMSO, the cation should

be dissolved in a solvent mixture that consists of a weakly and strongly

coordinating component. The method is based on monitoring the resonance

frequency of 7Li nucleus as a function of DMSO-to-metal ion molar ratio while

keeping the metal ion concentration constant and varying the concentration of

DMSO over a wide range. This kind of study might be hampered by limited

solubilities of Li salts in solvent mixtures or limited miscibility of mixture


48

components. There are a few conditions that must be obeyed: 1) A weakly

coordinating solvent must be available, which will dissolve all solution

components. 2) A measurable parameter must exist which is a function of

solvent composition. 3) The formation constant of the solvate must be

sufficiently large that a limiting value of the parameter may be obtained. In

this case, the parameter was the Li+ NMR signal. Its position in the NMR

spectrum depends on the solvation shell of the solvated ion. Typical alkali-

metal salts are not appreciably soluble in truly non-coordinating solvents.

Therefore, it was necessary to select a solvent with a weak interaction toward

DMSO and the alkali-metal ions.101

γ-Butyrolactone was found to represent a good compromise for the various

requirements. Contrary to DMSO, it is a rather weak donor (DN = 18.0)102 that

appreciably dissolves alkali-metal salts, and its interaction with DMSO is much

weaker than that between DMSO and a metal cation. The molar ratio data

confirmed the “inertness” of the solvent and indicated that it is in fact

sufficiently unreactive for the purpose of this investigation.

Figure 2.1. 7Li NMR chemical-shift variation as a function of the DMSO/LiClO4 molar

ratio in γ-butyrolactone solution.


49

Figure 2.1 shows the molar-ratio plot of the DMSO/LiClO4 system in γ-

butyrolactone solution, where the LiClO4 concentration was kept constant at

0.05 M. A clear break can be observed at the DMSO/Li+ ratio of 4:1, indicating

a solvation number of 4 for Li+ in DMSO.

2.5.2. Selective Solvation of Li+

Selective solvation of Li+ in H2O/DMSO mixture was studied by Emons et

al.,73 using Raman spectroscopy. Their results suggest that Li+-ion

coordination by DMSO is stronger than by water. On the other hand, the

Gutmann donor number for water (3330) is higher than for DMSO (29.8),

which would suggest the opposite.

We performed 7Li NMR measurements keeping the concentration of LiClO4

constant (0.1 M) and varying the composition of the mixture of H2O/DMSO by

increasing the percentage of H2O from 0 to 100 % (molar ratio). For each

sample, the 7Li NMR spectrum was recorded. The chemical shift varied

gradually from 4.35 ppm (DMSO) to 5.07 ppm (water; see Figure 2.2), and the

plot of chemical shift versus solvent composition (Figure 2.3) shows no clear

inflection like that observed for the DMSO/γ-butyrolactone system in Figure

2.1. Thus, the composition of the first coordination sphere around Li+ in

mixtures of H2O and DMSO depends only on the ratio of these solvents. DMSO

and water are equally strong ligands in the coordination of Li+.


50

Figure 2.2. 7Li NMR spectra of LiClO4 in H2O/DMSO for different mixture compositions.

Figure 2.3. Chemical-shift variation as a function of water content in LiClO4 solutions of

H2O/DMSO mixtures.


51

Model calculations (B3LYP(CPCM)/6-311+G**//B3LYP/6-311+G**)

corroborate these results (Table 2.1). The exchange reactions in the stepwise

conversion of [Li(H2O)4]+ + 4 DMSO to [Li(DMSO)4]+ + 4 water are computed

to be essentially thermoneutral. Only for the exchange of the last water

molecule by DMSO is the computed reaction enthalpy somewhat larger (–3.4

kcal/mol).

Table 2.1. Calculated energies for the stepwise conversion of [Li(H2O)4]+ to [Li(DMSO)4]+.[a]

ΔEDFT [b] ΔHCPCM [c] ΔGCPCM [c]

1 [Li(H2O)4]+ + 4 DMSO 0.0 0.0 0.0

2 [Li(DMSO)(H2O)3]+ + H2O + 3 DMSO –13.3 –0.5 2.6

3 [Li(DMSO)2(H2O)2]+ + 2 H2O + 2 DMSO –24.6 0.8 3.6

4 [Li(DMSO)3(H2O)]+ + 3 H2O + DMSO –33.4 –0.8 6.0

5 [Li(DMSO)4]+ + 4 H2O –42.6 –4.2 4.7

[a] Energies in kcal/mol relative to [Li(H2O)4]+ and 4 DMSO; [b] B3LYP/6-311+G**, including ∆ZPE computed at B3LYP/6-311+G**; [c] B3LYP(CPCM)/6 311+G**//B3LYP/6-311+G** with thermodynamic contributions calculated at B3LYP/6-311+G**

The Gibbs free energy values might be interpreted as predicting the

exchange reactions to be endergonic. However, due to inherent inaccuracies,

the error interval of the computed entropy contributions is too large to draw

any conclusions here. Note also the importance of including the effect of the

bulk solvent: in the gas-phase calculations, all solvent-exchange steps are

significantly exothermic.

Calculations at the B3LYP(IPCM)/6-311+G**//B3LYP/6-311+G** level,

that is, employing the isodensity surface polarizable continuum model, were

successful only for [Li(H2O)4]+ and [Li(DMSO)4]+ (Table 2.1, entries 1 and

5). However, the computed reaction enthalpy for the complete solvent

exchange, –5.3 kcal/mol, agrees well with the CPCM result.


52

2.5.3. Ligand Exchange on Solvated Li+

Exchange of a solvent molecule between the first and the second

coordination shell is the simplest reaction that can occur on a metal ion in

solution. This reaction is of fundamental importance in understanding not

only the reactivity of the metal ion in chemical and biological systems but also

the interaction between the metal ion and the solvent molecule. The

displacement of a solvent molecule from the first coordination shell presents

an important step in complex-formation reactions of metal cations and in

many redox processes.103 Mechanistically, associative (A), interchange (I) and

dissociative (D) pathways are conceivable. Solvent exchange on Li+ in water is

extremely fast, making it difficult to study directly. An exchange rate constant

of approximately 109 s-1 at 25 °C was predicted on the basis of complex-

formation data.104 As shown above, DMSO coordinates to Li+ with a similar

strength as water. Its exchange process is also very fast and therefore

impossible to be studied experimentally. All alkali-metal ions and the large

alkaline-earth-metal ions are very labile due to the low surface charge density

and the absence of ligand field stabilization effects.104

2.5.4. Computed Mechanisms of Solvent Exchange

Experimental (vide supra) and previous theoretical62 studies agree with the

present study that the first coordination shell around Li+ consists of four

tightly bound DMSO molecules. The computed Li-O distance is 1.96 Å. An

additional DMSO molecule does not coordinate to the lithium cation but

instead is only weakly bound, mainly by electrostatic interactions. The Li…O

distance is well over 5 Å (see Figure 2.4), and the computed gas-phase binding

energy for the fifth DMSO molecule in [Li(DMSO)4(DMSO)]+ is 7.6 kcal/mol.

When the influence of the bulk solvent is taken into account, at

B3LYP(CPCM)/6-311+G**//B3LYP/6-311+G**, this binding energy

(computed for the gas-phase structure) vanishes. Optimization of the

precursor complex at B3LYP(CPCM)/6-311+G** resulted in only minimal

changes in structure and energy.


53

Figure 2.4. Reaction profile for DMSO exchange on solvated Li+; enthalpy: B3LYP(CPCM)/6-311+G** including thermodynamic contributions at B3LYP/LANL2DZp; gas-phase energy values: B3LYP/6-311+G** including ΔZPE at B3LYP/LANL2DZp, in parentheses. T.S. = transition state.

In a previous study, we found a limiting associative (A) mechanism for

water exchange around Li+.7 Mechanistic details are very similar for DMSO

exchange. The reaction proceeds through a distorted trigonal-bipyramidal

reactive intermediate [Li(DMSO)5]+ that is reached via a late transition state.

The enthalpy profile (Figure 2.4, B3LYP(CPCM)/6-311+G**, thermodynamic

contributions computed at B3LYP/LANL2DZp) is in line with the experimental

observation (vide supra) of a very fast exchange process: the five-coordinate

intermediate is computed to be 7.9 kcal/mol less stable than [Li(DMSO)4]+ and

free DMSO, while an overall activation barrier of only 8.4 kcal/mol is

computed. Obviously, the intermediate will be extremely short-lived, since its

barrier toward loss of DMSO is only 0.5 kcal/mol. Note that with the obvious

exception of the binding energy for the fifth DMSO molecule in the precursor

complex (vide supra), the influence of the bulk solvent on the energy profile is

minor.


54

One might argue that this energy difference is smaller than the accuracy of

the computational method and thus meaningless. However, whereas the value

of the (small) barrier may indeed vary, the shape of the potential-energy

surface usually does not (see also ref.7). In the present case, zero-point energy

and thermodynamic corrections do not change the energetic order of the

structures. Inclusion of entropy contributions (we believe these to be

sufficiently accurate, in view of the similarity of the structures calculated for

intermediate and transition states) even stabilises the intermediate: a free

energy barrier of 1.8 kcal/mol is computed. We therefore consider the

mechanistic conclusion of associative (A) exchange to be reliable.

We also investigated in detail the mechanism of the exchange of a water

ligand by DMSO. The enthalpy profile, calculated at B3LYP(CPCM)/6-311+G**

with thermodynamic contributions computed at B3LYP/6-311+G**, is shown

in Figure 2.5. B3LYP/6-311+G** gas-phase energy values, corrected for ∆ZPE

at this level, are included for comparison.

Figure 2.5. Reaction profile for the water/DMSO exchange around [Li(H2O)4]+; enthalpy: B3LYP(CPCM)/6-311+G** including thermodynamic contributions at B3LYP/6-311+G**; gas-phase energy values: B3LYP/6-311+G** including ΔZPE, in parentheses. T.S. = transition state.


55

The reaction involves a precursor complex, in which the incoming DMSO

molecule is weakly bound to two water ligands through hydrogen bonds. As in

the case of water exchange7 and DMSO exchange (vide supra), the gas-phase

binding energy for this complex is exaggerated. Otherwise, the influence of the

bulk solvent on the energy profile is minor.

The two H2O ligands with hydrogen bonds to the DMSO oxygen move to

axial positions while DMSO approaches the lithium cation (Figure 2.5). In the

distorted trigonal-bipyramidal transition structure, one hydrogen bond to the

leaving water molecule remains. Note that the reaction proceeds with cis

stereochemistry. The activation barrier is 6.9 kcal/mol, relative to the

precursor complex, or 3.3 kcal/mol relative to separated [Li(H2O)4]+ and

DMSO. The computed normal mode vector for the imaginary frequency clearly

corresponds to a ligand substitution: the Li–Owater bond length increases upon

shortening of the Li–ODMSO bond (2.29 Å; 0.4 Å longer than in the product).

The Li–Owater bond is elongated by only 0.16 Å in the transition structure. We

conclude that the reaction will be very fast (cf. ref.105) and proceeds through an

associative interchange mechanism. No stable five-coordinate intermediate

structure could be found, despite extensive searches.

The reaction proceeds to a product complex in which the leaving water

molecule has hydrogen bonds to two water ligands and to DMSO; this

arrangement is 3.6 kcal/mol less stable than the precursor complex. Breaking

the bond to DMSO leads to a second, somewhat more stable complex, from

which the water molecule is released.

2.6. Conclusions

It was confirmed both experimentally and theoretically that Li+ has a

coordination number of four in pure DMSO and in water/DMSO mixtures. In

the binary mixture of water and DMSO, both solvents coordinate equally

strongly by the Li+ ion. Displacement of DMSO by water in the coordination

sphere of solvated Li+ occurs gradually, indicating that the two compounds

have similar donor strengths. In the solvent mixture consisting of DMSO and


56

γ-butyrolactone, the latter component is weakly coordinating and can be

considered as a diluent for DMSO and alkali-metal ions. Computational

studies suggest that ligand exchange on [Li(DMSO)4]+ follows a limiting

associative mechanism, whereas displacement of coordinated water by DMSO

seems to follow an associative interchange mechanism.


57

2.7. References

1. Birch, N. J. Chem. Rev. 1999, 99, 2659.

2. Erras-Hanauer, H.; Clark, T.; van Eldik, R. Coord. Chem. Rev. 2003,

238-239, 233.

3. Hartmann, M.; Clark, T.; van Eldik, R. J. Mol. Model. 1996, 2, 354.

4. Hartmann, M.; Clark, T.; van Eldik, R. J. Am. Chem. Soc. 1997, 119,

7843.

5. Hartmann, M.; Clark, T.; Van Eldik, R. J. Phys. Chem. A 1999, 103,

9899.

6. Puchta, R.; van Eikema Hommes, N.; van Eldik, R. Helv. Chim. Acta

2005, 88, 911.

7. Puchta, R.; Galle, M.; Van Eikema Hommes, N.; Pasgreta, E.; Van Eldik,

R. Inorg. Chem. 2004, 43, 8227.

8. Schlafer, H. L.; Schaffernicht, W. Angew. Chem. 1960, 72, 618.

9. Martin, D.; Weise, A.; Niclas, H. J. Angew. Chem., Int. Ed. Engl. 1967,

6, 318.

10. Mandal, P. K.; Dutta, P.; Roy, S. C. Tetrahedron Lett. 1997, 38, 7271.

11. Kiselev, V. D.; Kashaeva, E. A.; Luzanova, N. A.; Konovalov, A. I. Russ. J.

Gen. Chem. 1997, 67, 1361.

12. Higashigaki, Y.; Christensen, D. H.; Wang, C. H. J. Phys. Chem. 1981,

85, 2531.

13. Sweat, F. W.; Epstein, W. W. J. Org. Chem. 1967, 32, 835.

14. Epstein, W. W.; Sweat, F. W. Chem. Rev. 1967, 67, 247.

15. Tidwell, T. T. Synthesis 1990, 10, 857.

16. Mancuso, A. J.; Swern, D. Synthesis 1981, 3, 165.

17. Pummerer, R. Ber. Dtsch. Chem. Ges. 1909, 42, 2282.

18. Laleu, B.; Machado, M. S.; Lacour, J. Chem. Commun. 2006, 26, 2786.

19. Li, Z.; Xu, X.; Peng, Y.; Jiang, Z.; Ding, C.; Qian, X. Synthesis 2005, 8,

1228.

20. Shearer, A. G.; Hampton, R. Y. EMBO J. 2005, 24, 149.


58

21. Parry, J. M.; Fowler, P.; Quick, E.; Parry, E. M. Cytogenet. Genome Res.

2004, 104, 283.

22. Scholl, P. F.; McCoy, L.; Kensler, T. W.; Groopman, J. D. Chem. Res.

Toxicol. 2006, 19, 44.

23. Coluccia, M.; Sava, G.; Loseto, F.; Nassi, A.; Boccarelli, A.; Giordano, D.;

Alessio, E.; Mestroni, G. Eur. J. Cancer 1993, 29A, 1873.

24. Bardouki, H.; Barcellos da Rosa, M.; Mihalopoulos, N.; Palm, W. U.;

Zetzsch, C. Atmos. Environ. 2002, 36, 4627.

25. Haschek, W. M.; Baer, K. E.; Rutherford, J. E. Toxicology 1989, 54, 197.

26. Wood, D. C.; Weber, F. S.; Palmquist, M. A. J. Pharmacol. Exp. Ther.

1971, 177, 520.

27. Noel, P. R.; Barnett, K. C.; Davies, R. E.; Jolly, D. W.; Leahy, J. S.;

Mawdesley-Thomas, L. E.; Shillam, K. W.; Squires, P. F.; Street, A. E.;

Tucker, W. C.; Worden, A. N. Toxicology 1975, 3, 143.

28. Jacob, S. W.; Wood, D. C. Am. J. Surg. 1967, 114, 414.

29. Karkhaneei, E.; Afkhami, A.; Shamsipur, M. Polyhedron 1996, 15, 1989.


31. Luzar, A.; Soper, A. K.; Chandler, D. J. Chem. Phys. 1993, 99, 6836.

32. Rao, B. G.; Singh, U. C. J. Am. Chem. Soc. 1990, 112, 3803.

33. Skaf, M. S. J. Chem. Phys. 1997, 107, 7996.

34. Skaf, M. S. Mol. Phys. 1997, 90, 25.

35. Liu, H.; Mueller-Plathe, F.; van Gunsteren, W. F. J. Am. Chem. Soc.

1995, 117, 4363.

36. Thomas, R.; Shoemaker, C. B.; Eriks, K. Acta Cryst. 1966, 21, 12.

37. Itoh, S.; Ohtaki, H. Z. Naturforsch., A: Phys. Sci. 1987, 42, 858.

38. Onthong, U.; Bako, I.; Radnai, T.; Hermansson, K.; Probst, M. Int. J.

Mass Spectrom. 2003, 223-224, 263.

39. Onthong, U.; Megyes, T.; Bako, I.; Radnai, T.; Grosz, T.; Hermansson,

K.; Probst, M. Phys. Chem. Chem. Phys. 2004, 6, 2136.

40. Gabrielian, L. S.; Markarian, S. A. J. Mol. Liq. 2004, 112, 137.

41. Markarian, S. A.; Gabrielian, L. S.; Bonora, S.; Fagnano, C. Spectrochim.

Acta, Part A 2003, 59, 575.


59

42. Vaisman, I. I.; Berkowitz, M. L. J. Am. Chem. Soc. 1992, 114, 7889.

43. Kirchner, B.; Reiher, M. J. Am. Chem. Soc. 2002, 124, 6206.

44. Luzar, A.; Chandler, D. J. Chem. Phys. 1993, 98, 8160.

45. Laria, D.; Skaf, M. S. J. Chem. Phys. 1999, 111, 300.

46. Skaf, M. S. J. Phys. Chem. A 1999, 103, 10719.

47. Borin, I. A.; Skaf, M. S. J. Chem. Phys. 1999, 110, 6412.

48. Vechi, S. M.; Skaf, M. S. J. Chem. Phys. 2005, 123, 154507/1.

49. Das, A. K.; Madhusoodanan, M.; Tembe, B. L. J. Phys. Chem. A 1997,

101, 2862.

50. Madhusoodanan, M.; Tembe, B. L. J. Phys. Chem. 1995, 99, 44.

51. Madhusoodanan, M.; Tembe, B. L. J. Phys. Chem. 1994, 98, 7090.

52. Mierke, D. F.; Kessler, H. J. Am. Chem. Soc. 1991, 113, 9466.

53. Zheng, Y.-J.; Ornstein, R. L. J. Am. Chem. Soc. 1996, 118, 4175.

54. Markarian, S. A.; Gabrielyan, L. S.; Grigoryan, K. R. J. Solution Chem.

2004, 33, 1005.

55. Markarian, S. A.; Gabrielian, L. S.; Zatikyan, A. L.; Bonora, S.;

Trinchero, A. Vib. Spectrosc. 2005, 39, 220.

56. Gabrielyan, L. S.; Markaryan, S. A. Zh. Fiz. Khim. 2001, 75, 1792.

57. Alia, J. M.; Edwards, H. G. M. Vib. Spectrosc. 2000, 24, 185.

58. Markarian, S. A.; Stockhausen, M. Z. Phys. Chem. 2000, 214, 139.

59. Wang, Z.; Huang, B.; Wang, S.; Xue, R.; Huang, X.; Chen, L.

Electrochim. Acta 1997, 42, 2611.

60. Kloss, A. A.; Ronald Fawcett, W. J. Chem. Soc., Faraday Trans. 1998,

94, 1587.

61. Das, A. K.; Tembe, B. L. J. Chem. Phys. 1998, 108, 2930.

62. Onthong, U.; Megyes, T.; Bako, I.; Radnai, T.; Hermansson, K.; Probst,

M. Chem. Phys. Lett. 2005, 401, 217.

63. Megyes, T.; Bako, I.; Radnai, T.; Grosz, T.; Kosztolanyi, T.; Mroz, B.;

Probst, M. Chem. Phys. 2006, 321, 100.

64. Burk, P.; Koppel, I. A.; Koppel, I.; Kurg, R.; Gal, J.-F.; Maria, P.-C.;

Herreros, M.; Notario, R.; Abboud, J.-L. M.; Anvia, F.; Taft, R. W. J.

Phys. Chem. A 2000, 104, 2824.


60

65. Westphal, E.; Pliego Josefredo, R., Jr. J. Chem. Phys. 2005, 123,

074508.

66. Goldenberg, M. S.; Kruus, P.; Luk, S. K. F. Can. J. Chem. 1975, 53, 1007.

67. Kruus, P.; Poppe, B. E. Can. J. Chem. 1979, 57, 538.

68. Frolov, Y. L.; Guchik, I. V.; Shagun, V. A.; Vaschenko, A. V.; Trofimov, B.

A. J. Struct. Chem. 2003, 44, 927.

69. Kalugin, O. N.; Adya, A. K.; Volobuev, M. N.; Kolesnik, Y. V. Phys. Chem.

Chem. Phys. 2003, 5, 1536.

70. Das, A. K.; Tembe, B. L. J. Chem. Phys. 1999, 111, 7526.

71. Zamir, T.; Khan, A. J. Chem. Soc. Pak. 2005, 27, 130.

72. Kriz, J.; Abbrent, S.; Dybal, J.; Kurkova, D.; Lindgren, J.; Tegenfeldt, J.;

Wendsjo, A. J. Phys. Chem. A 1999, 103, 8505.

73. Emons, H. H.; Birkeneder, F.; Pollmer, K. Z. Anorg. Allg. Chem. 1986,

534, 175.


75. Ali, A.; Nain, A. K. Phys. Chem. Liq. 1997, 34, 25.

76. Chauhan, M. S.; Sharma, K. C.; Kumar, G.; Chauhan, S. Collect. Czech.

Chem. Commun. 2002, 67, 1141.

77. Maciel, G. E.; Hancock, J. K.; Lafferty, L. F.; Mueller, P. A.; Musker, W.

K. Inorg. Chem. 1966, 5, 554.


1975, 79, 80.

79. Wuepper, J. L.; Popov, A. I. J. Am. Chem. Soc. 1970, 92, 1493.

80. Mandal, B.; Sooksimuang, T.; Griffin, B.; Padhi, A.; Filler, R. Solid State

Ionics 2004, 175, 267.

81. Abe, T.; Kawabata, N.; Mizutani, Y.; Inaba, M.; Ogumi, Z. J.

Electrochem. Soc. 2003, 150, A257.

82. Kumar, G.; Janakiraman; Namboodiri, N.; Gangadharan; Saminathan;

Kanagaraj; Sharief, N. J. Chem. Eng. Data 1991, 36, 467.

83. Barthel, J.; Gores, H. J.; Neueder, R.; Schmid, A. Pure Appl. Chem.

1999, 71, 1705.



61


Phys. 1988, 37, 785.

86. Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys.

Chem. 1994, 98, 11623.

87. Dunning, T. H.; Jr.; Hay, P. J. Modern Theoretical Chemistry; Plenum:

New York, 1976; Vol. 3, p 1.

88. Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270.

89. Wadt, W. R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284.

90. Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299.

91. Huzinaga, S.; Andzelm, J.; Klobukowski, M.; Radzio-Andzelm, E.; Sakai,

Y.; Tatewaki, H. Gaussian basis sets for molecular calculations;

Elsevier: Amsterdam, 1984.



528.

93. Davidson, E. R.; Feller, D. Chem. Rev. 1986, 86, 681.

94. Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995.

95. Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem. 2003,

24, 669.

96. Cammi, R.; Mennucci, B.; Tomasi, J. J. Phys. Chem. A 2000, 104, 5631.

97. Foresman, J. B.; Keith, T. A.; Wiberg, K. B.; Snoonian, J.; Frisch, M. J. J.

Phys. Chem. 1996, 100, 16098.

98. Illner, P.; Zahl, A.; Puchta, R.; van Eikema Hommes, N.; Wasserscheid,

P.; van Eldik, R. J. Organomet. Chem. 2005, 690, 3567.

99. Puchta, R.; Clark, T.; Bauer, W. J. Mol. Model. 2006, 12, 739.

100. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;

Cheeseman, J. R.; Montgomery, J. A.; Jr.; Vreven, T.; Kudin, K. N.;

Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;

Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;

Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa,

J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.;

Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.;


62

Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;

Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;

Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,

S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;

Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;

Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;

Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,

M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;

Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian

03, Revision B.03, Gaussian Inc.; Wallingford, CT, 2004.

101. Maxey, B. W.; Popov, A. I. J. Am. Chem. Soc. 1968, 90, 4470.

102. Marcus, Y. Chem. Soc. Rev. 1993, 22, 409.

103. Richens, D. T. The Chemistry of Aqua Ions; Wiley&Sons: Chichester,

1997.


105. Dunand, F. A.; Helm, L.; Merbach, A. E. Adv. Inorg. Chem. 2003, 54, 1.

3. Acetonitrile and Hydrogen Cyanide

63


Lithium Cations. Acetonitrile and Hydrogen

Cyanide

3.1. General remark



Pasgreta, E.; Puchta, R.; Zahl, A.; van Eldik, R. European Journal of Inorganic

Chemistry 2007, 1815-1822.

3.2. Abstract

Solutions of LiClO4 in solvent mixtures of acetonitrile and water, or

acetonitrile and nitromethane, were studied by 7Li NMR spectroscopy.

Measured chemical shifts indicate that the Li+ cation is coordinated by four

acetonitrile molecules. In the binary water/acetonitrile mixture, water

coordinates more strongly to Li+ than acetonitrile such that addition of water

immediately leads to the formation of [Li(H2O)4]+. The solvent-exchange

mechanism of [Li(L)4]+ (L = CH3CN and HCN) was studied using DFT

calculations (RB3LYP/6-311+G**). This process was found to follow a limiting

associative mechanism involving the formation of relatively stable five-

coordinate intermediates. The suggested mechanisms are discussed with

reference to available experimental and theoretical data.


64

3.3. Introduction

The ongoing discussion on sustainability in chemistry is partly focused on

the role of solvents and especially on alternatives for common organic solvents.

Such alternatives could be, for example, ionic liquids1-4 or supercritical carbon

dioxide.5-8 A detailed knowledge of the properties and structural motifs, as well

as of elementary reactions such as solvent-exchange processes in different

solvents, is an important prerequisite for selecting, optimizing and tailoring

the ideal solvent.

Acetonitrile is a common solvent in chemical laboratories, for example it is

used in reactions9, 10 or in the investigation of complexation phenomena and

complex-formation constants.11-25 Use of its hydrogen derivative HCN is mostly

prevented by its high toxicity and its boiling point of 25 °C. Experimental

studies dating back to the middle of the 20th century show that HCN is a

water-like solvent that is well suited for dissolving organic and inorganic

compounds. Salts such as LiCl, LiBr or LiClO4 are highly soluble in hydrogen

cyanide, most likely because of its high dielectric constant of 123 (at 15.6 °C).26-

30 However, HCN serves as an excellent simplified model of CH3CN and thus

can be used instead in quantum chemical studies of CH3CN.

The behaviour of pure acetonitrile and acetonitrile mixtures was studied

experimentally31, 32 and computationally.31, 33 Since an ion in solution strongly

disturbs the local solvent structure, a detailed knowledge of the environment is

an essential prerequisite for understanding reactions in the neighbourhood of

the ion. Several studies investigated solvated Li+ ions in acetonitrile both

experimentally with various spectroscopic techniques34-44 and computationally

with different classical molecular dynamics (MD) methods45-52 and Monte

Carlo simulations on solvated alkali and halide ions in CH3CN.53 Over the past

decade, DFT and ab initio calculations on [Li(CH3CN)n]+ and related clusters

were also performed by different groups.37, 54-57

There has been a special interest in the properties of Li salts, such as

conductivity,58-64 transport ability,65-67 viscosity,68 and osmotic behaviour,69 in


65

solvents and solvent mixtures because of the use of Li+ ions in batteries of high

energy density.70 Such studies are not restricted to liquid systems. Li+ mobility

in potential solid electrolytes was also investigated intensively both

experimentally and theoretically.71-73

To the best of our knowledge, however, investigations of solvent-exchange

reactions in acetonitrile and acetonitrile/water mixtures or in HCN and

HCN/water mixtures have not been performed. Whereas experimental

investigations study reactions under real conditions in a whole ensemble of

molecules,74, 75 quantum chemical calculations focus on one molecule and a

small number of solvent molecules, for example, the first solvent shell.76-82 This

approach enables a more detailed mechanistic insight.

In this contribution, we extend our combined quantum chemical and

experimental investigations on ligand coordination and ligand exchange on Li+

in acetonitrile and HCN, as well as in CH3CN/H2O and HCN/H2O mixtures.

The calculations mainly focus on HCN for two reasons: (1) to gain a deeper

insight into this solvent, which is not used often, and (2) HCN is a good

working model for CH3-CN as its use prevents well known problems associated

with rotating CH3 groups and to facilitate studies with solvent models.



All transformations were carried out under an Ar or N2 atmosphere by

using standard Schlenk techniques, under Ar atmosphere in a glove box, or

under vacuum. LiClO4 (Aldrich, battery grade) was vacuum dried at 120 °C for

48 h and was then stored under dry nitrogen. Acetonitrile (Acros, extra dry,

water < 10 ppm) was used as received. Nitromethane (Roth) was purified and

dried by a previously reported method.83 Solutions of LiClO4 were prepared

under dry nitrogen. For 7Li NMR spectroscopic studies, these solutions were

sealed under an Ar atmosphere in 5-mm NMR tubes.


66


Fourier transform 7Li NMR spectra were recorded at a frequency of 155

MHz on a Bruker Avance DRX 400WB spectrometer equipped with a

superconducting BC-94/89 magnet system. Spinning 5-mm tubes were used

with a 1-mm o.d. melting point capillary inserted coaxially in the spinning tube

and filled with an external reference solution (usually 1 M LiClO4 in DMF). The

experiments were conducted at room temperature and at ambient pressure.

For a typical kinetic run, 0.8 mL of solution was transferred under Ar to a

Wilmad screw cup NMR tube equipped with a poly(tetrafluoroethylene)

septum. All samples were prepared in the same way.


All structures were fully optimized at the B3LYP/6-311+G**84-87 level of

theory and characterized as minima or transition-state structures by

computation of vibrational frequencies (for minima, all frequencies are

positive, NIMAG = 0; for transition-state structures, exactly one imaginary

frequency is present, NIMAG = 1). Single-point energy calculations were

performed at the MP2(full)/6-311+G**//B3LYP/6-311+G** level;88 the

influence of th bulk solvent was probed by using the IPCM89 and CPCM90,91

formalism and water or acetonitrile as solvent, i.e. B3LYP(IPCM)/6-

311+G**//B3LYP/6-311+G** and B3LYP(CPCM)/6-311+G**//B3LYP/6-

311+G**.92 The Gaussian 03 suite of programs was used throughout.93


3.5.1. Coordination Number of Li+

The structures of Li+-acetonitrile solvates are in general less well known

than the structures of Li+-water solvates. The Li+ ion was found to be four

coordinate by NMR spectroscopic studies in which acetonitrile was gradually

replaced by water when water was added to an acetonitrile solution of LiClO4

(1.6 M),94 and on the basis of IR spectroscopic intensities measured for the

acetonitrile CN stretching vibrations.95, 96 Even compounds with mixed


67

coordination of a counterion and acetonitrile were reported to be four-

coordinate, namely [Li(CH3CN)3Br] formed from 0.58 M LiBr in CH3CN.42

Extensive concentration-dependent ion-pairing was also found for weak

coordinating counterions, such as in LiBF437 and LiClO4,95 by studying the CN

vibration.97 These findings for the coordination number are in good agreement

with published X-ray structures.98-100

However, Sajeevkumar and Singh found the average coordination number

to be 4.7,101 whereas Megyes et al. found up to six acetonitrile molecules

surrounding the Li+ ion in the gas phase by means of mass spectrometry.102

Therefore, we reinvestigated the maximum coordination number of

[Li(CH3CN)n]+ both experimentally and computationally, and purely quantum

chemically for [Li(NCH)n]+.

In order to determine the coordination number of Li+ by using NMR

techniques, the concentration of LiClO4 was kept constant and the CH3CN

concentration was varied over a wide range. To use this method, the cation

must be dissolved in solvent mixtures consisting of one coordinating and one

non-coordinating component. The measured chemical shifts are then plotted

against the mole ratio of cation to coordinating solvent. When such a plot

shows a marked discontinuity, the appropriate solvent/cation ratio can be

taken as the coordination number of the cation. There are, however, a few

conditions that must be obeyed. An inert non-coordinating solvent must be

available that will dissolve all solution components, a measurable parameter

must exist that is a function of the solvent composition, and the formation

constant of the solvate must be sufficiently large so that a limiting value of the

parameter can be reached. In this case, the selected parameter was the Li+

NMR signal; its position in the NMR spectrum depends on the composition of

the coordination sphere of the studied ion. Typical alkali metal salts are not

appreciably soluble in truly inert solvents. Therefore, it is necessary to select a

solvent that minimally interacts with CH3CN and the alkali metal ion.103

Nitromethane (Gutmann donor number DN = 2.7)104 represents a good

compromise for the various requirements. It appreciably dissolves alkali metal


68

salts and its interaction with Li+ is much weaker than that between CH3CN

(Gutmann donor number DN = 14.1) and a metal cation. The mole ratio data

obtained confirms the “inertness” of this solvent and indicates that it is in fact

sufficiently unreactive for the purpose of this study.

Figure 3.1 and Figure 3.2 illustrate the results obtained from the 7Li NMR

spectroscopic study. Figure 3.2 shows a plot of the chemical shift as a function

of the mole ratio of the CH3CN/LiClO4 system; the LiClO4 concentration was

kept at 0.05 M in nitromethane. A clear break can be observed at a

[CH3CN]/[Li+] ratio of 4:1, which indicates that Li+ is coordinated by 4

acetonitrile molecules under these conditions. It appears, therefore, that NMR

studies of alkali metal ion coordination can be carried out even in solvents that

are not totally inert toward the reacting metal ion provided that the interaction

between the metal ion and the stronger donor solvent dominates.

Figure 3.1. 7Li NMR spectra recorded as a function of CH3CN/LiClO4 mole ratio (0.05 M LiClO4 in the CH3CN/nitromethane mixture) at 25 ºC.

6.06.26.46.66.87.07.27.47.67.88.08.2(ppm)

6.6

5.7

4.7

3.8

3.3

2.4

1.9

1

8.5

mole ratio CH3CN/LiClO4

ref. 0.1 M LiClO4 in DMF

06.06.26.46.66.87.07.27.47.67.88.08.2

(ppm)6.06.26.46.66.87.07.27.47.67.88.08.2

(ppm)

6.6

5.7

4.7

3.8

3.3

2.4

1.9

1

8.5

mole ratio CH3CN/LiClO4


0


69

0 2 4 6 8 10

6.95

7.00

7.05

7.10

7.15

7.20

7.25

chem

ical

shi

ft, p

pm

mole ratio, CH3CN/LiClO4

Figure 3.2. 7Li NMR shift measured for 0.05 M LiClO4 in CH3CN/nitromethane solvent

mixtures.

3.5.2. Selective Coordination of Li+

Selective coordination occurs when the solvate prefers one of the solvent

components over the other. The exchange of ions between different

environments is usually rapid with respect to the NMR time scale, and this

results in only one resonance signal at an average frequency determined by the

magnetic shielding and lifetime of the nucleus in each of the sites. Variation of

parameters such as concentration, counterions and solvent produces changes

in the relative proportion and type of environment, which may be reflected by

the NMR spectra in terms of chemical shift, line shape, and/or line width of

the observed resonance.105

We performed the measurements by keeping the concentration of LiClO4

constant (0.1 M) and by changing the composition of the H2O/CH3CN

mixtures. The concentration of H2O (molar ratio) was increased from 0 to 100

%. For each sample, the 7Li NMR spectrum was recorded. The chemical shift

changed as the amount of water in the sample was varied.


70

It was observed that the position of the NMR signal moved significantly

(from 4.5 to 6.4 ppm) even when the amount of added water was very small

(only 10% H2O). The chemical shift value then increased slightly and remained

almost constant when the amount of water was changed from 10 to 100% (see

Figures 3.3 and 3.4). This observation confirms that water coordinates much

more strongly to the Li+ ion than acetonitrile. Thus, addition of water

immediately leads to the formation of the aquated cation.

Figure 3.3. 7Li NMR spectra of 0.1 M LiClO4 in H2O/CH3CN mixtures.

3.03.54.04.55.05.56.06.57.07.58.08.59.0(ppm)

percentage of H2O

30

20

10

8

6

4

2

0

90

80

70

60

50

40

100


3.03.54.04.55.05.56.06.57.07.58.08.59.0(ppm)

3.03.54.04.55.05.56.06.57.07.58.08.59.0(ppm)

percentage of H2O

30

20

10

8

6

4

2

0

90

80

70

60

50

40

100



71

0 20 40 60 80 100

4.5

5.0

5.5

6.0

6.5

7.0

chem

ical

shi

ft, p

pm

% mol H2O

Figure 3.4. 7Li NMR shifts for 0.1 M LiClO4 as a function of the H2O/CH3CN composition.

3.5.3. Ligand Exchange on Solvated Li+

It was not possible to follow the exchange process for CH3CN on Li+

experimentally. As reported before, the rate constant of water exchange on Li+

is extremely fast (approximately 109 s-1 at 25 ºC) although water is a relatively

strong donor (DN = 33.0). As CH3CN coordinates much more weakly (DN =

14.1), its exchange on Li+ is even faster, which makes it impossible to study

experimentally. Therefore, theoretical calculations are essential and may be

very useful in such investigations.


We studied the ligand exchange process in more detail and provide

computational evidence for a limiting associative exchange mechanism on the

HCN- and CH3CN-solvated lithium cation. As mentioned before, HCN was

employed as a simplified working model to perform quantum chemical

calculations on CH3CN. Our calculations corroborated the results of the

experimental study in solution (vide supra)35 and in the solid state,100 which


72

showed that the first coordination sphere around Li+ consists of four tightly

bound CH3CN and HCN molecules. The five-coordinate intermediates are 3.7

kcal/mol and 4.6 kcal/mol (3.2 kcal/mol B3LYP(IPCM: acetonitrile)/6-

311+G**//B3LYP/6-311+G**) higher in energy for acetonitrile and HCN,

respectively. The computed Li-N distance for Li(NCH)4 and Li(NCCH3)4 are

2.06 and 2.05 Å, respectively. Additional solvent molecules do not coordinate

to Li+, but instead form a second coordination sphere, mainly through

electrostatic interactions. The gas-phase binding energy for the fifth

acetonitrile molecule in [Li(NCCH3)4(NCCH3)]+ is 6.1 kcal/mol, with a Li-N

distance of 6.65 Å. The fifth molecule is stabilized only by a weak HCH3-NC-CH3

interaction of 2.32 Å. The fifth HCN molecule is bound through a weak

hydrogen bond (2.04 Å) in the second coordination sphere. This bond

stabilizes the system by 8.9 kcal/mol (see Figure 3.5).

Figure 3.5. Energy profile (RB3LYP/6-311+G**) for the exchange of HCN on [Li(NCH)4]+.

The HCN exchange itself proceeds through a trigonal-bipyramidal

intermediate [Li(NCH)5]+ that is reached via a late transition state. The


73

entering HCN molecule approaches the lithium cation directly and pushes

three coordinated solvent molecules away toward the equatorial positions. In

line with the experimental observation of a very fast exchange process (vide

supra), the computed activation barrier is only 5.5 kcal/mol, while the five-

coordinate intermediate is 4.6 kcal/mol less stable than the precursor complex.

The very low barrier toward dissociation of only 0.9 kcal/mol means that the

five-coordinate intermediate is very short-lived. The evaluation of the energy

at the MP2(full)/6-311+G**//B3LYP/6-311+G** and B3LYP/6-

311+G**//B3LYP/6-311+G** levels shows only minor changes in the computed

energies, except in those for the intermediate ([Li(L)4(L)]+) (see Table 3.1). We

attribute this to an incorrectly balanced description of the metal-ligand and

hydrogen-bond strengths by DFT methods as observed earlier by Rotzinger

and others in studies on water exchange reactions.106,107 Inclusion of solvent

effects at the B3LYP(CPCM)/6-311+G**//B3LYP/6-311+G** and

B3LYP(IPCM)/6-311+G**//B3LYP/6-311+G** level clearly favours a four-

coordinate structure with NCH in the second coordination sphere or

uncoordinated. Therefore, the transition state is also stabilized since the

entering NCH molecule can best be addressed as uncoordinated (see Table 3.1

and Figure 3.5). Unfortunately, the transition state is modelled as being too

stable in the B3LYP(IPCM)/6-311+G**//B3LYP/6-311+G** approximation,

irrespective of whether acetonitrile or water is used in the modelling by the

IPCM formalism.

In trigonal-bipyramidal [Li(NCH)5]+, the axial Li-N distances are 0.11 Å

longer than the equatorial distances, which may be taken as a clear indication

that one of the axial ligands will be expelled. The energy profile for the

exchange of acetonitrile on [Li(NCCH3)4]+ shows the same topology; only the

energy differences are somewhat smaller. We attribute this difference to

stabilization effects caused by the methyl group and to the absence of hydrogen

bonding. A detailed investigation was prevented since a local minimum was

not reached with [Li(NCCH3)4(NCCH3)]+ because of the rotation of methyl

groups, and IPCM calculations show the well-documented problems in

reaching an energy convergence.3,108


74

Table 3.1. Calculated relative energies (in kcal/mol) for the exchange of L at [Li(L)4]+ (L: HCN, CH3CN).[a]

L: HCN [Li(L)4]+ + L [Li(L)4…(L)]+ ts [Li(L)4(L)]+

B3LYP 8.9 0.0 5.5 4.6

MP2(full) 9.5 0.0 3.2 0.5

IPCM -2.8 0.0 -2.6 3.2

CPCM -0.4 0.0 1.8 1.4

L: H3CCN

B3LYP 6.1 0.0 4.1 3.7

[a] B3LYP: B3LYP/6-311+G**//B3LYP/6-311+G**; MP2(full): MP2(full)/6-311+G**//B3LYP/6-311+G**; IPCM: B3LYP/(IPCM: Acetonitrile)/6-311+G**//B3LYP/6-311+G**; IPCM: B3LYP/(CPCM: water)/6-311+G**//B3LYP/6-311+G**

These findings fully corroborate the results expected for small ions

coordinated by small ligands. The dissociation of a molecule of HCN from

[Li(NCH)4]+ results in the intermediate [Li(NCH)3]+ with a D3h symmetry,

which is 14.3 kcal/mol higher in energy and less favourable. In addition, there

is no visible steric crowding, which is necessary for a dissociative mechanism.

Interchange mechanisms are known for Li+, e.g. in the case of ammonia

exchange on Li+,82 but requires a five-coordinate transition state that would

contradict the five-coordinate intermediate in our mechanism. Analogous

arguments can also be applied for the second reaction under investigation.

Under actual experimental conditions water can never be excluded.

Therefore, and because of the importance of solvent mixtures in technical and

laboratory applications, we modelled the HCN/H2O exchange process

exemplarily for [Li(NCH)4]+ to [Li(NCH)3(H2O)]+ (Table 3.2). In this case,

NCH can again be considered as a model for acetonitrile.

From experiments (vide supra) and from the Gutmann donor number for

acetonitrile, which is only 50% of that for water, it is known that addition of

water to solutions of lithium ions in acetonitrile leads to the formation of a


75

water-coordinated Li+ complex. This can be reproduced for HCN as model by

quantum chemical calculations (RB3LYP/6-311+G**).

Table 3.2. Calculated energies for the subsequent reactions of [Li(NCH)4]+ with water.

B3LYP(IPCM)/6-311+G**//B3LYP/6-311+G**

[kcal/mol]

Gas

phase

IPCM:

acetonitrile

IPCM:

water

[Li(NCH)4]+ + H2O [Li(H2O)(H2O)3]+ + NCH -0.5 +1.7 +1.6

[Li(H2O)( NCH)3]+ + H2O [Li(H2O)2(NCH)2]+ +

NCH -0.5 -1.4 -1.6

[Li(H2O)2(NCH)2]+ + H2O [Li(H2O)3(NCH)]+ +

NCH -0.3 -3.6 -3.9

[Li(H2O)3(NCH)]+ + H2O [Li(H2O)4]+ + NCH +0.1 -0,9 -1.2

∑ -1.3 -4.2 -5.1

By including solvent effects at the B3LYP(IPCM)/6-311+G**//B3LYP/6-

311+G** level, the reaction energies are significantly increased, regardless of

whether water or acetonitrile is selected as solvent.

Figure 3.6. Energy profile (RB3LYP/6-311+G**) for the HCN/water exchange around

[Li(NCH)4]+.


76

We investigated the mechanism for the first reaction step in detail (see

Figure 3.6). As shown before,35,100 the first coordination sphere around Li+

consists of four tightly bound acetonitrile molecules, here NCH molecules. An

additional solvent molecule does not coordinate to the Li+ ion; a second

coordination sphere is formed in which water is bound by one linear hydrogen

bond (1.88 Å). The gas-phase hydrogen-bond energy for the H2O molecule in

[Li(NCH)4…(OH2)]+ is 9.5 kcal/mol. Again, the entering solvent, here water,

approaches the Li+ ion directly through a face of the [Li(NCH)4]+ tetrahedron

and pushes three coordinated HCN molecules away. According to the

movement of the water molecule in the transition state (7.1 kcal/mol higher in

energy), one would expect an intermediate with three HCN molecules in

equatorial positions and the fourth together with H2O in the axial position.

During the optimization steps, the structure rearranges to form an

intermediate 3.7 kcal/mol lower in energy, with two HCN molecules in the

axial position and the H2O molecule together with the two remaining HCN

molecules in the equatorial positions. The bonds for the axial molecules are

0.16 Å longer than those for the equatorial HCN ligands - it can be concluded

that one axial ligand will be expelled to reach a four-coordinate product. The

activation barrier for the release of an axial HCN molecule is rather low, (0.1

kcal/mol), which means that the five-coordinate intermediate will be very

short-lived. On the assumption that the expelled HCN will form a hydrogen

bond with the nearest hydrogen atom, one would expect the product

[Li(NCH)3(H2O)…(NCH)]+ to form at an energy that is 4.2 kcal/mol lower than

that of the second transition state. The energy required to destroy this

hydrogen bond is 9.7 kcal/mol (Table 3.3).


77

Table 3.3. Calculated relative energies for the exchange of H2O/NCH at [Li(NCH)4]+ (L = HCN; W = H2O).[a]

L:HCN; W: H2O [Li(L)4…W]+ ts [Li(L)2{W(L)2}]+ ts [Li(L)3(W)…L]+

B3LYP 0.7 7.8 4.1 4.2 0.0

MP2(full) 0.9 6.1 0.6 1.2 0.0

IPCM 0.1 2.8 1.3 1.8 0.0

CPCM 0.1 5.6 1.0 1.5 0.0

[a] B3LYP: B3LYP/6-311+G**// B3LYP/6-311+G**; MP2(full): MP2(full)/6-

311+G**//B3LYP/6-311+G**; IPCM: B3LYP/(IPCM: Acetonitrile)/6-311+G**//B3LYP/6-

311+G**

Whereas MP2(full)/6-311+G**//B3LYP/6-311+G** calculations show

typical discrepancies, the application of the IPCM- and CPCM-solvent models

dramatically lowers the energy for the intermediate and the transition states.

The barriers of 2.8 and 1.8 kcal/mol corroborate the experimental findings for

an efficient formation of a water coordinated Li+ complex.

3.6. Conclusions

It was confirmed both experimentally and theoretically that Li+ is

coordinated by four acetonitrile molecules. In the binary mixture of water and

acetonitrile, water is coordinated more strongly to the Li+ ion than acetonitrile.

Thus, addition of water immediately leads to the formation of the aquated

cation. Ligand exchange reactions on Li+ most probably follow an associative

mechanism as indicated by the formation of stable five-coordinate

intermediates.


78

3.7. References

1. Wasserscheid, P.; Welton, T. Ionic Liquids in Synthesis; Wiley-VCH

Verlag GmbH & Co. KGaA: Weinheim, Germany, 2003; p 364.

2. Wasserscheid, P. Chem. unserer Zeit 2003, 37, 52.

3. Illner, P.; Zahl, A.; Puchta, R.; van Eikema Hommes, N.; Wasserscheid,

P.; van Eldik, R. J. Organomet. Chem. 2005, 690, 3567.

4. Weber, C. F.; Puchta, R.; van Eikema Hommes, N. J. R.; Wasserscheid,

P.; van Eldik, R. Angew. Chem., Int. Ed. 2005, 44, 6033.

5. Clifford, A.; Clifford, T. Fundamentals of Supercritical Fluids; Oxford

University Press: Oxford, 1999; p 228.

6. Erkey, C. J. Supercrit. Fluids 2000, 17, 259.

7. Oakes, R. S.; Clifford, A. A.; Rayner, C. M. J. Chem. Soc., Perkin Trans. 1

2001, 9, 917.

8. Jessop, P. G.; Ikariya, T.; Noyori, R. Science 1995, 269, 1065.

9. Kinart, W. J.; Sniec, E.; Tylak, I.; Kinart, C. M. Phys. Chem. Liq. 2000,

38, 193.

10. Babu, B. S.; Balasubramanian, K. K. Synlett 1999, 8, 1261.

11. D'Aprano, A.; Salomon, M.; Mauro, V. J. Solution Chem. 1995, 24, 685.

12. Shamsipur, M.; Madrakian, T. Main Group Met. Chem. 2001, 24, 239.

13. Karkhaneei, E.; Zebarjadian, M. H.; Shamsipur, M. J. Solution Chem.

2001, 30, 323.

14. Shamsipur, M.; Madrakian, T. J. Coord. Chem. 2000, 52, 139.

15. Hojo, M.; Hisatsune, I.; Tsurui, H.; Minami, S.-I. Anal. Sci. 2000, 16,

1277.

16. Shamsipur, M.; Madrakian, T. Polyhedron 2000, 19, 1681.

17. Shamsipur, M.; Karkhaneei, E.; Afkhami, A. J. Coord. Chem. 1998, 44,

23.

18. Danil de Namor, A. F.; Pulcha Salazar, L. E.; Llosa Tanco, M. A.;

Kowalska, D.; Villanueva Salas, J.; Schulz, R. A. J. Chem. Soc., Faraday

Trans. 1998, 94, 3111.


79

19. Karkhaneei, E.; Zolgharnein, J.; Afkhami, A.; Shamsipur, M. J. Coord.

Chem. 1998, 46, 1.

20. de Namor, A. F. D.; Zapata-Ormachea, M. L.; Jafou, O.; Al Rawi, N. J.

Phys. Chem. B 1997, 101, 6772.

21. D'Aprano, A.; Vicens, J.; Asfari, Z.; Salomon, M.; Iammarino, M. J.

Solution Chem. 1996, 25, 955.

22. de Namor, A. F. D.; Ng, J. C. Y.; Tanco, M. A. L.; Salomon, M. J. Phys.

Chem. 1996, 100, 14485.

23. Fakhari, A. R.; Shamsipur, M. J. Inclusion Phenom. 1996, 26, 243.

24. Karkhaneei, E.; Afkhami, A.; Shamsipur, M. J. Coord. Chem. 1996, 39,

33.

25. de Namor, A. F. D.; Gil, E.; Tanco, M. A. L.; Tanaka, D. A. P.; Salazar, L.

E. P.; Schulz, R. A.; Wang, J. J. Phys. Chem. 1995, 99, 16776.

26. Jander, G.; Klemm, W. Die Chemie in wasserähnlichen Lösungsmitteln;

Springer-Verlag: Berlin, 1949; p 367.

27. Jander, J.; Lafrenz, C. Wasserähnliche Lösungsmittel; Springer-Verlag:

Berlin, 1968; Vol. 3, p 212.

28. Jander, J.; Tuerk, G. Chem. Ber. 1962, 95, 881.



31. Bako, I.; Megyes, T.; Palinkas, G. Chem. Phys. 2005, 316, 235.

32. Lu, H.; Wang, J.; Zhao, Y.; Xuan, X.; Zhuo, K. J. Chem. Eng. Data 2001,

46, 631.

33. Rivelino, R.; Cabral, B. J. C.; Coutinho, K.; Canuto, S. Chem. Phys. Lett.

2005, 407, 13.

34. Mollner, A. K.; Brooksby, P. A.; Loring, J. S.; Bako, I.; Palinkas, G.;

Fawcett, W. R. J. Phys. Chem. A 2004, 108, 3344.

35. Barthel, J.; Buchner, R.; Wismeth, E. J. Solution Chem. 2000, 29, 937.

36. Brouillette, D.; Irish, D. E.; Taylor, N. J.; Perron, G.; Odziemkowski, M.;

Desnoyers, J. E. Phys. Chem. Chem. Phys. 2002, 4, 6063.

37. Xuan, X.; Zhang, H.; Wang, J.; Wang, H. J. Phys. Chem. A 2004, 108,

7513.


80

38. Fawcett, W. R.; Brooksby, P.; Verbovy, D.; Bako, I.; Palinkas, G. J. Mol.

Liq. 2005, 118, 171.

39. Il'in, K. K.; Demakhin, A. G. Russ. J. Gen. Chem. 1999, 69, 703.

40. Loring, J. S.; Fawcett, W. R. J. Phys. Chem. A 1999, 103, 3608.

41. Matsubara, K.; Kaneuchi, R.; Maekita, N. J. Chem. Soc., Faraday Trans.

1998, 94, 3601.

42. Cartailler, T.; Kunz, W.; Turq, P.; Bellisent-Funel, M. C. J. Phys.:

Condens. Matter 1991, 3, 9511.

43. Alia, J. M.; Diaz de Mera, Y.; Edwards, H. G. M.; Garcia, F. J.; Lawson,

E. E. J. Mol. Struct. 1997, 408-409, 439.

44. Alia, J. M.; Diaz De Mera, Y.; Edwards, H. G. M.; Garcia, F. J.; Lawson,

E. E. Z. Phys. Chem. 1996, 196, 209.

45. Maroncelli, M. J. Chem. Phys. 1991, 94, 2084.

46. Troxler, L.; Wipff, G. J. Am. Chem. Soc. 1994, 116, 1468.

47. Richardi, J.; Fries, P. H.; Krienke, H. J. Chem. Phys. 1998, 108, 4079.

48. Cabaleiro-Lago, E. M.; Rios, M. A. Chem. Phys. 1998, 236, 235.

49. Guardia, E.; Pinzon, R. J. Mol. Liq. 2000, 85, 33.

50. Gregoire, G.; Brenner, V.; Millie, P. J. Phys. Chem. A 2000, 104, 5204.

51. Spångberg, D.; Hermansson, K. Chem. Phys. 2004, 300, 165.

52. Kolesnik, Y. V.; Kalugin, O. N. Russ. J. Electrochem. 2003, 39, 427.

53. Fischer, R.; Richardi, J.; Fries, P. H.; Krienke, H. J. Chem. Phys. 2002,

117, 8467.

54. Dimitrova, Y. THEOCHEM 1995, 334, 215.

55. Cabaleiro-Lago, E. M.; Rios, M. A. Chem. Phys. 2000, 254, 11.

56. Pejov, L. Int. J. Quantum Chem. 2002, 86, 356.

57. Bako, I.; Megyes, T.; Radnai, T.; Palinkas, G.; Probst, M.; Fawcett, W. R.

Z. Phys. Chem. 2004, 218, 643.

58. Hojo, M.; Hasegawa, H.; Hiura, N. J. Phys. Chem. 1996, 100, 891.

59. D'Aprano, A.; Iammarino, M.; Mauro, V.; Sesta, B. J. Electroanal. Chem.

1995, 392, 27.

60. Brouillette, D.; Perron, G.; Desnoyers, J. E. J. Solution Chem. 1998, 27,

151.


81

61. Hojo, M.; Ueda, T.; Nishimura, M.; Hamada, H.; Matsui, M.; Umetani,

S. J. Phys. Chem. B 1999, 103, 8965.

62. Izonfuo, W.-A. L.; Obunwo, C. C. Indian J. Chem., Sect. A: Inorg., Bio-

inorg., Phys., Theor. Anal. Chem. 1999, 38A, 939.

63. Barthel, J.; Kleebauer, M.; Buchner, R. J. Solution Chem. 1995, 24, 1.

64. Croce, F.; D'Aprano, A.; Nanjundiah, C.; Koch, V. R.; Walker, C. W.;

Salomon, M. J. Electrochem. Soc. 1996, 143, 154.

65. D'Aprano, A.; Sesta, B.; Capalbi, A.; Iammarino, M.; Mauro, V. J.

Electroanal. Chem. 1996, 403, 257.

66. Mauro, V.; D'Aprano, A.; Croce, F.; Salomon, M. J. Power Sources

2005, 141, 167.

67. Faverio, C. L.; Mussini, P. R.; Mussini, T. Anal. Chem. 1998, 70, 2589.

68. Brouillette, D.; Perron, G.; Desnoyers, J. E. Electrochim. Acta 1999, 44,

4721.

69. Barthel, J.; Neueder, R.; Poepke, H.; Wittmann, H. J. Solution Chem.

1999, 28, 489.

70. Mandal, B.; Sooksimuang, T.; Griffin, B.; Padhi, A.; Filler, R. Solid State

Ionics 2004, 175, 267.

71. Murugavel, S.; Roling, B. J. Phys. Chem. B 2004, 108, 2564.

72. Mariappan, C. R.; Govindaraj, G.; Roling, B. Solid State Ionics 2005,

176, 723.

73. Heuer, A.; Murugavel, S.; Roling, B. Phys. Rev. B: Condens. Matter

Mater. Phys. 2005, 72, 174304/1.


75. Helm, L.; Merbach, A. E. Coord. Chem. Rev. 1999, 187, 151.

76. Erras-Hanauer, H.; Clark, T.; van Eldik, R. Coord. Chem. Rev. 2003,

238-239, 233.

77. Hartmann, M.; Clark, T.; van Eldik, R. J. Mol. Model. 1996, 2, 354.

78. Hartmann, M.; Clark, T.; van Eldik, R. J. Am. Chem. Soc. 1997, 119,

7843.

79. Hartmann, M.; Clark, T.; Van Eldik, R. J. Phys. Chem. A 1999, 103,

9899.


82

80. Rotzinger, F. P. Chem. Rev. 2005, 105, 2003.

81. Puchta, R.; van Eikema Hommes, N.; van Eldik, R. Helv. Chim. Acta

2005, 88, 911.


R. Inorg. Chem. 2004, 43, 8227.

83. Coetzee, J. F.; Chang, T. H. Pure Appl. Chem. 1986, 58, 1541.


Chem. 1994, 98, 11623.



Phys. 1988, 37, 785.



528.

88. Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab Initio

Molecular Orbital Theory; John Wiley & Sons: New York, 1986; p 548.

89. Foresman, J. B.; Keith, T. A.; Wiberg, K. B.; Snoonian, J.; Frisch, M. J. J.

Phys. Chem. 1996, 100, 16098.

90. Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995.

91. Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem. 2003,

24, 669.

92. In the CPCM formalism some molecules failed while using symmetry in

the molecular cavity generation. This problem was solved successfully by

the keyword “nosymm”.










83










94. Keil, R. G.; Johnson, D. W.; Fryling, M. A.; O'Brien, J. F. Inorg. Chem.

1989, 28, 2764.

95. Fawcett, W. R.; Liu, G.; Faguy, P. W.; Foss, C. A., Jr.; Motheo, A. J. J.

Chem. Soc., Faraday Trans. 1993, 89, 811.

96. Barthel, J.; Deser, R. J. Solution Chem. 1994, 23, 1133.

97. Solovieva, L. A.; Akopyan, S. K.; Vilaseca, E. Spectrochim. Acta, Part A

1994, 50A, 683.

98. Brooks, N. R.; Henderson, W. A.; Smyrl, W. H. Acta Crystallogr., Sect.

E: Struct. Rep. Online 2002, E58, m176.

99. Chitsaz, S.; Pauls, J.; Neumuller, B. Z. Naturforsch., B: Chem. Sci. 2001,

56, 245.

100. Yokota, Y.; Young, V. G., Jr.; Verkade, J. G. Acta Crystallogr., Sect. C:

Cryst. Struct. Commun. 1999, C55, 196.

101. Sajeevkumar, V. A.; Singh, S. J. Mol. Struct. 1996, 382, 101.

102. Megyes, T.; Radnai, T.; Wakisaka, A. J. Mol. Liq. 2003, 103-104, 319.

103. Maxey, B. W.; Popov, A. I. J. Am. Chem. Soc. 1968, 90, 4470.



1975, 79, 80.

106. Rotzinger, F. P. J. Phys. Chem. B 2005, 109, 1510.

107. Hanauer, H.; Puchta, R.; Clark, T.; van Eldik, R. Inorg. Chem. 2007, 46,

1112.

108. Puchta, R.; Clark, T.; Bauer, W. J. Mol. Model. 2006, 12, 7

4. Complexation by Cryptands in Acetone as Solvent

84


Lithium Cations. Complexation by

Cryptands in Acetone as Solvent

4.1. General remark



Pasgreta, E.; Puchta, R.; Zahl, A.; van Eldik, R. European Journal of Inorganic

Chemistry 2007, 3067-3076.

4.2. Abstract

Kinetic studies on Li+ exchange between the cryptands C222 and C221 and

acetone as solvent were performed as a function of ligand-to-metal ratio,

temperature and pressure using 7Li NMR spectroscopy. Temperature and

pressure dependence measurements were performed in the presence of an

excess of Li+. The rate and activation parameters for the exchange process are:

C222: k298 = (1.7 ± 0.5) x 103 s-1, ∆H# = 24 ± 1 kJ mol-1 and ∆S# = -105 ± 4 J K-1

mol-1; C221: k298 = 2.24 ± 0.09 s-1, ∆H# = 29 ± 2 kJ mol-1 and ∆S# = -141 ± 5 J

K-1 mol-1. The influence of pressure on the exchange rate is insignificant for

both ligands, such that the activation volume is around zero within the

experimental error limits. The reported activation parameters suggest that the

exchange of Li+ between solvated and chelated ions follows an associative

interchange mechanism.


85

4.3. Introduction

Following the work of Pedersen on two-dimensional crowns, Lehn and

coworkers developed in the late 1960s a range of three-dimensional, polycyclic

ligand systems which they named cryptands (Greek: cryptos = cave).1, 2 Until

today this class of compounds has been intensively explored and extended in

different directions3, 4 in terms of cavity size,5, 6 donor atoms,7 and metalla-

topomers.8-11 Because cryptands readily form complexes with a wide range of

metal ions and NH4+,12 provided the ion involved is not too large to be

contained in the macrocyclic cavity,13 they find their application in various

areas like catalysis, medicine, biomimetic research,14-21 and as receptors22 and

electrides.23 Studies on the complexes (cryptates) formed between alkali-metal

ions and cryptands, have revealed a substantial understanding of the effects of

metal-ion size and cryptand-cavity size on the structure, stability and lability of

cryptates, and the mechanism of the cryptate complex-formation processes.24,

25 Such selective complexation has led to many biological and chemical studies.

Kinetic studies on macrocyclic complex-formation reactions with alkali

cations not only result in important information on the rates and mechanisms

of complex-formation reactions, but also lead to a better understanding of the

high selectivity of these ligands toward different cations. The broad impact of

such studies spans from the theory of complexation to cation transport

through membranes, and makes the ability to control cation selectivity via

tactical structural changes in the complexing agent of prime importance.18

Polyoxo macrobicyclic diamines, represented by the structure in Figure 4.1, are

able to form metal-ion complexes in which the metal ion is located in the cavity

of the macromolecule and two nitrogen donors along with the oxygen donors

participate in binding the metal cation. All complexes have a 1:1 stoichiometry

with the metal ion positioned at the center of the ligand cavity.24


86

O

N

OO

O

N O O

m

m

n Figure 4.1. Chelate structures: C211: m = 0, n = 1; C221: m = 1, n = 0; C222: m = 1, n = 1.

In case of crown ethers and cryptands in non-aqueous solutions, the rates

of complex-formation (kf) with alkali-metal cations are in general diffusion-

controlled and, consequently, the complexation selectivities are governed by

the decomplexation rates (kd).26 These reactions have received significant

attention, initially from Lehn and co-workers and more recently from Popov et

al.27-29 and Lincoln et al..30-37 Many examples have been included in a series of

review articles.38-41 It has been reported that ligand structure and solvent

properties have considerable effects on the reaction rates, activation

parameters, and exchange mechanism of cations between the solvated and

complexed sites. We have now extended our earlier studies42 on Li+ exchange

reactions to acetone as solvent.

Non-aqueous solvents and their mixtures combined with Li+ are often used

in secondary lithium batteries.43, 44 Li+ in acetone is also known to catalyze

Diels-Alder-reactions45 and the ene reaction.46 The behavior of Li+-ions in

different solvents is therefore an important research area.47-57 A variety of

studies on solutions of Li+ in acetone have been mainly studied

experimentally52, 55, 58-63 and some theoretically,64, 65 as well as for

solvent/acetone combinations53 or chelator/acetone combinations.53, 66 In this

contribution, we focus on acetone as a solvent and the competition between

solvated and endohedral complexed Li+, report kinetic parameters for the

exchange of Li+ between solvated and complexed states in acetone under

ambient and elevated pressure conditions, and discuss the mechanistic

implications.


87



The preparation of test solutions was carried out under an Ar or N2 using

standard Schlenk techniques, under an Ar in a glove box, or under vacuum.

Lithium perchlorate (Aldrich, battery grade) was vacuum-dried at 120 °C for

48 h and then stored under dry nitrogen. The cryptands C222 (purchased from

Fluka), C221 and C211 (purchased from Aldrich) were dried under vacuum for

48 h and stored under dry nitrogen. Acetone (Acros, extra dry, water < 50

ppm) was used as received. Solutions of LiClO4, C222, C221 and C211 in

acetone were prepared under dry nitrogen. For the 7Li NMR measurements the

solutions were sealed in 5-mm NMR tubes under Ar. In order to study the

kinetics of the exchange of Li+ between C222 and C221 and acetone as solvent,

three solutions of different concentration ratio were prepared for each system.

The concentration of lithium perchlorate in each solution was kept constant at

0.04 M, whereas the concentration of ligand was varied between 0.01 M and

0.032 M. In this way the measurements were always performed in the

presence of an excess of Li+.


Variable-temperature and variable-pressure Fourier transform 7Li NMR

spectra were recorded at a frequency of 155 MHz on a Bruker Avance DRX

400WB spectrometer equipped with a superconducting BC-94/89 magnet

system. Spinning tubes (5 mm) were used with a 1-mm o.d. melting point

capillary inserted coaxially in the spinning tube and filled with an external

reference solution (1 M LiClO4 in acetonitrile). The temperature dependence

measurements were performed over a wide as possible temperature range for

each system. We did not use a temperature calibration by insertion of a

capillary filled with methanol (low temperature calibration) or ethylene glycol

(high temperature calibration) into the sample tube, as the corresponding

signals of these solvents would disturb the NMR spectra in the region of

intrest. In such a case a temperature-calibrated NMR probe has to be used in


88

order to determine the temperature of the sample accurately. Methanol was

used for the low-temperature calibration of the NMR probe, whereas ethylene

glycol was used for high-temperature calibration. The actual sample

temperature can be calculated from the chemical shift differences. If there is

any difference between the sample temperature and the temperature detected

by the thermocouple of the probe, the probe can be calibrated by using

standard Bruker Topspin software.

A homemade high-pressure probe described in the literature67 was used for

the variable-pressure experiments which were conducted at a selected

temperature and at ambient, 30, 60, 75, 90, 120 and 150 MPa pressure. A

standard 5-mm NMR tube cut to a length of 45 mm was used for the sample

solution. The high-pressure NMR tube was sealed with a MACOR plug and an

O-ring made of VITON. The plug and O-ring were shown to be stable in the

employed solvent. The pressure was transmitted by the movable MACOR

piston to the sample, and the temperature of the sample was controlled with a

Pt-100 resistor.67 The probe itself was externally heated with a thermostatted

circulating fluid that was pumped through a double helix that was cut into the

outer surface of the autoclave. Because the sample is embedded in the high-

pressure transmitting fluid that serves as a thermostating bath, a reliable

temperature setting is guaranteed.

For a typical kinetic run, 0.8 ml of solution was transferred under Ar to a

Wilmad screw-cap NMR tube equipped with a poly(tetrafluoroethylene)

septum. All test samples were prepared in the same way.

4.4.3. Programs

The line widths of the NMR signals were obtained by fitting Lorentzian

functions to the experimental spectra using the NMRICMA 2.7 program68 for

Matlab.69 The adjustable parameters were the resonance frequency, intensity,

line width and baseline. Complete line-shape analysis based on the Kubo-Sack

formalism using modified Bloch equations70 was also performed with

NMRICMA 2.7 to extract rate constants from experimental spectra. The

temperature and pressure dependent 7Li line widths and chemical shifts


89

employed in the line shape analysis were extrapolated from low temperatures

where no exchange-induced modification occurred.


All structures were optimized with the B3LYP71-73 and the basis set

D95Vp,74, 75 or with 6-311+G** with no constrains other than symmetry, and

confirmed to be minima by computation of vibrational frequencies. Relative

energies were corrected for zero point vibrational energy differences. The

D95V basis set, augmented with polarisation functions (D95Vp), was used.76 If

not otherwise mentioned, calculations were done without a solvent model. The

influence of bulk solvent was probed using the CPCM formalism and water as

solvent, i.e., B3LYP(CPCM)/D95Vp//B3LYP/D95Vp. The Gaussian 03 suite of

programs was used throughout.77


4.5.1. Spectroscopic and Kinetic Observations

In preliminary experiments we investigated the Li+ exchange in different

solvents. In water and methanol the exchange process was too fast to be

studied by 7Li NMR (only a single signal was observed). The process became

slower in acetonitrile, and in acetone two completely separated signals were

observed at low temperature. We, therefore, investigated Li+ exchange between

the selected cryptands and acetone as solvent. A similar system was also

investigated before by Popov et al..28 Although they reported a significantly

negative value for the activation entropy, they suggested the operation of a

dissociative exchange mechanism. We realized that it would be worth to

perform a reinvestigation of this system in acetone in order to clarify this

apparent discrepancy, in particular by means of high pressure 7Li NMR

spectroscopy.

Initially, a series of 7Li NMR measurements for the Li+-cryptand system in

acetone were performed. For each of the cryptands C222, C221 and C211,

molar ratio dependent measurements were performed by variation of both the


90

cryptand/Li+ and Li+/cryptand molar ratios. The results enabled us to observe

the distribution of particular species on variation of the concentration ratio.

Basically, two signals of varying intensity were observed in the 7Li NMR

spectra that refer to solvated Li+ and Li+ bound inside the cryptand cavity. In

case of the smallest chelate C211, however, an additional signal appeared

which can be ascribed to Li+ bound outside the cryptand and being partially

solvated. In the case of Li(C211)+, the complex is very stable and almost no

exchange takes place. A detailed study of the observed rate constant, kobs, for

the exchange of Li+ as a function of temperature and pressure was therefore

undertaken only for C221 and C222. The lithium exchange reaction between

cryptand and acetone was studied for three solutions with different ligand-to-

metal ratios in the presence of an excess of Li+.

Three signals can be observed in Figure 4.2; viz. one coming from the

reference (δ = 5.8 ppm) and the other two from the exchanging sites (δ = 9.0

ppm for solvated Li+ and 6.6 ppm for complexed Li+). At low temperature the

signals are sharp and completely separated due to very slow or no exchange.

Because the formation constant for Li(C221)+ is very high,24 the concentration

of bound lithium equals the amount of cryptand added. At ca. 283 K, the

signals start to broaden and the exchange process takes place. The linear

temperature dependence for the temperature range in which exchange takes

place can be seen in Figure 4.5. In the case of C222 (Figure 4.3), the exchange

process is faster due to the larger cavity of the cryptand. The signals that refer

to solvated and complexed Li cation (8.7 and 6.8 ppm, respectively at 223 K)

coalescence at ca. 263 K and at high temperature (>313 K) only one sharp

signal at 7.7 ppm was observed as a consequence of a very fast exchange

process. An accurate temperature dependence for C222 over a large

temperature range can be observed in Figure 4.6.


91

Figure 4.2. 7Li NMR spectra recorded as a function of temperature for the mole ratio

C221/LiClO4 = 0.5.

Figure 4.3. 7Li NMR spectra recorded as a function of temperature for the mole ratio

C222/LiClO4 = 0.5.

The rate of the exchange reaction increased not only with increasing cavity

size of the cryptand but also with increasing concentration of the chelate, as


92

can be recognized from the shape of the signals at 223 K in Figure 4.4 and

from the values of kobs in Table 4.1 (see Figure 4.5 and Figure 4.6).

Figure 4.4. 7Li NMR spectra of C222 and LiClO4 in acetone. Temperature dependence for

the sample ratios C222/LiClO4 = 0.25, 0.5 and 0.75, respectively.

Figure 4.5. Eyring plots for kobs as a function of temperature for C221 in the temperature range in which exchange was observed for the concentration ratios C221/LiClO4 = 0.5, 0.65 and 0.8, respectively.

Figure 4.6. Eyring plots for kobs as a function of temperature for C222 for the concentration ratios C222/LiClO4 = 0.25, 0.5 and 0.75, respectively.

0.0031 0.0032 0.0033 0.0034-4.5

-4.2

-3.9

-3.6

-3.3

ln (k/T)

1/T, K-10.0030 0.0031 0.0032 0.0033 0.0034

-3.0

-2.7

-2.4

-2.1

-1.8

ln (k/T)

1/T, K-10.0031 0.0032 0.0033

-4.8

-4.4

-4.0

-3.6

ln (k/T)

1/T, K-10.0031 0.0032 0.0033 0.0034

-4.5

-4.2

-3.9

-3.6

-3.3

ln (k/T)

1/T, K-10.0030 0.0031 0.0032 0.0033 0.0034

-3.0

-2.7

-2.4

-2.1

-1.8

ln (k/T)

1/T, K-10.0031 0.0032 0.0033

-4.8

-4.4

-4.0

-3.6

ln (k/T)

1/T, K-1

0.0032 0.0036 0.0040 0.0044-4

-3

-2

-1

0

1

2

ln (k

/T)

1/T, K-10.0032 0.0036 0.0040 0.0044

-1

0

1

2

3

4

ln (k

/T)

1/T, K-1

0.0032 0.0036 0.0040 0.0044-3

-2

-1

0

1

2

3

ln (k

/T)

1/T, K-10.0032 0.0036 0.0040 0.0044

-4

-3

-2

-1

0

1

2

ln (k

/T)

1/T, K-10.0032 0.0036 0.0040 0.0044

-1

0

1

2

3

4

ln (k

/T)

1/T, K-1

0.0032 0.0036 0.0040 0.0044-3

-2

-1

0

1

2

3

ln (k

/T)

1/T, K-1


93

Table 4.1. Values of the observed rate constants (kobs) and activation parameters calculated from the line-shape analysis for the exchange of Li+ between solvated and chelated sites at 25 °C.

Molar ratio [L]:[Li+] kobs298, s-1 ΔH#, kJ mol-1 ΔS#, J K-1 mol-1

C221

0.5 2.10 ± 0.04 41 ± 3 -102 ± 12

0.65 4.5 ± 0.1 37 ± 1 -109 ± 3

0.8 19.4 ± 0.8 31 ± 1 -116 ± 4

C222

0.25 1001 ± 20 33.3 ± 0.5 -76 ± 1

0.5 3263 ± 98 34.1 ± 0.4 -64 ± 1

0.75 6346 ± 254 29 ± 1 -75 ± 2

4.5.2. Suggested Exchange Mechanism

Two mechanisms should be considered for the decomplexation processes of

cryptand-alkali-metal complexes: unimolecular (dissociative) and bimolecular

(associative) cation exchange. The solvent plays a major role in the

unimolecular mechanism and this mechanism is therefore favored in strong

donor solvents.40

Lincoln et al.30-37 investigated many similar systems where the Li+ cation

was exchanged between different cryptands or crown ethers and common

solvents. In case of crown ethers and cryptands in non-aqueous solutions, the

rates of formation (kc) for their interaction with alkali-metal cations are

generally diffusion controlled and, consequently, the complexation selectivities

are governed by the decomplexation rates (kd). They proposed the operation of

a monomolecular mechanism for the decomplexation of Li+ from the cryptand,

notwithstanding the fact that decomplexation is characterized by very negative

activation entropies.


94

According to Popov et al.,28 there are two possible mechanisms that need to

be considered for the exchange of Li+ cation between the solvated and the

complexed sites, viz. bimolecular (associative) (I) and unimolecular

(dissociative) (back reaction in II) cation exchange processes, where S presents

solvent and L the employed chelate.28, 78 (In this description we have added the

terms ‘associative’ and ‘dissociative’ because this is implied by the mechanistic

interpretation offered by the authors).

At equilibrium

]Li(S)][Li(L)[*]Li(L)][Li(S)[* 11 −= kk

]S][Li(L)[]L][Li(S)[ 22 −= kk

Therefore see Equation (4.1) and Equation (4.2)

]S][Li(L)[]Li(L)][Li(S)[]Li(S)[]Li(S)[ 21 −+==− kkkdtd

obs (4.1)

and

]Li(S)[

]S][Li(L)[]Li(L)[ 2

1−+=

kkkobs (4.2)

An alternative form of Equation (4.2) is Equation (4.3), which was used to fit

the data in Figure 4.7 and Figure 4.9.

]Li(S)['

]Li(L)[1

21 −+= kkkobs (4.3)

k1

k-1

*Li(S)+ + Li(L)+ Li(S)+ + *Li(L)+ (I)

k2

k-2

Li(S)+ + L Li(L)+ + S (II)

k1

k-1

*Li(S)+ + Li(L)+ Li(S)+ + *Li(L)+ (I)

k2

k-2

Li(S)+ + L Li(L)+ + S (II)


95

Plots of kobs/[Li(L)] vs. 1/[Li(S)] were found to be linear with varying slopes

(k-2’ = k-2[S]) and intercepts (k1) (see Figure 4.7 and Figure 4.9). The plots

indicate that both pathways compete with each other and the competition is

controlled by temperature and the nature of the cryptand. In the case of C222

(Figure 4.7), the plots at low temperature exhibit no meaning intercept,

whereas the plots at high temperature exhibit very large intercepts. Thus the

contribution from k1 increases significantly with temperature. In the case of

C221 (Figure 4.9), all the plots exhibit almost no intercept, such that the

exchange process is mainly controlled by k-2. The corresponding Eyring plots

are shown in Figure 4.8 and Figure 4.10, respectively, and the thermal

activation parameters are summarized in Table 4.2.

0 10 20 30 40 50 60 70 80 90 1000.0

2.0x104

4.0x104

6.0x104

1.6x105

2.4x105

3.2x105

4.0x105

k obs

/[Li(L

)], M

-1S

-1

1/[Li(S)], M-1

313 K 298 K 273 K 253 K 243 K 233 K

Figure 4.7. Plots of kobs/[Li(L)] vs. 1/[Li(S)] for C222 as a function of temperature.


96

0.0032 0.0036 0.0040 0.0044-2

0

2

4

6 k1 k-2

ln (k

/T)

1/T, K-1

Figure 4.8. Eyring plots for k1 and k-2 as a function of temperature for C222.

0 50 100 150 200 2500

200

400

600

800

1000

1200 318 K 313 K 308 K 303 K 298 K

k obs

/[Li(L

)], M

-1s-1

1/[Li(S)], M-1

Figure 4.9. Plots of kobs/[Li(L)] vs. 1/[Li(S)] for C221 as a function of temperature.


97

0.00315 0.00320 0.00325 0.00330 0.00335

-4.9

-4.8

-4.7

-4.6

-4.5

-4.4

-4.3

-4.2

-4.1

k-2

ln (k

/T)

1/T, K-1

Figure 4.10. Eyring plot for k-2 as a function of temperature for C221.

Table 4.2. Thermal activation parameters obtained from the values of k1 and k-2 as a function of temperature.

Chelate k1298, M-1 s-1 k-2298, s-1 ∆H#, kJ mol-1 ∆S#, J K-1 mol-1

C221 2.24 ± 0.09 29 ± 2 -141 ± 5

C222 (1.7 ± 0.5) x 103 24 ± 1 -105 ± 4

C222 (6 ± 3) x 104 54 ± 4 +29 ± 15

The activation parameters derived from the temperature dependence of k1

are rather inaccurate due to the large error limits of the intercepts and the

reported activation entropy has no mechanistic meaning. On the basis of the

values of k-2 and the significantly negative activation entropies obtained for

this pathway, the activation parameters support an associative type of

exchange process. k-2 for C222 is almost three orders of magnitude larger than

for C221. Furthermore, the activation parameters are very similar to those

obtained directly from the values of kobs as a function of temperature. This


98

would be in a good agreement with that postulated by Lincoln, who assumed

that only the decomplexation process contributes to the observed reaction

under the given experimental conditions. The negative activation entropies

found for the decomplexation reactions of both C221 and C222 strongly

suggest that attack of the solvent on the complexed Li+ must play an important

role in the decomplexation process, i.e. decomplexation follows an associative

mechanism (see theoretical calculations).

The reported activation parameters are within the range of those reported

by Popov et al. for closely related systems.28 However, it should be pointed out

that they used a different rate law than the one proposed by Shchori et al.79

and the one used in the present study, which complicates a direct comparison

of the reported activation parameters. Nevertheless, although they proposed

another mechanism, the values of the activation parameters obtained from

their study are very similar to ours, as shown in Table 4.3.below.

Table 4.3. Comparison of activation parameters.

kb, s-1 ΔH#, kJ mol-1 ΔS#, J K-1 mol-1

Popov et al.28 794 ± 42 20 ± 1 -121 ± 3

This work 1724 ± 484 24 ± 1 -105 ± 4

It is reasonable to expect that solvents such as acetonitrile, acetone and γ-

butyrolactone, due to their similar properties as expected from their rather

similar Gutmann donor numbers (DN = 14.1, 17.0 and 18.0, respectively,80)

should coordinate to Li+ in a similar way and should have a similar influence

on the exchange mechanism. This observation corresponds to that reported by

Shamsipur et al.81 for the exchange of Li+ between C221 and

acetonitrile/nitromethane mixtures. In all solvent mixtures used, Li+ exchange

was found to occur through an associative mechanism, as expected for weakly

donor solvents (DN = 14.1 and 2.7 for acetonitrile and nitromethane,

respectively).


99

4.5.3. High Pressure 7Li NMR Measurements

Additional mechanistic information on the intimate nature of the exchange

mechanism can in principle be obtained from the effect of pressure on the Li+

exchange process.82-85 A systematic pressure dependence study was performed

to determine the volume of activation for the Li+ exchange process for both

chelates. Pressure-dependent NMR spectra were recorded for each sample

over the pressure range of 0.1 to 150 MPa at 313.2 K. The higher temperature

had to be chosen, since in case of C221 the exchange process hardly occurs at

room temperature, whereas in case of C222 the signals are broad at room

temperature due to coalescence and it was difficult to observe line broadening

that resulted from increasing pressure. There are two signals to be seen in

Figure 4.11: one for the solvated Li at 8.8 ppm and the one for Li complexed by

C221 at 6.5 ppm. It can be observed that there is a very small difference in the

shape of the NMR spectra recorded at various pressures. The same effect was

observed for the second ligand, C222 (Figure 4.12), where only one signal

appears due to the fast exchange. Its shape almost does not change with the

varying pressure for any of the three samples. The rate constants calculated for

each pressure were very similar within the experimental error limits and

resulted in a practically zero value for ∆V#obs for both ligands, which points to a

concerted interchange type of exchange mechanism.86 In terms of an

interchange mechanism, bond formation and bond cleavage occur in a

concerted fashion without the formation of a distinct intermediate of higher or

lower coordination number. In addition, decomplexation will be accompanied

by charge creation due to the release of Li+, which in turn will lead to an

increase in electrostriction which will offset the intrinsic volume changes.


100

Figure 4.11. 7Li NMR spectra recorded as a function of pressure at 313.2 K for the mole ratio C221/LiClO4 = 0.5.

Figure 4.12. 7Li NMR spectra of C222 and LiClO4 in acetone. Pressure dependence for the sample ratios C222/LiClO4 = 0.25, 0.5, and 0.75, respectively, at 313.2 K.

4.5.4. Theoretical Calculations

For the closely related γ-butyrolactone and Li+/γ-butyrolactone high level

ab initio studies were recently performed by Rey et al..87 For acetone and

Li+/acetone only Hartree-Fock calculations were performed.65 In our DFT

calculations we found that the Li+ cation coordinates to the carbonyl oxygen

with an almost co-linear configuration relative to the carbon-oxygen bond, as

for γ-butyrolactone. This configuration holds for clusters of up to four


101

molecules and also in the solid phase, where a tetrahedral first solvation

(coordination) sphere was found (see Table 4.4).66

Table 4.4. Calculated bond length for Li+ solvated in actone.

Bond length [Å] PG B3LYP/D95Vp B3LYP/6-311+G**

[Li(acetone)]+ C2 1.77 1.75

[Li(acetone)2]+ D2 1.81 1.79

[Li(acetone)3]+ D3 1.88 1.86

[Li(acetone)4]+ C1 1.96 (averaged) 1.96 (averaged)

Experimentally, a decrease in the cavity size of the cryptands hampers the

exchange of Li+ between the cryptand and the solvent. We were able to observe

it in the case of γ-butyrolactone and again we can see it in the present case. The

improved binding of Li+ in a smaller cavity is visible from the structure and the

following model reaction [Equation (4.4)] that describes the energy required

for the dechelation process. Application of the CPCM solvent model

strengthens this trend significantly (see Table 4.5).

[Li ⊂ Cryptand]+ + 4 solvent [Li(solvent)4]+ + Cryptand (4.4)

In agreement with previous studies by Wipff et al.,5 we also found only a

limited flexibility for the investigated cryptands in our DFT calculations,

therefore all cryptands were calculated within the highest possible symmetry

(Table 4.5). We recently demonstrated that reaction (4.4) is a valuable tool to

determine ion selectivity as a function of the ionic radii88 and to examine cavity

size of the cryptands as a variable.42


102

Table 4.5. Dechelation energy, Edechelation, for the reaction outlined in Equation (4.4).[a]

Acetone C222 C221 C211

Edechelation (DFT) [kcal/mol] -8.6 +2.0 +3.0

Edechelation (CPCM) [kcal/mol] +5.4 +13.6 +20.4

γ-Butyrolactone C222 C221 C211

Edechelation (DFT) [kcal/mol]42 -16.8 -11.6 -5.3

[a] DFT: B3LYP/D95Vp, CPCM: B3LYP(CPCM)/D95Vp//B3LYP/D95Vp.

Table 4.6. Calculated (RB3LYP/D95Vp) and X-ray bond lengths (in Å).

d [Å] C222 C221 C211 [Li(OH2)4]+ [Li(acetone)4]+ [Li(NH3)4]+

Li-O

2.14

2.06

-

2.08

2.10

2.14

2.11

1.95

1.96

-.--

Li-N 2.22 2.28 2.38 -.-- -.-- 2.13

PG C3 C1 C2 T C1 C1

X-ray 89 90 91 92 93

Li-O

-.--

1.99

2.06

2.10

2.17

2.08

1.96

1.94, 1.91, 1.93,

1.95

-.--

Li-N -.-- 2.40

2.44

2.29 -.-- -.-- 2.08

In agreement with the experimental kinetic data, the smaller cryptand

binds Li+ better and slows down its exchange rate (Table 4.2). On comparing

the calculated Li-O and Li-N bond lengths of the three structures with the

bond length in [Li(OH2)4]+ and [Li(NH3)4]+, the trend can easily be

understood (see Table 4.6 and Figure 4.11). In C222, Li+ is four-fold weakly


103

coordinated in a trigonal-pyramidal shape by one Li-N and three Li-O

interactions all clearly elongated (0.2 and 0.1 Å, respectively). The Li+ cation in

C221 is five-fold coordinated (distorted trigonal-bipyramidal structure) by one

nitrogen and four oxygen bonds all extended by around 0.15 Å. The

coordinating bonds in [Li ⊂ C211]+ are even more extended, but the Li-O bonds

form a somewhat distorted tetrahedral coordination sphere for the lithium

monocation, while the stronger elongated Li-N bonds form a kind of second

coordination sphere. Therefore, the lithium ion can formally adopt a six-

coordinate state, which strongly supports the proposed associative character of

the decomplexation process that involves binding of external solvent molecules

as suggested by the very negative activation entropies. A comparison of the

calculated and X-ray bond lengths shows a good agreement. A comparison

between the calculated data for γ-butyrolactone and acetone shows the same

trend for both solvents. Li+ is better coordinated in the smaller cavities of the

smaller cryptands, as expected.

To draw a conclusion on solvation abilities from a direct comparison of the

dechelating energies in acetone and γ-butyrolactone, is risky. The calculated

dechelating energies can not be considered as a proof of the rule of Madrakian

et al., that there is an inverse relationship between the stability of the

complexes and the solvating ability of the solvent.66 The different values can be

attributed to the different number of non-hydrogen atoms in both solvents.

Acetone has only four non-hydrogen atoms, while γ-butyrolactone has six and

can therefore stabilize charge much better.


104

Figure 4.13. Optimized structures for [Li ⊂ C222]+, [Li ⊂ C221]+, [Li ⊂ C211]+ and

[Li(Acetone)4]+.

4.6. Conclusions

It can be concluded from the results of the present study that for the Li+

exchange process between C222, C221 and acetone as solvent, two pathways,

viz. bimolecular (associative) and unimolecular (dissociative) compete with

each other. The pressure dependence of the exchange process suggests the

operation of a pure interchange mechanism, whereas the negative activation

entropy values favor an associative interchange process for the decomplexation

reaction. The basic idea of an interchange process involves concerted bond


105

formation and bond cleavage, which is in agreement with the consideration

that acetone is a rather weak donor. The results are in good agreement with

that expected on the basis of a systematic comparison with data reported in the

literature for solvents of similar donor strength.


106

4.7. References

1. Dietrich, B.; Lehn, J. M.; Sauvage, J. P. Tetrahedron Lett. 1969, 34,

2889.


2885.

3. Lehn, J. M. Acc. Chem. Res. 1978, 11, 49.

4. Lehn, J. M. Angew. Chem. 1988, 100, 91.

5. Wipff, G.; Wurtz, J. M. New J. Chem. 1989, 13, 807.

6. Zhang, X. X.; Izatt, R. M.; Krakowiak, K. E.; Bradshaw, J. S. Inorg.

Chim. Acta 1997, 254, 43.

7. Lehn, J. M.; Montavon, F. Helv. Chim. Acta 1976, 59, 1566.

8. Puchta, R.; Seitz, V.; van Eikema Hommes, N. J. R.; Saalfrank, R. W. J.

Mol. Model. 2000, 6, 126.

9. Saalfrank, R. W.; Dresel, A.; Seitz, V.; Trummer, S.; Hampel, F.;

Teichert, M.; Stalke, D.; Stadler, C.; Daub, J.; Schunemann, V.;

Trautwein, A. X. Chem.-Eur. J. 1997, 3, 2058.

10. Saalfrank, R. W.; Seitz, V.; Caulder, D. L.; Raymond, K. N.; Teichert, M.;

Stalke, D. Eur. J. Inorg. Chem. 1998, 9, 1313.

11. Saalfrank, R. W.; Seitz, V.; Heinemann, F. W.; Gobel, C.; Herbst-Irmer,

R. J. Chem. Soc., Dalton Trans. 2001, 5, 599.

12. Dietrich, B.; Kintzinger, J. P.; Lehn, J. M.; Metz, B.; Zahidi, A. J. Phys.

Chem. 1987, 91, 6600.

13. Lindoy, L. F. The Chemistry of Macrocyclic Ligand Complexes;

Cambridge University Press: Cambridge, 1992; p 269.

14. Farahbakhsh, M.; Schmidt, H.; Rehder, D. Chem. Ber. Recl. 1997, 130,

1123.

15. Housecroft, C. E. Cluster Compounds of Main Group Elements; VCH:

Weinheim, Germany, 1996; p 94.

16. Kirch, M.; Lehn, J. M. Angew. Chem. 1975, 87, 542.


18. Lehn, J. M.; Sauvage, J. P. J. Chem. Soc., Chem. Commun. 1971, 9, 440.


107

19. Morf, W. E.; Simon, W. Helv. Chim. Acta 1971, 54, 2683.

20. Pregel, M. J.; Buncel, E. J. Am. Chem. Soc. 1993, 115, 10.

21. Simon, W.; Morf, W. E.; Meier, P. C. Struct. Bond. 1973, 16, 113.

22. Severin, K. Coord. Chem. Rev. 2003, 245, 3.

23. Huang, R. H.; Wagner, M. J.; Gilbert, D. J.; Reidy-Cedergren, K. A.;

Ward, D. L.; Faber, M. K.; Dye, J. L. J. Am. Chem. Soc. 1997, 119, 3765.


25. Wipff, G.; Kollman, P. Nouv. J. Chim. 1985, 9, 457.

26. Cox, B. G.; Garcia-Rosas, J.; Schneider, H. J. Am. Chem. Soc. 1981, 103,

1054.

27. Rhinebarger, R. R.; Popov, A. I. Polyhedron 1988, 7, 1341.

28. Shamsipur, M.; Popov, A. I. J. Phys. Chem. 1986, 90, 5997.

29. Smetana, A. J.; Popov, A. I. J. Solution Chem. 1980, 9, 183.

30. Abou-Hamdan, A.; Lincoln, S. F. Inorg. Chem. 1991, 30, 462.

31. Clarke, P.; Gulbis, J. M.; Lincoln, S. F.; Tiekink, E. R. T. Inorg. Chem.

1992, 31, 3398.

32. Clarke, P.; Lincoln, S. F.; Tiekink, E. R. T. Inorg. Chem. 1991, 30, 2747.

33. Lincoln, S. F.; Abou-Hamdan, A. Inorg. Chem. 1990, 29, 3584.

34. Lincoln, S. F.; Rodopoulos, T. Inorg. Chim. Acta 1991, 190, 223.


36. Lincoln, S. F.; Stephens, A. K. W. Inorg. Chem. 1991, 30, 3529.

37. Stephens, A. K. W.; Dhillon, R. S.; Madbak, S. E.; Whitbread, S. L.;

Lincoln, S. F. Inorg. Chem. 1996, 35, 2019.


J.; Sen, D. Chem. Rev. 1985, 85, 271.

39. Izatt, R. M.; Bradshaw, J. S.; Pawlak, K.; Bruening, R. L.; Tarbet, B. J.

Chem. Rev. 1992, 92, 1261.


1991, 91, 1721.


Reviews 1995, 95, 2529.


108

42. Pasgreta, E.; Puchta, R.; Galle, M.; van Eikema Hommes, N.; van Eldik,

R. J. Inclusion Phenom. Macrocyclic Chem. 2007, 58, 81.

43. Sonoda, K.; Deguchi, M.; Koshina, H.; Armand, M.; Michot, C.;

Gauthier, M. Proc. - Electrochem. Soc. 2003, 2002-26, 512.

44. Takami, N.; Sekino, M.; Ohsaki, T.; Kanda, M.; Yamamoto, M. J. Power

Sources 2001, 97-98, 677.

45. Handy, S. T.; Grieco, P. A.; Mineur, C.; Ghosez, L. Synlett 1995, Spec.

Issue, 565.

46. Kinart, W. J.; Sniec, E.; Tylak, I.; Kinart, C. M. Phys. Chem. Liq. 2000,

38, 193.

47. Barthel, J.; Buchner, R.; Wismeth, E. J. Solution Chem. 2000, 29, 937.

48. Barthel, J.; Gores, H. J.; Neueder, R.; Schmid, A. Pure Appl. Chem.

1999, 71, 1705.

49. Croce, F.; D'Aprano, A.; Nanjundiah, C.; Koch, V. R.; Walker, C. W.;

Salomon, M. J. Electrochem. Soc. 1996, 143, 154.

50. Matsubara, K.; Kaneuchi, R.; Maekita, N. J. Chem. Soc., Faraday Trans.

1998, 94, 3601.


52. Himeno, S.; Takamoto, M.; Ueda, T.; Santo, R.; Ichimura, A.

Electroanalysis 2004, 16, 656.

53. Karkhaneei, E.; Zebrajadian, M. H.; Shamsipur, M. J. Inclusion Phenom.

Macrocyclic Chem. 2001, 40, 309.

54. Kumar, G.; Janakiraman; Namboodiri, N.; Gangadharan; Saminathan;

Kanagaraj; Sharief, N. J. Chem. Eng. Data 1991, 36, 467.

55. Hojo, M.; Hasegawa, H.; Hiura, N. J. Phys. Chem. 1996, 100, 891.


R. Inorg. Chem. 2004, 43, 8227.

57. Spångberg, D.; Hermansson, K. Chem. Phys. 2004, 300, 165.

58. Alia, J. M.; Diaz de Mera, Y.; Edwards, H. G. M.; Garcia, F. J.; Lawson,

E. E. J. Mol. Struct. 1997, 408-409, 439.

59. Chabanel, M.; Legoff, D.; Touaj, K. J. Chem. Soc., Faraday Trans. 1996,

92, 4199.


109

60. Hojo, M.; Hasegawa, H.; Morimoto, Y. Anal. Sci. 1996, 12, 521.

61. Barthel, J.; Neueder, R.; Poepke, H.; Wittmann, H. J. Solution Chem.

1999, 28, 489.

62. Kiselev, V. D.; Kashaeva, E. A.; Luzanova, N. A.; Konovalov, A. I.

Thermochim. Acta 1997, 303, 225.

63. Kameda, Y.; Kudoh, N.; Suzuki, S.; Usuki, T.; Uemura, O. Bull. Chem.

Soc. Jpn. 2001, 74, 1009.

64. Blint, R. J. J. Electrochem. Soc. 1995, 142, 696.

65. Jarek, R. L.; Miles, T. D.; Trester, M. L.; Denson, S. C.; Shin, S. K. J.

Phys. Chem. A 2000, 104, 2230.

66. Madrakian, T.; Afkhami, A.; Ghasemi, J.; Shamsipur, M. Polyhedron

1996, 15, 3647.

67. Zahl, A.; Neubrand, A.; Aygen, S.; van Eldik, R. Rev. Sci. Instrum. 1994,

65, 882.


69. Matlab, version 5.3.1.; Mathworks Inc.: 1999.





Phys. 1988, 37, 785.


Chem. 1994, 98, 11623.


New York, 1976; Vol. 3, p 1.




76. The performance of the computational level employed in this study is

well documented, see for example: Galle, M.; Puchta, R.; van Eikema

Hommes, N.; van Eldik, R. Z. Phys. Chem. 2006, 220, 511; Puchta, R.;

van Eikema Hommes, N. J. R.; Meier, R.; van Eldik, R. Dalton Trans.


110

2006, 28, 3392; Klaus, S.; Neumann, H.; Jiao, H.; Jacobi von Wangelin,

A.; Goerdes, D.; Struebing, D.; Huebner, S.; Hateley, M.; Weckbecker,

C.; Huthmacher, K.; Riermeier, T.; Beller, M. J. Organomet. Chem.

2004, 689, 3685; Saalfrank, R. W.; Deutscher, C.; Maid, H.; Ako, A. M.;

Sperner, S.; Nakajima, T.; Bauer, W.; Hampel, F.; Hess, B. A.; van

Eikema Hommes, N. J. R.; Puchta, R.; Heinemann, F. W. Chem.-Eur. J.

2004, 10, 1899; Scheurer, A.; Maid, H.; Hampel, F.; Saalfrank, R. W.;

Toupet, L.; Mosset, P.; Puchta, R.; van Eikema Hommes, N. J. R. Eur. J.

Org. Chem. 2005, 12, 2566; Weber, C. F.; Puchta, R.; van Eikema

Hommes, N. J. R.; Wasserscheid, P.; van Eldik, R. Angew. Chem. 2005,

117, 6187; Weber, C. F.; Puchta, R.; van Eikema Hommes, N. J. R.;

Wasserscheid, P.; van Eldik, R. Angew. Chem. (Int. Ed. Eng.) 2005, 44,

6033.


















78. Shchori, E.; Jagur-Grodzinski, J.; Shporer, M. J. Am. Chem. Soc. 1973,

95, 3842.


111

79. Shchori, E.; Jagur-Grodzinski, J.; Luz, Z.; Shporer, M. J. Am. Chem. Soc.

1971, 93, 7133.



23.

82. Davis, A. V.; Fiedler, D.; Seeber, G.; Zahl, A.; Van Eldik, R.; Raymond, K.

N. J. Am. Chem. Soc. 2006, 128, 1324.

83. van Eldik, R.; Hubbard, C. D. Chemistry at Extreme Conditions;

Elsevier: Amsterdam, 2005; p 109.

84. Wolak, M.; van Eldik, R. J. Am. Chem. Soc. 2005, 127, 13312.

85. Zahl, A.; Van Eldik, R.; Matsumoto, M.; Swaddle, T. W. Inorg. Chem.

2003, 42, 3718.


87. Masia, M.; Rey, R. J. Phys. Chem. B 2004, 108, 17992.

88. Galle, M.; Puchta, R.; van Eikema Hommes, N. J. R.; van Eldik, R. Z.

Phys. Chem. 2006, 220, 511.

89. Choua, S.; Sidorenkova, H.; Berclaz, T.; Geoffroy, M.; Rosa, P.;

Mezailles, N.; Ricard, L.; Mathey, F.; Le Floch, P. J. Am. Chem. Soc.

2000, 122, 12227.

90. Moras, D.; Weiss, R. Acta Crystallogr.,Sect. B: Struct. Crystallogr.

Cryst. Chem. 1973, 29, 400.

91. Meske, W.; Babel, D. Z. Anorg. Allg. Chem. 1998, 624, 1751.

92. Cotton, F. A.; Daniels, L. M.; Murillo, C. A.; Zhou, H. C. Inorg. Chim.

Acta 2000, 300-302, 319.

93. Scotti, N.; Zachwieja, U.; Jacobs, H. Z. Anorg. Allg. Chem. 1997, 623,

1503.

5. Complexation by Cryptands in γ-Butyrolactone as Solvent

112

5. Ligand exchange processes on solvated

lithium cations. Complexation by Cryptands

in γ-Butyrolactone as Solvent

5.1. General remark



Pasgreta, E.; Puchta, R.; Galle, M.; van Eikema-Hommes N.; Zahl A.; van Eldik

R. Journal of Inclusion Phenomena and Macrocyclic Chemistry 2007, 58, 81-

88.

5.2. Abstract

Kinetic studies on Li+ exchange between the cryptands C222 and C221 and

γ-butyrolactone as solvent were performed as a function of ligand-to-metal

ratio, temperature and pressure using 7Li NMR. The thermal rate and

activation parameters are: C222: k298 = (3.3 ± 0.8) x 104 M-1 s-1, ∆H# = 35 ± 1

kJ mol-1 and ∆S# = -41 ± 3 J K-1 mol-1; C221: k298 = 105 ± 32 M-1 s-1, ∆H# = 48 ±

1 kJ mol-1 and ∆S# = -45 ± 2 J K-1 mol-1. Temperature and pressure dependence

measurements were performed in the presence of an excess of Li+. The

influence of pressure on the exchange rate is insignificant for both ligands,

such that the value of activation volume is around zero within the experimental

error limits. The activation parameters obtained in this study indicate that the

exchange of Li+ between solvated and chelated Li+ ions follows an associative

interchange mechanism.


113

5.3. Introduction

Following the work of Pedersen on two-dimensional crowns, Lehn and

coworkers developed in 1969 a range of three-dimensional, polycyclic ligand

systems which they named cryptands (Greek: cryptos = cave).1, 2 Until today

this class of compounds has been intensively explored and extended in

different directions3, 4 in terms of cavity size,5, 6 donor atoms,7 and metalla-

topomers.8-11 Since cryptands readily form complexes with a wide range of

metal ions and NH4+,12 provided the ion involved is not too large to be

contained in the macrocyclic cavity,13 they find their application in various

areas like catalysis, medicine, biomimetic research,14-21 and as receptors22 and

electrides.23 Studies on the complexes (cryptates) formed between alkali metal

ions and cryptands, have revealed a substantial understanding of the effects of

metal ion size and cryptand cavity size on the structure, stability and lability of

cryptates, and the mechanism of the cryptate complex-formation processes.24-

28 Such selective complexation has led to many biological and chemical studies.

Kinetic studies on macrocyclic complex-formation reactions with alkali

cations not only result in important information on the rates and mechanisms

of complex-formation reactions, but also lead to a better understanding of the

high selectivity of these ligands toward different cations. The broad impact of

such studies spans from the theory of complexation to cation transport

through membranes, and makes the ability to control cation selectivity via

tactical structural changes in the complexing agent of prime importance.18

Polyoxo macrobicyclic diamines, represented by the structure in Figure 5.1, are

able to form metal ion complexes in which the metal ion is located in the cavity

of the macromolecule and two nitrogen atoms along with the oxygen atoms

participate in binding the metal cation. All complexes have 1:1 stoichiometry

with the metal ion positioned in the center of the ligand cavity.24


114

O

N

OO

O

N O O

m

m

n Figure 5.1. Chelate structures. C211: m = 0, n = 1. C221: m = 1, n = 0. C222: m = 1, n = 1.

In case of crown ethers and cryptands in non-aqueous solutions, the rates

of complex-formation (kf) for the interaction with alkali-metal cations are

generally diffusion-controlled and, consequently, the complexation

selectivities are governed by the decomplexation rates kd.29 These reactions

have received significant attention, initially from Lehn and co-workers and

more recently from Popov et al.30-32 and Lincoln et al..33-40 These and other

groups performed a series of detailed kinetic studies and found that two

mechanisms need to be considered for the decomplexation process of

cryptand-alkali metal cation complexes, viz., bimolecular associative (I) and

unimolecular dissociative (back reaction in II) cation exchange processes,

where S presents solvent and L the employed chelate.41, 42

The solvent plays a major role in the unimolecular mechanism II,43, 44

which is therefore favored in strong donor solvents. Many examples have been

included in a series of review articles.41, 45-47 It has been reported that ligand

structure and solvent properties have considerable effects on the reaction

rates, activation parameters, and exchange mechanism of cations between the

k1

k-1

*Li(S)+ + Li(L)+ Li(S)+ + *Li(L)+ (I)

k2

k-2

Li(S)+ + L Li(L)+ + S (II)

k1

k-1

*Li(S)+ + Li(L)+ Li(S)+ + *Li(L)+ (I)

k2

k-2

Li(S)+ + L Li(L)+ + S (II)


115

solvated and complexed sites. We have now extended the study of Li+ exchange

to γ-butyrolactone (see Figure 5.2) as solvent.

O

O

Figure 5.2. Structure of γ-butyrolactone.

γ-Butyrolactone was chosen as solvent due to its relevance in lithium

batteries and capacitors.48, 49 It is a model cyclic ester used as a major chemical

compound with extensive application in pharmaceuticals, pesticides and

petrochemicals. The butyrolactone ring is an integral building block of many

natural products of biological activity,50, 51 like the sesquiterpene lactones,

flavor components, alkaloids, antileukemics, and pheromones. γ-Butyrolactone

is one of the important intermediates in fine chemical industrial practices, for

example in the synthesis of pyrrolidine, herbicides and rubber additives.52 It is

also used as a solvent for surface treatment of textiles and metal coated

plastics, as a paint remover, in photochemical etching, in vitamin and

pharmaceutical preparations.

Recently, computational studies on γ-butyrolactone and Li+/γ-

butyrolactone have been performed by Rey et al..53 It was found that the Li+

cation coordinates to the carbonyl oxygen with an almost co-linear

configuration relative to the carbon-oxygen bond but with a slight tilting

toward the lactone oxygen. This configuration holds for clusters of up to four

molecules and also in the liquid phase, where a tetrahedral first solvation

(coordination) sphere was found.

In this contribution, we report kinetic parameters for the exchange of Li+ in

γ-butyrolactone under ambient and elevated pressure conditions, and discuss

their mechanistic implications.


116



The preparation of test solutions was carried out under an Ar or N2

atmosphere using standard Schlenk techniques, under an Ar atmosphere in a

glove box, or under vacuum. Lithium perchlorate (Aldrich, battery grade) was

vacuum-dried at 120 °C for 48 h and then stored under dry nitrogen. The

cryptands C222 (purchased from Fluka), C221 and C211 (purchased from

Aldrich) were dried under vacuum for 48 h and stored under dry nitrogen. γ-

Butyrolactone (Aldrich) was dried over anhydrous CaSO4, then fractionally

destilled under vacuum and stored under nitrogen over Linde 3Å molecular

sieves. Solutions of LiClO4, C222, C221 and C211 in γ-butyrolactone were

prepared under dry nitrogen. For the 7Li NMR measurements the solutions

were sealed in 5 mm NMR tubes under an Ar atmosphere. In order to study the

kinetics of the exchange of Li+ between C222 and C221 and γ-butyrolactone as

solvent, four solutions of different concentration ratio were prepared for each

system. The concentration of lithium perchlorate in each solution was kept

constant at 0.05 M, whereas the concentration of ligand was varied between

0.0175 M and 0.04 M. In this way the measurements were always performed in

the presence of an excess of Li+.


Variable-temperature and variable-pressure Fourier transform 7Li NMR

spectra were recorded at a frequency of 155 MHz on a Bruker Avance DRX

400WB spectrometer equipped with a superconducting BC-94/89 magnet

system. 5 mm spinning tubes were used with a 1 mm o.d. melting point

capillary inserted coaxially in the spinning tube and filled with an external

reference solution (usually 1 M LiClO4 in DMF). The temperature dependence

measurements were performed over a wide as possible temperature range for

each system. A homemade high-pressure probe described in the literature 54

was used for the variable-pressure experiments which were conducted at a

selected temperature and at ambient, 30, 60, 90, 120 and 150 MPa pressure. A


117

standard 5 mm NMR tube cut to a length of 45 mm was used for the sample

solution. The high-pressure NMR tube was sealed with a MACOR plug and an

O-ring made of VITON. The plug and O-ring were shown to be stable in the

employed solvent. The pressure was transmitted by the movable MACOR

piston to the sample, and the temperature was controlled as described

elsewhere.54 For a typical kinetic run, 0.8 ml of solution was transferred under

an Ar atmosphere to a Wilmad screw-cap NMR tube equipped with a

poly(tetrafluoroethylene) septum. All test samples were prepared in the same

way.

5.4.3. Programs

The line widths of the NMR signals were obtained by fitting Lorentzian

functions to the experimental spectra using the NMRICMA 2.7 program55 for

Matlab.56 The adjustable parameters were the resonance frequency, intensity,

line width and baseline. Complete line-shape analysis based on the Kubo-Sack

formalism using modified Bloch equations57 was also performed with

NMRICMA 2.7 to extract rate constants from experimental spectra. The

temperature and pressure dependent 7Li line widths and chemical shifts

employed in the line shape analysis were extrapolated from low temperatures

where no exchange-induced modification occurred.


All structures were optimized at the B3LYP/D95Vp hybrid DFT level 58-60 61,

62 with no constrains other than symmetry and confirmed to be minima by

computation of vibrational frequencies. Relative energies were corrected for

zero point vibrational energy differences. The D95V basis set, augmented with

polarisation functions (D95Vp), was used.63 The Gaussian suite of programs

was used throughout.64


118


5.5.1. Spectroscopic Observations

In preliminary experiments, a series of 7Li NMR measurements for the Li+-

cryptand system in γ-butyrolactone as solvent were performed. For each of the

cryptands C222, C221 and C211, mole ratio dependent measurements were

performed by variation of both the cryptand:Li+ and Li+:cryptand molar ratios.

The results enabled us to observe the distribution of particular species on

increasing the amount of chelate or lithium salt. Basically, two signals of

varying intensity were observed in the 7Li NMR spectra that refer to solvated

Li+ and Li+ bound inside the cryptand cavity. In case of the smallest chelate

C211, however, an additional signal appeared which can be ascribed to Li+

bound outside the cryptand and being partially solvated. The distribution plot

for the various species in a solution of C211 and LiClO4 in γ-butyrolactone is

presented in Figure 5.3. The concentration of LiClO4 was kept constant at 0.01

M and the concentration of C211 was varied in the range 0 to 0.04 M. Initially

only the signal of free (solvated) Li+ is observed, and then with increasing

amount of C211, the signals referring to the complexed Li+ appear. Contrary to

the larger cryptands, there are two signals that refer to bound Li+; one for Li+

inside the cavity and another of low intensity (max. 15 %) for Li+ presumably

bound outside the cryptand and being partially solvated. The latter signal,

however, disappears once the C211/LiClO4 molar ratio gets to 1, and then all

the Li+ is bound as internal complex. Due to the facts that the exchange

process among the three species cannot be clearly defined, Li(C211)+ is very

stable and there is almost no exchange taking place, a detailed study of the

observed rate constant, kobs, for the exchange of Li+ as a function of

temperature and pressure was undertaken only for C221 and C222. The

lithium exchange reaction between cryptand and γ-butyrolactone was studied

for four solutions with different ligand-to-metal ratios in the presence of an

excess of Li+.


119

0.00 0.01 0.02 0.03 0.04

0

20

40

60

80

100

free lithium internal complex external complex

%

[C211], M

Figure 5.3. Distribution plot for various species in a solution of C211 and LiClO4 in γ-

butyrolactone.

Three signals can be observed in Figure 5.4; one comes from the reference

(6.6 ppm) and the other two from the exchanging sites (6.0 ppm solvated Li+

and 4.8 ppm complexed Li+). At low temperature the signals are narrow and

completely separated due to very slow or no exchange. Since the formation

constant for Li(C221)+ is very high,24 the concentration of bound lithium is

equal to the amount of cryptand added. At ca. 358 K the signals start to

coalescence and at 388 K only one broad signal (at 5.3 ppm) can be observed

due to fast exchange. In the case of C222 (Figure 5.5), the exchange process is

faster due to the larger cavity of the cryptand. The signals that refer to solvated

and complexed Li cation (8.3 ppm and 6.8 ppm, respectively at 218 K)

coalescence at a lower temperature (ca. 268 K), and at high temperature (368

K) only one sharp signal at 7.7 ppm was observed as a consequence of a very

fast exchange process. An accurate temperature dependence for C222 over a

large temperature range can be observed in Figure 5.7.


120

Figure 5.4. 7Li NMR spectra recorded as a function of temperature for the mole ratio C221:LiClO4 = 0.65.

Figure 5.5. 7Li NMR spectra recorded as a function of temperature for the mole ratio C222:LiClO4 = 0.65.

The rate of the exchange reaction increased not only with increasing cavity

size of the cryptand but also with increasing concentration of the chelate, as

can be recognized from the shape of the signals at 338 K in Figure 5.6 and from

the values of kobs in Table 5.1.


121

Figure 5.6. 7Li NMR spectra of C222 and LiClO4 in GBL. Temperature dependence for the sample ratios C222:LiClO4 = 0.35, 0.45, 0.55 and 0.65.

Figure 5.7. Eyring plots for kobs as a function of temperature for C222 for the concentration ratios C222:LiClO4 = 0.35, 0.45, 0.55 and 0.65, respectively.

6.57.07.58.0(ppm)

6.57.07.58.08.5(ppm)

6.57.07.58.08.5(ppm)

6.57.07.58.08.5(ppm)

248 K

268 K

283 K

298 K

318 K

338 K

6.57.07.58.0(ppm)

6.57.07.58.0 6.57.07.58.0(ppm)

6.57.07.58.08.5(ppm)

6.57.07.58.08.5 6.57.07.58.08.5(ppm)

6.57.07.58.08.5(ppm)

6.57.07.58.08.5 6.57.07.58.08.5(ppm)

6.57.07.58.08.5(ppm)

6.57.07.58.08.5 6.57.07.58.08.5(ppm)

248 K

268 K

283 K

298 K

318 K

338 K

0,0026 0,0028 0,0030 0,0032 0,0034 0,0036 0,0038 0,0040 0,0042 0,0044-4

-2

0

2

4 Y = A + B * X

Parameter Value Error------------------------------------------------------------A 14,22877 0,21476B -4047,99144 63,1407-----------------------------------------------------------

ln k

/T

1/T0,0026 0,0028 0,0030 0,0032 0,0034 0,0036 0,0038 0,0040 0,0042

-2

0

2

4

Y = A + B * X

Parameter Value Error------------------------------------------------------------A 16,71606 0,11624B -4674,27696 34,83709----------------------------------------------------------

ln k

/T

1/T

0,0026 0,0028 0,0030 0,0032 0,0034 0,0036 0,0038 0,0040 0,0042

-2

0

2

4

Y = A + B * X

Parameter Value Error------------------------------------------------------------A 16,78361 0,19905B -4604,44033 59,65542------------------------------------------------------------

ln k

/T

1/T0,0026 0,0028 0,0030 0,0032 0,0034 0,0036 0,0038 0,0040 0,0042 0,0044

-2

0

2

4

Y = A + B * X


ln k

/T

1/T

0,0026 0,0028 0,0030 0,0032 0,0034 0,0036 0,0038 0,0040 0,0042 0,0044-4

-2

0

2

4 Y = A + B * X

Parameter Value Error------------------------------------------------------------A 14,22877 0,21476B -4047,99144 63,1407-----------------------------------------------------------

ln k

/T

1/T0,0026 0,0028 0,0030 0,0032 0,0034 0,0036 0,0038 0,0040 0,0042

-2

0

2

4

Y = A + B * X

Parameter Value Error------------------------------------------------------------A 16,71606 0,11624B -4674,27696 34,83709----------------------------------------------------------

ln k

/T

1/T

0,0026 0,0028 0,0030 0,0032 0,0034 0,0036 0,0038 0,0040 0,0042

-2

0

2

4

Y = A + B * X


ln k

/T

1/T0,0026 0,0028 0,0030 0,0032 0,0034 0,0036 0,0038 0,0040 0,0042 0,0044

-2

0

2

4

Y = A + B * X


ln k

/T

1/T


122

Table 5.1. Values of the observed rate constants (kobs) calculated from the line shape analysis for the exchange of Li+ between solvated and chelated sites at 25°C.

Mole ratio C221:Li+ kobs, s-1 Mole ratio C222:Li+ kobs, s-1

0.4 2.68 ± 0.05 0.35 654 ± 13

0.5 4.3 ± 0.1 0.45 851 ± 25

0.65 6.9 ± 0.3 0.55 1282 ± 51

0.8 9.6 ± 0.5 0.65 1352 ± 68

5.5.2. Suggested Exchange Mechanism

As mentioned above, there are two possible mechanisms for the exchange

of the lithium ion between the solvated and the complexed sites, namely

bimolecular associative (I) and unimolecular dissociative (II) mechanisms.

At equilibrium

]Li(S)][Li(L)[*]Li(L)][Li(S)[* 11 −= kk

]S][Li(L)[]L][Li(S)[ 22 −= kk

Therefore

]S][Li(L)[]Li(L)][Li(S)[]Li(S)[]Li(S)[ 21 −+==− kkkdtd

obs (5.1)

and

]Li(S)[

]S][Li(L)[]Li(L)[ 2

1−+=

kkkobs (5.2)

An alternative form of Equation (5.2) is Equation (5.3), which was used to fit

the data in Figure 5.8 and Figure 5.10.

]Li(S)['

]Li(L)[1

21 −+= kkkobs (5.3)


123

Plots of kobs/[Li(L)] vs. 1/[Li(S)] were found to be linear with negligible

slopes (k-2’ = k-2[S]) (especially at low temperature) and significant intercepts

(k1) (see Figures 5.8 and 5.10). This indicates that the exchange mechanism

has a more associative character for both C221 and C222. This led us to take

only the values of k1 into consideration, from which it follows that the exchange

process is much faster for C222 than for C221. The corresponding Eyring plots

for k1 are shown in Figures 5.9 and 5.11, respectively, and the thermal

activation parameters are summarized in Table 5.2.

30 35 40 45 50 550

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

k obs/[

Li(L

)], M

-1s-1

1/[Li(S)], M-1

248K 258K 268K 278K 283K 288K 298K

Figure 5.8. Plots of kobs/[Li(L)] vs 1/[Li(S)] for C222 as a function of temperature.


124

0.0030 0.0032 0.0034 0.0036 0.0038 0.0040 0.0042

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0 Y = A + B * X

Parameter Value Error------------------------------------------------------------A 18.85629 0.38719B -4237.51545 108.61204------------------------------------------------------------

ln (k

1/T)

1/T, K-1

Figure 5.9. Eyring plot for k1 as a function of temperature for C222.

30 40 50 60 70 80 90 1000

500

1000

1500

2000

2500

3000

3500

4000

4500

k obs/[L

i(L)],

M-1s-1

1/[Li(S)], M-1

298K 308K 318K 328K 338K 348K 358K

Figure 5.10. Plots of kobs/[Li(L)] vs 1/[Li(S)] for C221 as a function of temperature.


125

0.0027 0.0028 0.0029 0.0030 0.0031 0.0032 0.0033 0.0034-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5 Y = A + B * X

Parameter Value Error------------------------------------------------------------A 18.36172 0.29829B -5766.54595 97.29293------------------------------------------------------------

ln (k

1/T)

1/T, K-1

Figure 5.11. Eyring plot for k1 as a function of temperature for C221.

Table 5.2. Thermal activation parameters obtained from the values of k1 as a function of temperature.

Chelate k1298, M-1 s-1 ∆H#, kJ mol-1 ∆S#, J K-1 mol-1

C221 105 ± 32 48 ± 1 -45 ± 2

C222 (3.3 ± 0.8) x 104 35 ± 1 -41 ± 3

Based on the significantly negative activation entropy, the activation

parameters support an associative interchange type of exchange process.

Furthermore, the reported activation parameters are within the range of those

reported by Popov et al. for closely related systems,31 viz. k298 = 680 ± 42 M-1 s-

1, ∆H# = 28 ± 1 kJ mol-1 and ∆S# = -101 ± 2 J K-1 mol-1 for Li+/C221 in

acetonitrile, and k298 = 892 ± 50 M-1 s-1, ∆H# = 26 ± 1 kJ mol-1 and ∆S# = -109 ±

2 J K-1 mol-1 for Li+/C221 in propylene carbonate, for which the authors also

postulated an associative character for the Li+ exchange process. In the case of

Li+ exchange between C222 and solvents such as acetonitrile, propylene

carbonate and acetone, these authors found the mechanism to have more of a

dissociative character. However, it should be pointed out that they used a

different rate law than the one proposed by Shchori et al.65 and that used in the

present study, which complicates a direct comparison of the reported


126

activation parameters with ours. It is reasonable to expect that solvents such as

propylene carbonate, ethylene carbonate and γ-butyrolactone, together

considered as a class of solvents that play an important role in polymer-gel

lithium-ion baterries,66-70 due to their similar structure and properties as

expected from their rather similar Gutmann donor numbers (DN = 15.1, 16.4

and 18.0, respectively71), should coordinate to Li+ in a similar way and should

have a similar influence on the exchange mechanism. This observation

corresponds to that reported by Shamsipur et al.72 for the exchange of Li+

between C221 and acetonitrile-nitromethane mixtures. In all solvent mixtures

used, Li+ exchange was found to occur via an associative mechanism, as

expected for weakly donor solvents (DN = 14.1 and 2.7 for acetonitrile and

nitromethane, respectively). The same group reported another case, where Li+

exchange occurred between C221 and mixtures of nitromethane (low donor

ability) and dimethylformamide (high DN = 26.6), where it was found that the

associative mechanism predominates in nitromethane, whereas a dissociative

pathway prevails in dimethylformamide solution.

5.5.3. High Pressure 7Li NMR Measurements



exchange process73-76. A systematic pressure dependence study was performed

to determine the volume of activation for the Li+ exchange process for both

chelates. Pressure dependent NMR spectra were recorded for each sample over

the pressure range of 0.2 to 150 MPa at 316.5 K. The effect of pressure was

found to be not very significant for these systems. There was a very small

difference in the shape of the NMR spectra recorded at various pressures as

can be seen in Figure 5.12. The rate constants calculated for each pressure were

very similar within the experimental error limits and resulted in a practically

zero value for ∆V#obs, which points to a concerted interchange type of exchange

mechanism.77 In terms of an interchange mechanism, bond formation and

bond cleavage occur in a concerted fashion without the formation of a distinct

intermediate of higher or lower coordination number.


127

Figure 5.12. 7Li NMR spectra recorded as a function of pressure at 316.5 K for the mole

ratio C221:LiClO4 = 0.65.

5.5.4. Influence of Water

The rate of the exchange process was also investigated in a mixture of γ-

butyrolactone and water in order to probe the impact of the solvent polarity

and donor ability on the rate of the exchange reaction. The exchange

measurements were repeated for one of the samples (C222:LiClO4 = 0.35) with

1μl H2O added to the 0.8 ml of γ-butyrolactone solution. It was observed that

the exchange rate increased significantly as compared to the value obtained for

the sample in the absence of water. The values of kobs increase from 27 to 75 s-1

at 248 K, from 654 to 1282 s-1 at 298 K, and from 5945 to 12895 s-1 at 358 K, on

addition of water. This can be accounted for by the higher donor strength of

water compared to γ-butyrolactone. It follows that the presence of water plays

an important role in the exchange rate of Li+ between cryptand and solvent.

5.5.5. Theoretical Calculations

Experimentally a decrease in the cavity size hampers the exchange of Li+

between the cryptand and the solvent. The improved binding of Li+ in a

smaller cavity should become visible from the structure and the following

model reaction that describes the energy required for the dechelation process

(Table 5.3).


128

[Li⊂Cryptand]+ + 4γ-Butyrolactone [Li(γ-Butyrolactone)4]+ + Cryptand (5.4)

It is noteworthy that in agreement with previous studies by Wipff et al.,5 we

also found only a limited flexibility for the investigated cryptands in our DFT

calculations, justifying that all cryptands were calculated within the highest

possible symmetry. In earlier work we demonstrated that eq. (5.4) is a valuable

tool to determine ion selectivity as a function of the ionic radii.78 We therefore

examined the cavity size of the cryptands as a variable.

Table 5.3. Dechelation energy, Edechelation, for the reaction outlined in (5.4).

γ-Butyrolactone C222 C221 C211

Edechelation [kJ/mol] -70.3 -48.6 -22.2

Table 5.4. Calculated (RB3LYP/D95Vp) and x-ray bond lengths (in Å).

d [Å] C222 C221 C211 [Li(OH2)4]+ [Li(γ-butyrolactone)4]+ [Li(NH3)4]+

Li-O

2.14

2.06

-

2.08

2.10

2.14

2.11

1.95

1.94

-.--

Li-N 2.22 2.28 2.3

8

-.-- -.-- 2.13

PG C3 C1 C2 T C1 C1

X-ray 79 80 81 82

Li-O

-.--

1.99

2.06

2.10

2.17

2.0

8

1.96

-.--

-.--

Li-N -.-- 2.40

2.44

2.29 -.-- -.-- 2.08

In agreement with the experimental kinetic data, the smaller cryptand

binds Li+ better and slows down its exchange rate (Table 5.2). On comparing


129

the calculated Li-O and Li-N bond lengths of the three structures with the

bond distance in [Li(OH2)4]+ and [Li(NH3)4]+, the trend can easily be

understood (see Table 5.4 and Figure 5.13). In C222, Li+ is four times weakly

coordinated in a trigonal pyramidal shape by one Li-N and three Li-O

interactions all clearly elongated (0.2 and 0.1 Å, respectively). The Li+ cation in

C221 is coordinated five times (distorted trigonal bipyramidal structure) by

one nitrogen and four oxygen bonds all extended by around 0.15 Å. The

coordinating bonds in [Li ⊂ C211]+ are even more extended, but the Li-O bonds

form a somewhat distorted tetrahedral coordination sphere for the lithium

mono cation, while the stronger elongated Li-N bonds form a kind of second

coordination sphere. Therefore, the lithium ion is formally six-coordinate. A

comparison of the calculated and x-ray bond lengths shows a good agreement.

Figure 5.13. Optimized structures for [Li ⊂ C222]+, [Li ⊂ C221]+, [Li ⊂ C211]+ and [Li(γ-Butyrolactone)4]+.


130

5.6. Conclusions

It can be concluded from the results of the present study that for the Li+

exchange process between C222, C221 and γ-butyrolactone as solvent, two

pathways, viz. bimolecular associative and unimolecular dissociative, compete

with each other. The pressure dependence of the exchange process suggests the

operation of a pure interchange mechanism, whereas the negative activation

entropy values favor an associative interchange process. The basic idea of an

interchange process involves concerted bond-formation and bond-cleavage,

which is in agreement with the consideration that γ-butyrolactone is a rather

weak donor. The results are in good agreement with that expected on the basis

of a systematic comparison with data reported in the literature for solvents of

similar donor strength.


131

5.7. References


2889.


2885.

3. Lehn, J. M. Acc. Chem. Res. 1978, 11, 49.

4. Lehn, J. M. Angew. Chem. 1988, 100, 91.

5. Wipff, G.; Wurtz, J. M. New J. Chem. 1989, 13, 807.

6. Zhang, X. X.; Izatt, R. M.; Krakowiak, K. E.; Bradshaw, J. S. Inorg.

Chim. Acta 1997, 254, 43.


8. Puchta, R.; Seitz, V.; van Eikema Hommes, N. J. R.; Saalfrank, R. W. J.

Mol. Model. 2000, 6, 126.

9. Saalfrank, R. W.; Dresel, A.; Seitz, V.; Trummer, S.; Hampel, F.;

Teichert, M.; Stalke, D.; Stadler, C.; Daub, J.; Schunemann, V.;

Trautwein, A. X. Chem.-Eur. J. 1997, 3, 2058.

10. Saalfrank, R. W.; Seitz, V.; Caulder, D. L.; Raymond, K. N.; Teichert, M.;

Stalke, D. Eur. J. Inorg. Chem. 1998, 9, 1313.

11. Saalfrank, R. W.; Seitz, V.; Heinemann, F. W.; Gobel, C.; Herbst-Irmer,

R. J. Chem. Soc., Dalton Trans. 2001, 5, 599.

12. Dietrich, B.; Kintzinger, J. P.; Lehn, J. M.; Metz, B.; Zahidi, A. J. Phys.

Chem. 1987, 91, 6600.

13. Lindoy, L. F. The Chemistry of Macrocyclic Ligand Complexes;

Cambridge University Press: Cambridge, 1992; p 269.

14. Farahbakhsh, M.; Schmidt, H.; Rehder, D. Chem. Ber. Recl. 1997, 130,

1123.

15. Housecroft, C. E. Cluster Compounds of Main Group Elements; VCH:

Weinheim, Germany, 1996; p 94.

16. Kirch, M.; Lehn, J. M. Angew. Chem. 1975, 87, 542.


18. Lehn, J. M.; Sauvage, J. P. J. Chem. Soc., Chem. Commun. 1971, 9, 440.


132

19. Morf, W. E.; Simon, W. Helv. Chim. Acta 1971, 54, 2683.

20. Pregel, M. J.; Buncel, E. J. Am. Chem. Soc. 1993, 115, 10.

21. Simon, W.; Morf, W. E.; Meier, P. C. Struct. Bond. 1973, 16, 113.

22. Severin, K. Coord. Chem. Rev. 2003, 245, 3.

23. Huang, R. H.; Wagner, M. J.; Gilbert, D. J.; Reidy-Cedergren, K. A.;

Ward, D. L.; Faber, M. K.; Dye, J. L. J. Am. Chem. Soc. 1997, 119, 3765.


25. Wipff, G.; Kollman, P. Nouv. J. Chim. 1985, 9, 457.

26. Shamsipur, M.; Karkhaneei, E.; Afkhami, A. Polyhedron 1998, 17, 3809.

27. Chen, Q.; Cannell, K.; Nicoll, J.; Dearden, D. V. J. Am. Chem. Soc. 1996,

118, 6335.

28. Chantooni, M. K., Jr.; Kolthoff, I. M. J. Solution Chem. 1985, 14, 1.

29. Cox, B. G.; Garcia-Rosas, J.; Schneider, H. J. Am. Chem. Soc. 1981, 103,

1054.

30. Rhinebarger, R. R.; Popov, A. I. Polyhedron 1988, 7, 1341.

31. Shamsipur, M.; Popov, A. I. J. Phys. Chem. 1986, 90, 5997.

32. Smetana, A. J.; Popov, A. I. J. Solution Chem. 1980, 9, 183.

33. Abou-Hamdan, A.; Lincoln, S. F. Inorg. Chem. 1991, 30, 462.

34. Clarke, P.; Gulbis, J. M.; Lincoln, S. F.; Tiekink, E. R. T. Inorg. Chem.

1992, 31, 3398.

35. Clarke, P.; Lincoln, S. F.; Tiekink, E. R. T. Inorg. Chem. 1991, 30, 2747.

36. Lincoln, S. F.; Abou-Hamdan, A. Inorg. Chem. 1990, 29, 3584.



39. Lincoln, S. F.; Stephens, A. K. W. Inorg. Chem. 1991, 30, 3529.

40. Stephens, A. K. W.; Dhillon, R. S.; Madbak, S. E.; Whitbread, S. L.;

Lincoln, S. F. Inorg. Chem. 1996, 35, 2019.


1991, 91, 1721.

42. Shchori, E.; Jagur-Grodzinski, J.; Shporer, M. J. Am. Chem. Soc. 1973,

95, 3842.


133

43. Strasser, B. O.; Hallenga, K.; Popov, A. I. J. Am. Chem. Soc. 1985, 107,

789.

44. Strasser, B. O.; Popov, A. I. J. Am. Chem. Soc. 1985, 107, 7921.


J.; Sen, D. Chem. Rev. 1985, 85, 271.

46. Izatt, R. M.; Bradshaw, J. S.; Pawlak, K.; Bruening, R. L.; Tarbet, B. J.

Chem. Rev. 1992, 92, 1261.


Reviews 1995, 95, 2529.

48. Sonoda, K.; Deguchi, M.; Koshina, H.; Armand, M.; Michot, C.;

Gauthier, M. Proc. - Electrochem. Soc. 2003, 2002-26, 512.

49. Takami, N.; Sekino, M.; Ohsaki, T.; Kanda, M.; Yamamoto, M. J. Power

Sources 2001, 97-98, 677.

50. Hoffmann, H. M. R.; Rabe, J. Angew. Chem. 1985, 97, 96.

51. Mandal, S. K.; Amin, S. R.; Crowe, W. E. J. Am. Chem. Soc. 2001, 123,

6457.

52. Zhu, Y.-L.; Xiang, H.-W.; Wu, G.-S.; Bai, L.; Li, Y.-W. Chem. Commun.

2002, 3, 254.

53. Masia, M.; Rey, R. J. Phys. Chem. B 2004, 108, 17992.

54. Zahl, A.; Neubrand, A.; Aygen, S.; van Eldik, R. Rev. Sci. Instrum. 1994,

65, 882.


56. Matlab, version 5.3.1.; Mathworks Inc.: 1999.





Phys. 1988, 37, 785.


Chem. 1994, 98, 11623.


New York, 1976; Vol. 3, p 1.


134




63. The performance of the computational level employed in this study is

well documented, see for example; Galle, M.; Puchta, R.; van Eikema

Hommes, N.; van Eldik, R. Z. Phys. Chem. 2006, 220, 511; Puchta, R.;

van Eikema Hommes, N. J. R.; Meier, R.; van Eldik, R. Dalton Trans.

2006, 28, 3392; Klaus, S.; Neumann, H.; Jiao, H.; Jacobi von Wangelin,

A.; Goerdes, D.; Struebing, D.; Huebner, S.; Hateley, M.; Weckbecker,

C.; Huthmacher, K.; Riermeier, T.; Beller, M. J. Organomet. Chem.

2004, 689, 3685; Saalfrank, R. W.; Deutscher, C.; Maid, H.; Ako, A. M.;

Sperner, S.; Nakajima, T.; Bauer, W.; Hampel, F.; Hess, B. A.; van

Eikema Hommes, N. J. R.; Puchta, R.; Heinemann, F. W. Chem.-Eur. J.

2004, 10, 1899; Scheurer, A.; Maid, H.; Hampel, F.; Saalfrank, R. W.;

Toupet, L.; Mosset, P.; Puchta, R.; van Eikema Hommes, N. J. R. Eur. J.

Org. Chem. 2005, 12, 2566; Weber, C. F.; Puchta, R.; van Eikema

Hommes, N. J. R.; Wasserscheid, P.; van Eldik, R. Angew. Chem. 2005,

117, 6187; Weber, C. F.; Puchta, R.; van Eikema Hommes, N. J. R.;

Wasserscheid, P.; van Eldik, R. Angew. Chem. (Int. Ed. Eng.) 2005, 44,

6033.














135






65. Shchori, E.; Jagur-Grodzinski, J.; Luz, Z.; Shporer, M. J. Am. Chem. Soc.

1971, 93, 7133.

66. Song, J. Y.; Wang, Y. Y.; Wan, C. C. J. Power Sources 1999, 77, 183.

67. Matsuda, Y.; Fukushima, T.; Hashimoto, H.; Arakawa, R. J.

Electrochem. Soc. 2002, 149, A1045.

68. Adebahr, J.; Forsyth, M.; MacFarlane, D. R.; Gavelin, P.; Jacobsson, P. J.

Mater. Chem. 2003, 13, 814.

69. Ding, M. S.; Xu, K.; Zheng, J. P.; Jow, T. R. J. Power Sources 2004, 138,

340.

70. Ue, M. J. Electrochem. Soc. 1994, 141, 3336.



23.

73. Davis, A. V.; Fiedler, D.; Seeber, G.; Zahl, A.; Van Eldik, R.; Raymond, K.

N. J. Am. Chem. Soc. 2006, 128, 1324.

74. van Eldik, R.; Hubbard, C. D. Chemistry at Extreme Conditions;

Elsevier: Amsterdam, 2005; p 109.

75. Wolak, M.; van Eldik, R. J. Am. Chem. Soc. 2005, 127, 13312.

76. Zahl, A.; Van Eldik, R.; Matsumoto, M.; Swaddle, T. W. Inorg. Chem.

2003, 42, 3718.


78. Galle, M.; Puchta, R.; van Eikema Hommes, N. J. R.; van Eldik, R. Z.

Phys. Chem. 2006, 220, 511.

79. Choua, S.; Sidorenkova, H.; Berclaz, T.; Geoffroy, M.; Rosa, P.;

Mezailles, N.; Ricard, L.; Mathey, F.; Le Floch, P. J. Am. Chem. Soc.

2000, 122, 12227.


136

80. Moras, D.; Weiss, R. Acta Crystallogr.,Sect. B: Struct. Crystallogr.

Cryst. Chem. 1973, 29, 400.

81. Meske, W.; Babel, D. Z. Anorg. Allg. Chem. 1998, 624, 1751.

82. Scotti, N.; Zachwieja, U.; Jacobs, H. Z. Anorg. Allg. Chem. 1997, 623,

1503.

6. Summary

137

6. Summary

In this work the solvation process of Li+ ion, as well as solvent and ligand

exchange reactions on Li+ ion were studied. Li+ ions possess interesting

properties and like other alkali metal ions are known to form complexes with

macrocyclic ligands called cryptands. In this summary, an overview over the

insights gained in the factors that control the reactivity of Li+ complexes with

respect to the solvent and cryptand properties is presented.

Three main questions were addressed:

• How does the nature of the solvent influence the first coordination

sphere around the Li+ ion?

• How is solvent exchanged between the first and second coordination

spheres of the Li+ ion?

• How is the Li+ ion complexed by different cryptands and what is the

nature of the Li+ exchange mechanism between cryptand and

solvent?

In this thesis these aims were approached from the microscopic point of

view, making use of both theoretical calculations and experimental techniques.

The first two topics mentioned above constitute the contents of Chapters 2

and 3, whereas the third objective was approached and the results are reported

in Chapters 4 and 5.

The experimental method used throughout was 7Li NMR. It was shown that

7Li NMR is a very powerful technique for the study of alkali complexes with

macrocyclic ligands, particularly in non-aqueous solutions. Particular attention

was given to high pressure 7Li NMR, which represents the first example

reported for this kind of research. It could be demonstrated that 7Li NMR is a

very convenient method for the estimation of the coordination number on Li+

ion as well as for investigation of the selective solvation of Li+ in a mixture of

solvents. This method was also used for mechanistic studies, specifically for Li+

exchange between two environments: solvated and chelated species.

6. Summary

138

Solvents of different properties and their influence on the solvation shell

were investigated. Questions concerning selective solvation around Li+ and the

solvation number of the Li+ cation were studied as a function of the solvent

properties. The Gutmann Donor Number (DN) was mostly considered as a

factor to describe the donor properties of the solvent. It reflects the ease of

donation of the solvent molecule’s lone electron pair.

In order to determine the coordination number of Li+, the cation was

dissolved in a solvent mixture that consisted of a weakly and strongly

coordinating component. It could be shown in Chapter 2 that the Li+ ion in

DMSO is coordinated by 4 solvent molecules. γ-Butyrolactone was found to be

a convenient diluent in this kind of study. Contrary to DMSO (DN = 29.8), it is

rather a weak donor (DN = 18.0) which appreciably dissolves alkali metal salts

and its interaction with DMSO is much weaker than that between DMSO and a

metal cation. By plotting the chemical shift of the obtained 7Li NMR signal vs.

the DMSO/LiClO4 mole ratio, a clear brake point could be observed at a

DMSO/Li+ ratio of 4:1, indicating that the solvation number of Li+ in DMSO is

4. Using the same method for the solvent mixture of acetonitrile (DN = 14.1)

and nitromethane (DN = 2.7), it was found that also in this case the Li+ ion is

coordinated by four acetonitrile molecules, as reported in Chapter 3.

Selective solvation of alkali metal ions in mixed solvents can be monitored

from the extent to which the chemical shift of the alkali metal nucleus varies

with solvent composition. In the binary solvent mixture of water and DMSO

reported in Chapter 2, no selective solvation was observed, indicating that on

increasing the water content of the solvent mixture, DMSO is gradually

displaced by water in the coordination sphere of Li+. We did not observe any

rapid change in the position of the NMR signal when the amount of water was

increased. DMSO is a solvent of high solvating ability which can strongly

coordinate to the Li+ ion and can compete with water in the first coordination

sphere of the Li+ ion. It could be concluded that the composition of the first

coordination sphere in mixtures of H2O/DMSO gradually changes from

[Li(DMSO)4]+ to [Li(H2O)4]+ on increasing the water content of the solvent.

6. Summary

139

This is a consequence of the rather similar donor strength of these solvents as

expressed by their donor numbers.

In the binary water/acetonitrile mixture, however, water is coordinated

stronger to Li+ than acetonitrile such that addition of water immediately leads

to the formation of [Li(H2O)4]+. It was observed that the position of the NMR

signal moved significantly (from 4.5 to 6.4 ppm) even if the amount of added

water is very small (only 10 % H2O). Then the chemical shift value increased

slightly and stayed almost constant while the amount of water was changed

from 10 to 100 %. This observation confirms that water coordinates much

stronger to the Li+ ion than acetonitrile (Chapter 3).

Once the solvation shell of the Li+ ion was known, we wanted to investigate

the exchange of certain solvent molecules around Li+. The Li+ ion, however, is

very labile due to the low surface charge density and the absence of ligand field

stabilization effects. Solvent exchange on Li+ is therefore extremely fast,

making it difficult to study experimentally. The ligand exchange mechanism of

Li+ ions solvated by DMSO and water/DMSO mixtures was studied using DFT

calculations. Ligand exchange on [Li(DMSO)4]+ was found to follow a limiting

associative mechanism, whereas the displacement of coordinated H2O by

DMSO in [Li(H2O)4]+ follows an associative interchange mechanism.

In Chapter 3, we have extended our combined quantum chemical and

experimental investigations of the ligand coordination and ligand exchange on

Li+ in acetonitrile (CH3CN) and HCN, as well as in CH3CN/H2O and HCN/H2O

mixtures.

We studied the ligand exchange process in more detail and report

computational evidence for a limiting associative exchange mechanism on the

HCN and CH3CN solvated lithium cation. HCN was employed here as a

simplified model to perform quantum chemical studies on CH3CN. Our

calculations corroborated the experimental work that the first coordination

sphere around Li+ consists of four tightly bound CH3CN and HCN molecules.

Ligand exchange reactions on Li+ most probably follow an associative

6. Summary

140

mechanism as indicated by the formation of stable five-coordinate

intermediates.

The class of ligands that were found to form stable complexes with alkali

metals are macrocyclic ligands called cryptands. They were designed and

shown to form stable alkali metal (i.e. lithium) cryptate complexes by inclusion

into the three-dimensional molecular cavity. Here three cryptands, viz. C211,

C221, and C222 were investigated.

In this work, we were particularly interested in the mechanistic aspects of

the Li+ exchange process between cryptands and solvent. In preliminary

experiments we investigated the Li+ exchange in different solvents. It was

shown that the stability of the cryptate varies with the nature of the solvent.

The stability tends to decrease as DN increases. In water and methanol the

exchange process was too fast to be studied by 7Li NMR (only a single signal

was observed). The process became slower in acetonitrile, and in acetone and

γ-butyrolactone two completely separated signals were observed at low

temperature. Therefore, these two solvents, viz. acetone and γ-butyrolactone,

were chosen for more detailed mechanistic investigations.

In Chapter 4 we focused on acetone as a solvent and the competition

between solvated and endohedral complexed Li+, reported kinetic parameters

for the exchange of Li+ between solvated and complexed states in acetone

under ambient and elevated pressure conditions, and discussed the

mechanistic implications.

Initially, a series of 7Li NMR measurements for the Li+-cryptand system in

acetone were performed. The results enabled us to observe the distribution of

particular species on variation of the concentration ratio. Basically, two signals

of varying intensity were observed in the 7Li NMR spectra that refer to solvated

Li+ and Li+ bound inside the cryptand cavity. In the case of the smallest chelate

(C211), however, an additional signal appeared which could be ascribed to Li+

bound outside the cryptand and being partially solvated. This complex is very

stable and almost no exchange took place. A detailed study of the observed rate

6. Summary

141

constant, kobs, for the exchange of Li+ as a function of temperature and

pressure was therefore undertaken only for C221 and C222.

Stabilities of the cryptates increase in the sequence Li(C222)+ < Li(C221)+ <

Li(C211)+, consistent with the six donor atoms of the C211 cavity and the

optimal fit of Li+ (r = 1.48 Å) into the C211 cavity (r = 1.6 Å) in inclusive

Li(C211)+, generating the largest stability. The increase in stability Li(C222)+ <

Li(C221)+ indicates that the effect of the decrease in electrostatic interaction

with Li+ resulting from the decrease in the number of donor atoms from 8 in

C222 to 7 in C221 is offset by the smaller cavity size of C221, with the

consequence that Li(C221)+ is the more stable cryptate. The more open and

flexible structure of C222 accounts for the larger kobs values that characterize

Li(C222)+ by comparison with those of the relatively rigid inclusive Li(C211)+.

These comparisons illustrate the interplay of the effects of the number of

donor atoms and cryptand flexibility which produce variations in cryptate

stability and lability.

Two mechanisms were considered for the decomplexation processes of

cryptand-alkali metal complexes: bimolecular and unimolecular cation

exchange.

The obtained results indicate that both pathways compete with each other

and the competition is controlled by temperature and the nature of the

cryptand. However, the unimolecular pathway seems to dominate, indicating

the nonparticipation of solvated Li+ in the rate-determining step, resulting in

the first order rate constant k-2. Based on the values of k-2 and the significantly

negative activation entropies obtained for this pathway, the activation

parameters support an associative type of exchange process. k-2 for C222 is

almost three orders of magnitude larger than for C221. The negative activation

entropies found for the decomplexation reactions of both C221 and C222

strongly suggest that attack of the solvent on the complexed Li+ must play an

important role in the decomplexation process, i.e. decomplexation follows an

associative mechanism.

6. Summary

142



exchange process. It could be observed that there was a very small difference in

the shape of the NMR spectra recorded at various pressures. The observed rate

constants calculated for each pressure were very similar within the

experimental error limits and resulted in a practically zero value for ∆V#obs for

both ligands, which points to a concerted interchange type of exchange

mechanism. In terms of an interchange mechanism, bond formation and bond

cleavage occur in a concerted fashion without the formation of a distinct

intermediate of higher or lower coordination number. In addition,

decomplexation will be accompanied by charge creation due to the release of

Li+, which in turn will lead to an increase in electrostriction which will offset

the intrinsic volume changes.

It is not invariably the case that the unimolecular mechanism operates in

cryptate systems. This is exemplified by the system in γ-butyrolactone, for

which the Li+ exchange process follows rather the bimolecular mechanism in

which also solvated Li+ is involved and therefore the concentration of Li+ is

taken into account in the rate law.

In Chapter 5, we report kinetic parameters for the exchange of Li+ in γ-

butyrolactone under ambient and elevated pressure conditions, and discuss

their mechanistic implications. γ-Butyrolactone was chosen as solvent due to

its relevance in lithium batteries where it is used as plasticizer.

It could also be seen in this system that both pathways, bimolecular and

unimolecular compete with each other. However, the bimolecular pathway

seems to predominate, indicating participation of solvated Li+ in the rate-

determining step, resulting in the second order rate constant k1. Involvement

of the solvent in the decomplexation process was found to be less pronounced

than it was in the case of acetone. This is probably due to the fact that although

both solvents, acetone and γ-butyrolactone should coordinate to the naked Li+

cation comparatively strongly, it cannot be observed in the process in which

Li+ must be released from the cryptand cavity. In this case the steric hindrance

6. Summary

143

should be taken into account, i.e. acetone as a sterically less demanding

molecule has probably better access to the Li+ complexed by the cryptand as γ-

butyrolactone does. Therefore, second order rate constants were obtained for

both ligands, as a consequence of participation of the solvated Li+ in the

decomplexation process.

Based on the values of k1 and the significantly negative activation entropies

obtained for this pathway, the activation parameters support an associative

interchange type of exchange process. The effect of pressure was found to be

not very significant for these systems and resulted in a practically zero value

for ∆V#obs, which points to a concerted interchange type of exchange

mechanism.

This work clearly demonstrated that the ligand structure and solvent

properties have considerable effects on the reaction rate and the exchange

mechanism of the cation between the solvated and the complexed sites. The

stability of the cryptate complexes depends to a large extent on the size

relationship between the three-dimensional cavity of the cryptand and the

diameter of the complexed ion. The relative stabilities of cryptates depend on

the electrostatic interactions between metal ions and cryptands, the solvation

energies of metal ions and cryptands, and the fit of the metal ion into the

cryptand cavity.

In general it can be concluded that the present thesis has shown in a

systematic manner that the combination of experimental (dynamic 7Li NMR)

and theoretical (DFT) methods represents a powerful tool that enabled us to

gain valuable qualitative and quantitative insight into mechanistic details of

the coordination behavior of lithium cations and cryptands in different

solvents. To the best of our knowledge, this was for the first time that

mechanistic investigations on complex-formation reactions by cryptands were

carried out using high pressure NMR techniques, in particular 7Li NMR. The

results presented here give a promising outlook in terms of the possibility to

investigate further lithium ion technologically relevant systems by the methods

used in this study.

7. Zusammenfassung

144

7. Zusammenfassung

In der vorliegenden Arbeit wurde der Solvatisierungsprozess des

Lithiumkations, sowie der Lösungsmittel- und Ligandenaustausch am

Lithiumion, untersucht. Das Lithiumkation bildet ebenso wie die anderen

Alkalimetallkationen, Komplexe mit Kryptanden. In dieser Arbeit lag der

Schwerpunkt auf C211, C221 und C222. Diese Zusammenfassung soll einen

Überblick über die Faktoren geben, die die Reaktivität der Lithium-(I)-

komplexe in Abhängigkeit von Lösungsmittel- und Kryptandeneigenschaften

steuern.

Folgende drei grundsätzliche Fragen wurden angegangen:

Wie wird die erste Koordinationssphäre um das Lithiumion von der Natur

des Lösungsmittels beeinflusst?

Wie wird das Lösungsmittel zwischen der ersten und der zweiten

Koordinationssphäre des Lithiumions ausgetauscht?

Wie wird das Lithiumion von verschiedenen Kryptanden komplexiert und

wie ist der Mechanismus des Lithiumionaustausches zwischen Kryptand und

Lösungsmittel.

Die gesteckten Ziele wurden aus der mikroskopischen Perspektive

angegangen. Anwendung fanden hier sowohl experimentelle Methoden (7Li

NMR) als auch quantenchemische Rechnungen. Die ersten beiden oben

genannten Fragestellungen bilden den Kern der Kapitel 2 und 3, während die

erhaltene Ergebnisse des dritten Themas in den Kapitel 4 und 5 beschrieben

sind.

Wie aus der Literatur bereits bekannt war, stellt 7Li NMR Spektroskopie

eine sehr leistungsstarke Methode zur Untersuchung von Alkalimetallionen in

makrozyklischen Liganden, besonders in nichtwässrigen Lösungen dar. In der

vorgelegten Arbeit wird hierbei besondere Aufmerksamkeit der Hochdruck 7Li

NMR Spektroskopie gewidmet. Sie wird hier zum ersten Mal für

Untersuchungen an Kryptaten benutzt. Es wurde weiterhin gezeigt, dass die 7Li

7. Zusammenfassung

145

NMR Spektroskopie eine sehr leistungsstarke Methode zur Bestimmung der

Koordinationszahl des Lithiumions als auch zur Untersuchung der selektiven

Solvatation in gemischten Lösungsmitteln ist. Diese Methode wurde auch für

mechanistische Untersuchungen benutzt, hier konkret für den

Lithiumionenaustausch zwischen solvatisierten und chelatisierten Spezies.

Lösungsmittel von verschiedenen Eigenschaften und ihr Einfluss auf die

Solvatationshülle wurden untersucht. Das Problem der selektiven Solvatation

um das Lithiumion und die Koordinationszahl des Lithiumions wurden als

Funktion der Lösungsmitteleigenschaften studiert. Die Donorzahl nach

Gutmann (DN) wurde meistens als der Faktor eingesetzt, der die Fähigkeit des

Lösungsmittels als Donor zu agieren charakterisiert.

Um die Koordinationszahl des Lithiumkations zu bestimmen, wurde eine

Lösung vorbereitet, die aus einem Gemisch von schwach und stark

koordinierenden Lösungsmitteln bestand. In Kapitel 2 wurde gezeigt, dass

das Lithiumion in DMSO von 4 Lösungsmittelmolekülen koordiniert wird.

Wir konnten zeigen, dass γ-Butyrolakton ein gut geeignetes

Verdünnungsmittel für derartige Untersuchungen ist. Im Gegensatz zu DMSO

(DN = 29.8), stellt γ-Butyrolakton einen eher schwachen Donor (DN = 18.0)

dar. Es löst die Salze der Alkalimetalle in merklicher Weise, wobei die

Wechselwirkung mit den DMSO-Molekülen wesentlich schwächer ist als die

zwischen DMSO und den Metallkationen. Durch graphische Auftragung der

chemischen Verschiebung des 7Li NMR Signals gegen das Molverhältniss von

DMSO/LiClO4 konnte ein deutlicher Bruch in der Kurve bei einem

Molverhältniss von 4:1 beobachtet werden. Dies deutet auf eine

Koordinationszahl von 4 für Li+ in DMSO hin. Die Anwendung dieser Methode

auf ein Lösungsmittelgemisch bestehend aus Acetonitril (DN = 14.1) und

Nitromethan (DN = 2.7) ergab ebenfalls eine vierfache Koordination des

Lithiumkations in Acetonitril Lösung, wie im Kapitel 3 berichtet wird.

Selektive Solvatation der Alkalimetallionen in Lösungsmittelgemischen

kann anhand der Veränderung der chemischen Verschiebung des

Alkalimetallkerns in Abhängigkeit mit der Zusammensetzung des

7. Zusammenfassung

146

Lösungsmittelgemisches beobachtet werden. Im binären

Lösungsmittelgemisch Wasser/DMSO, über das im Kapitel 2 berichtet wird,

wurde keine selektive Solvatation beobachtet. Dies deutet darauf hin, dass, mit

wachsendem Wassergehalt im Lösungsmittelgemisch, DMSO rein statistisch

sukzessiv durch Wasser in der Koordinationssphäre des Lithiumions ersetz

wird. Wir beobachteten keine sprunghaften Änderungen des NMR-Signals bei

Erhöhung des Wasseranteils.

DMSO ist ein Lösungsmittel mit großer Solvatationsfähigkeit, das das

Lithiumion stark koordinieren kann, und deswegen mit Wasser in der ersten

Koordinationssphäre des Lithiumions konkurriert. Es kann die

Schlussfolgerung gezogen werden, dass sich die Zusammensetzung der ersten

Koordinationssphäre in Gemischen von H2O/DMSO mit zunehmendem

Wassergehalt im Lösungsmittelgemisch schrittweise von [Li(DMSO)4]+ zu

[Li(H2O)4]+ ändert. Dies lässt auf eine ähnliche Donorstärke dieser

Lösungsmittel schließen, wie sie sich auch aus ihren Donorzahlen ergibt.

In binären Gemischen von Wasser und Acetonitril ist jedoch Wasser zum

Lithiumion stärker koordiniert als Acetonitril, so dass die Zugabe von Wasser

sofort zur Bildung von [Li(H2O)4]+ führt. Wir beobachteten, dass sich die

Position des NMR-Signals signifikant ändert (4.5 zu 6.4 ppm) sogar wenn die

Menge des zugegebenen Wasser sehr gering ist (nur 10 % H2O). Danach

erhöhte sich die chemische Verschiebung nur moderat und blieb konstant

während die Menge des Wassers von 10 % auf 100 % gesteigert wurde. Diese

Beobachtung bestätigt, dass Wasser viel stärker als Acetonitril an das

Lithiumion koordiniert (Kapitel 3).

Nachdem die Solvatationshülle des Lithiumions aufgeklärt wurde, wollten

wir den Austausch von bestimmten Lösungsmittelmolekülen um das

Lithiumion untersuchen. Das Lithiumkation ist jedoch wegen seiner geringen

Oberflächenladungsdichte und fehlenden Ligandenfeldstabilisierungseffekten

sehr labil. Der Lösungsmittelaustausch am Lithiumion ist deswegen extrem

schnell und experimentell schwierig zu untersuchen. Der

Ligandenaustauschmechanismus an dem von DMSO und Wasser/DMSO-

7. Zusammenfassung

147

Gemischen solvatisierten Lithiumion wurde mittels DFT-Rechnungen

untersucht. Es wurde gefunden, dass der Lösungsmittelaustausch an

[Li(DMSO)4]+ durch einen reinen assoziativen Mechanismus (A) erfolgt,

während die Substitution von koordiniertem Wasser durch DMSO in

[Li(H2O)4]+ durch einen assoziativen Interchange Mechanismus erfolgt.

Im Kapitel 3 erweiterten wir unsere kombinierten quantenchemischen

und experimentellen Untersuchungen der Ligandenkoordination und des

Ligandenaustausches auf das Lithiumion in CH3CN und HCN, wie auch auf

CH3CN/H2O und HCN/H2O Gemische.

Wir studierten die Ligandenaustauschprozesse detaillierter und konnten

quantenchemische Hinweise für einen reinen assoziativen

Austauschmechanismus an dem von HCN und CH3CN solvatisierten

Lithiumion finden. HCN wurde hier als ein vereinfachtes Modell für

quantenchemische Untersuchungen von CH3CN eingesetzt. Unsere

Rechnungen untermauerten zusätzlich die experimentellen Ergebnisse, dass

die erste Koordinationssphäre um das Lithiumion aus vier koordinierten

CH3CN- und HCN-Molekülen besteht. Ligandenaustauschreaktionen am

Lithiumion erfolgen am wahrscheinlichsten durch einen assoziativen

Mechanismus, worauf die Bildung von stabilen fünffach-koordinierten

Intermediaten hindeutet.

Wie bereits ausgeführt, bilden Kryptanden mit Alkalimetallionen stabile

Komplexe, sog. Kryptate. Wir haben drei Kryptanden untersucht: C211, C221

und C222.

In dieser Arbeit waren wir insbesondere an mechanistischen Aspekten des

Austauschprozesses am Lithiumion zwischen Kryptanden und Lösungsmittel

interessiert. In experimentellen Vorarbeiten untersuchten wir den

Lithiumionenaustausch in verschiedenen Lösungsmitteln. Es konnte gezeigt

werden, dass die Stabilität des Kryptats von der Natur des Lösungsmittels

abhängt. Die Stabilität nimmt mit zunehmendem DN ab. In Wasser und

Methanol war der Austauschprozess zu schnell, um mit 7Li NMR

Spektroskopie studiert werden zu können (nur ein Signal wurde beobachtet).

7. Zusammenfassung

148

In Acetonitril ist der Prozess langsamer geworden, und in Aceton und γ-

Butyrolakton konnten zwei vollständig getrennte Signale bei tiefen

Temperaturen beobachtet werden. Infolgedessen wurden diese zwei

Lösungsmittel (Aceton und γ-Butyrolakton) für weitere detaillierte

mechanistische Untersuchungen ausgewählt.

Im Kapitel 4 konzentrierten wir uns auf Aceton als Lösungsmittel und auf

die Konkurenz zwischen solvatisirtem und endohedral komplexiertem

Lithiumion. Darüber hinaus wird in diesem Kapitel über kinetische Parameter

für den Lithiumionenustausch zwischen solvatisiertem und komplexiertem

Zustand in Aceton unter Umgebungs- und Hochdruckbedingungen berichtet

und mechanistische Folgerungen werden gezogen.

Zunächst wurde eine Serie von 7Li NMR-Messungen für das System

Li+/Kryptand durchgeführt. Die erhaltenen Ergebnisse ermöglichten uns die

Verteilung konkreter Spezies in Abhängigkeit von Konzentrationsänderungen

zu beobachten. Es wurden zwei Signale von variirender Intensität in den 7Li

NMR-Spektren beobachtet, die auf das solvatisierte und das im

Kryptandenhohlraum gebundene Lithiumion zurückgeführt werden können.

Im Falle des kleinsten Kryptanden (C211) erschien aber auch ein zusätzliches

drittes Signal, das auf das von außen gebundene und damit nur teilweise

solvatisierte Lithiumion zurückgeführt werden könnte. Der Li(C211)-Komplex

ist sehr stabil und es fand fast kein Austausch statt. Eine detaillierte Studie der

beobachtenen Geschwindigkeitskonstante (kobs) für den

Lithiumionenaustausch als Funtion von Temperatur und Druck wurde

deswegen nur für C221 und C222 durchgeführt.

Die Stabilität der Kryptate nimmt in der Reihnfolge Li(C222)+ < Li(C221)+

< Li(C211)+ zu. Dies ist im Einklang mit den sechs Donoratomen des

Hohlraums von C211 und damit, dass das Lithiumion (r = 1.48 Å) optimal in

den Hohlraum von C211 (r = 1.6 Å) passt. Die zunehmende Stabilität Li(C222)+

< Li(C221)+ < Li(C211)+ weist darauf hin, dass die Abnahme der Anzahl der

Donoratome von 8 in C222 auf 7 in C221 durch den kleineren Hohlraum von

C221 ausgeglichen wird. Dementsprechend ist Li(C221)+ der stabilere Kryptat.

7. Zusammenfassung

149

Die flexiblere Struktur von C222 hat den größeren kobs-Wert zu Folge, der für

Li(C222)+ charakteristisch ist, im Vergleich zu den Werten für den relativ

starren Li(C211)+ Einschlußkomplex. Diese Vergleiche stellen die Folgen der

unterschiedlichen Zahl von Donoratomen in Kombination mit der Flexibilität

des Kryptanden gut dar. Sie haben eine unterschiedliche Stabilität bzw.

Labilität der Kryptate zur Folge.

Es wurden zwei Mechanismen für den Dekomplexierungsprozess von

Alkalimetal-Kryptatkomplexen in Aceton in Betracht gezogen: der

bimolekulare und der unimolekulare Kationenaustausch. Die Ergebnisse

deuten darauf hin, dass beide Prozesse miteinander konkurrieren und von der

Temperatur und der Natur des Ligandes gesteuert werden. Allerdings scheint

der unimolekulare Prozess dominant zu sein, was auf die Nichtbeteiligung des

solvatisierten Lithiumions im geschwindigkeitsbestimmenden Schritt

hindeutet und was in einer Geschwindigkeitskonstante erster Ordnung

resultiert. Basierend auf den für diesen Prozess erhaltenen

Geschwindigkeitskonstanten und signifikant negativen Aktivierungsentropien,

kann auf einen assoziativen Austauschprozess geschlossen werden. Die

Geschwindigkeitskonstante für C222 ist fast um drei Größenordnungen größer

als für C221. Die negativen Aktivierungsentropien, die für die

Dekomplexierungsreaktionen von C221 und C222 beobachtet wurden, deuten

stark darauf hin, dass der Angriff des Lösungsmittels auf das komplexierte

Lithiumion eine wichtige Rolle im Dekomplexirungsprozess spielt, d.h. die

Dekomplexierung erfolgt durch einen assoziativen Mechanismus.

Weitere detaillierte mechanistische Informationen über die Natur des

Austauschmechanismuses kann aus dem Einfluss von hohen Drücken auf den

Li+-Austauschprozess gewonnen werden. Es wurde beobachtet, dass nur sehr

kleine Änderungen in den bei verschidenem Druck aufgenommenen NMR-

Spektren auftreten. Die in Abhängigkeit vom Druck berechneten

Geschwindigkeitskonstanten waren im Rahmen der experimentellen

Fehlergrenzen sehr ähnlich und resultierten in ΔV#obs-Werten von Null für

C221 und C222. Dies spricht eher für einen Interchangemechanismus (I). Die

7. Zusammenfassung

150

Bildung und Spaltung der Li-Donor-Bindungen erfolgen konzertiert ohne

Bildung eines eindeutigen Intermediats mit höherer oder niedrigerer

Koordinationszahl. Durch die Dekomplexierung der Li+-Ionen erfolgt eine

Freisetzung von beweglichen Landungsträgen, was auf der anderen Seite zu

zunehmender Elektrostriktion führt, die die intrinsische Volumenänderungen

ausgleicht.

Im Gegensatz zum unimolekularen Lithiumionenaustauschprozess bei in

Aceton gelösten Li-Kryptatkomplexen finden wir im Fall von in γ-Butyrolacton

gelösten Krytptaten eher einen bimolekularen Mechanismus. Dieser wird auch

durch die Konzentration der solvatisierten Lithiumionen beeinflusst,

dementsprechend muß auch die Lithiumkonzentration in die

Geschwindigkeitsberechnung einbezogen werden.

Im Kapitel 5 berichten wir über kinetische Parameter für den

Lithiumionenaustausch an in γ-Butyrolacton gelösten Krytptaten unter

verschiedenen Drücken und diskutieren ihre Folgen für den Mechanismus. Die

Wahl von γ-Butyrolacton als Lösungsmittel ergab sich aus dessen Bedeutung

für Lithiumbatterien.

Hier wurde ebenfalls beobachtet, dass beide Prozesse, der bimolekulare

und der unimolekulare, miteinander konkurrieren. Allerdings, ist in diesem

Fall der bimolekulare Prozess dominant, was auf die Beteiligung des

solvatisierten Lithiumions am geschwindigkeitsbestimmenden Schritt

hindeutet und in einer Geschwindigkeitskonstante zweiter Ordnung resultiert.

Die Beteiligung des Lösungsmittels am Dekomplexierungsprozess war geringer

als im Falle des Acetons. Beide Lösungsmittel, Aceton und γ-Butyrolacton,

koordinieren nicht komplexierte Lithiumkationen gleich gut. In Konkurrenz

zwischen den hier untersuchten Kryptanden und den Lösungsmittelmolekülen

um die Komplexierung kann dies jedoch nicht beobachtet werden. Da γ-

Butyrolacton im Gegensatz zu Aceton sterisch anspruchsvoller ist, muß der

Angriff auf ein im Kryptanden endohedral komplexiertes Li+-Ion als sterisch

gehindert angesehen werden. Somit spielen solvatisierte Li+-Ionen bei den

Austauschprozessen eine größere Rolle als im Fall des Acetons. Dies zeigt die

7. Zusammenfassung

151

für beide Liganden, C221 und C222, erhaltene Geschwindigkeitskonstanten

zweiter Ordnung.

Die für diesen Prozess erhaltenen Geschwindigkeitskonstanten und die

signifikant negativen Entropien, untermauern den assoziativen Interchange

Charakter der untersuchten Reaktionen. Es konnte hierbei kein signifikanter

Druckeinfluss beobachtet werden. Die ΔV#obs-Werten sind nahezu Null und

legen zusätzlich einen Interchange-Mechanismus (I) nahe.

Diese Arbeit hat eindeutig gezeigt, dass die Ligandenstruktur und die

Lösungsmitteleigenschaften einen erheblichen Einfluss auf die

Reaktionsgeschwindigkeit und den Mechanismus des Kationenaustausches

zwischen den solvatisierten und komplexierten Spezies hat. Die Stabilität der

Kryptatkomplexe hängt weitgehend vom Größenverhältnis zwischen dem

dreidimensionalen Hohlraum des Kryptanden und dem Durchmesser des

komplexierten Ions ab. Zusätzlich wird die relative Stabilität der Kryptate

durch die Wechselwirkungen zwischen den Li+-Ionen und den Kryptanden

sowie deren Solvatatisierungsenergien bedingt.

Allgemein kann zusammengefasst werden, dass die vorliegende

Dissertation auf systematische Art und Weise zeigt, dass das Zusammenspiel

der experimentellen (dynamische 7Li NMR Spektroskopie) und

quantenchemischen (DFT-Rechnungen) Methoden wichtige qualitative und

quantitative Einblicke in mechanistische Details des Koordinationsverhaltens

von Lithiumionen und Kryptanden in verschiedenen Lösungsmitteln

ermöglicht. Dies sind nach unserem Wissen die ersten mechanistischen

Untersuchungen an Kryptaten mittels Hochdruck-NMR (7Li NMR). Die hier

vorgestellten Ergebnisse sind vielversprechend für weitere Untersuchungen an

anwendungsbezogenen Systemen mit Lithiumionen.

Lebenslauf

Persönliche Daten:

Name: Ewa Maria Pasgreta

Geburtsdatum: 11. März 1976

Geburtsort: Więcbork (Polen)

Eltern: Urszula und Stefan Pasgreta

Staatsangehörigkeit: polnisch

Familienstand: ledig

Schulbildung:

September 1983 - Juni 1991 Grundschule in Runowo Krajeńskie, Polen

September 1991 - Juni 1995 Gymnasium in Bydgoszcz, Polen (abgeschlossen

mit Abitur)

Studium:

Oktober 1995 - Juni 2001 Universität in Toruń, Polen

Chemiestudium mit dem Abschluss Dipl.-Chem.

November 2001 – December 2006 Friedrich-Alexander Universität Erlangen-

Nürnberg,

Anfertigung der Promotion bei Prof. Dr. Dr. h.c.

mult. Rudi van Eldik an der Institut für

Anorganische Chemie

März 2007 Abschluss mit Promotion zum Dr. rer. nat.

Ligand Exchange Processes on Solvated Lithium Cations · • “Ligand Exchange Processes on...

Documents

Transcript of Ligand Exchange Processes on Solvated Lithium Cations · • “Ligand Exchange Processes on...