Citation
Alkali cation binding to nonionic polymers

Material Information

Title:
Alkali cation binding to nonionic polymers
Creator:
Cambron, Ronald Edward, 1948- ( Dissertant )
Esch, Thieo E. Hogan ( Thesis advisor )
Bates, Roger G. ( Reviewer )
Butler, George B. ( Reviewer )
Gabbay, Edmond J. ( Reviewer )
Shah, Dinesh O. ( Reviewer )
Place of Publication:
Gainesville, Fla.
Publisher:
University of Florida
Publication Date:
Copyright Date:
1976
Language:
English
Physical Description:
xii, 129 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Acetates ( jstor )
Alkalies ( jstor )
Anions ( jstor )
Cations ( jstor )
Ethers ( jstor )
Ions ( jstor )
Molecular weight ( jstor )
Picrates ( jstor )
Polymers ( jstor )
Sodium ( jstor )
Cations ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Polymers and polymerization ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Abstract:
The alkali cation binding of nonelectrolyte polyether type polymers in several nonaqueous solvents was measured by viscometric, spec tropho to.me tr i c , potentiometric and conductance techniques. An attempt was made, to correlate these findings to the ability of those polymers to catalyze nucleophilic substitution reactions to acetate salts with n-butyl bromide. UV-visible studies indicated that the complexation of sodium salts to polyethers in chloroform increased with increasing chain length up to tetraglyme, which contains five oxygen atoms. Further increase in chain length had no apparent effect on the complexing power of the polyethers. Viscometric and potent io metric studies indicated that in these systems many cations were bound to a single high molecular weight chain. These studies also showed that cation binding initially led to chain expansion due to chain bound cation-cation repulsion. Addition of more salt, however, gave a reduction in viscosity as more anions within the polymer domain shielded the positive charges. From conductance results it was shown that the number of cations bound per chain increased with increasing molecular weight. Further addition of polymer led to a redistribution of cations from polymer chains containing more than one cation to chains containing no cations. Binding of cations was strongest in THF, followed by acetoni tri 1 e , and acetone. Binding constants were calculated for the complexation of NaBPh to PEO 1000 and PEO 6000 in aceton i tr i 1 e . Distribution equilibria were used to study the interaction between alkali cations and poly (aery 1 oyltyroci di ne ) , a polymer containing a cyclic antibiotic. It was found that the ion selectivity of the polymer was similar to that of tyrocidine itself. The effect of po ly ( ethyl ene oxide) and similar polymers in the pha se- transfer reaction of sodium and potassium acetate with n-butyl bromide in acetonitrile was investigated as well as the corresponding homogeneous reaction in 5% H 2 / 9 5 ?i CH 3 CN (v/v) medium. In both cases a correlation was found between the complexing abilities and catalytic efficiencies of the lower molecular weight polyethers, although the effects were more pronounced showed that cation binding initially led to chain expansion due to chain bound cation-cation repulsion. Addition of more salt, however, gave a reduction in viscosity as more anions within the polymer domain shielded the positive charges. From conductance results it was shown that the number of cations bound per chain increased with increasing molecular weight. Further addition of polymer led to a redistribution of cations from polymer chains containing more than one cation to chains containing no cations. Binding of cations was strongest in THF, followed by acetoni tri 1 e , and acetone. Binding constants were calculated for the complexation of NaBPh to PEO 1000 and PEO 6000 in aceton i tr i 1 e . Distribution equilibria were used to study the interaction between alkali cations and poly (aery 1 oyltyroci di ne ) , a polymer containing a cyclic antibiotic. It was found that the ion selectivity of the polymer was similar to that of tyrocidine itself. The effect of poly ( ethyl ene oxide) and similar polymers in the pha se- transfer reaction of sodium and potassium acetate with n-butyl bromide in acetonitrile was investigated as well as the corresponding homogeneous reaction in 5% H 2 / 9 5 ?i CH 3 CN (v/v) medium. In both cases a correlation was found between the complexing abilities and catalytic efficiencies of the lower molecular weight polyethers, although the effects were more pronounced
Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 124-128.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
Ronald E. Cambron.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
025790286 ( AlephBibNum )
03324919 ( OCLC )
AAV1380 ( NOTIS )

Downloads

This item has the following downloads:


Full Text













ALKALI CATION BINDING TO NONIONIC POLYMERS


By
RONALD E. CAMBRON












A DISSERTTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMETNT OF THE REQUIREMENTS FOR THE
DEGREE OF UCCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1976

































Copyright 1976

by

Ronald E. Cambron















ACKNOWLEDGEMENTS


The author wishes to express thanks to his com-

mittee, Dr. Thieo Hogen Esch, chairman, Dr. George Butler,

Dr. Edmond Gabbay, Dr. Roger Bates, and Dr. Dinesh Shah,

for their assistance and support in the completion of

this research project. In addition, a debt of gratitude

is owed to the secretary, postdoctoral associates, and

fellow graduate students of the fourth floor of SSRB

for their understanding and moral support. Other people

who deserve thanks are the library staff and the personnel

of the glass shop, electronics shop, and machine shop for

their invaluable assistance with the technical aspects of

this project.

Special thanks are extended to the author's wife,

Nelda, for her encouragement and support during this

period of graduate study.




















TABLE OF CONTENTS








ACKNOWLEDGEMENTS ----------------------------------- liii

LIST OF TABLES ------------------------------------- vi

LIST OF FIGURES ------------------------------------ viii

ABSTRACT ------------------------------------------ x

CHAPTER

I INTRODUCTION -- -- - -- - - 1


Monomeric Complexing Agents ---------------- 1
Polymeric Comiplexing Agents ---------------- 10
Objectives --------------------------------- 12

II CATION BINDING ----------------------------- 18

Results ------------------------------------ 20
Viscomnetry ----------------------------- 20
Extraction Equilibria ------------------ 23
Potentiometry -------------------------- 27
UV-visible Spectrometry ---------------- 30
Conductance ---------------------------- 39
Discussion of Glymes and Poly(ethylene
oxides) ------------------------------------ 61
UV-visible Spectroscopy ---------------- 61
Viscometry ----------------------------- 63
Potentiometry -------------------------- 68
Conductance ---------------------------- 69
Discussion of Poly(3,6,9,12,15-pentaoxa-1-
heptadecene) ------------------------------- 87
Solvent Effects ------------------------ 89
Cation Effects ------------------------- 90
Discussion of Poly(acryloyltyrocidine) ----- 90





Pag



III PHASE TRANSFER REACTIONS -------------------- 91


Homogeneous Reactions ----------------------- 94
Results --------------------------------- 94
Discussion ------------------------------ 98
Heterogeneous Reactions --------------------- 99
Results --------------------------------- 99
Discussion ------------------------------ 105

IV EXPERIMENTAL PROCEDURES --------------------- 108


Preparation and Purification of Materials --- 108
Solvents -------------------------------- 108
Glymes, PEO, and PVP -------------------- 108
Vinyl Glyme ----------------------------- 109
Crown Ethers ---------------------------- 109
Antibiotic Comlpounds -------------------- 110
Salts ----------------------------------- 114
Cation Binding Measurements ----------------- 116
Viscometry ------------------------------ 116
Distribution Equilibria ----------------- 117
Potentiometry --------------------------- 118
UV-visible Spectromietry ----------------- 119
Conductance ----------------------------- 120
Reactions ----------------------------------- 121
Reaction Mlethods ------------------------ 121
Product Analysis ------------------------ 122

REFERENCES ----------------------------------------- 124


BIOGRAPHICAL SKETCH --------------------------------- 129
















LIST OF TABLES


Table Page


1 Extraction of Alkali Picrates from
Aqueous Layer into Chloroform Layer
Containing Poly(acryloyltyrocidine) 26

2 Logarithms of Stability Constants of
Crown Ethers Added to NaC1 in Methanol 27

3 Binding of 1.0 X 103 M KC1 by PEO
(molecular weight 100,000) in Methanol 29

4 Absorption Maxima of Picrate Salts in
Chloroform in the Presence of Excess
Crown Ethers 31

5 Binding Constant of Tetraglyme Addi-
tion to 1.0 X 10-4 M Sodium Picrate
in 10% THF/90% CHC13 from Benesi-
Hildebrande Plots 37

6 Absorption Maxima After Addition of
Ethers of Various Molecular Weights
to 1.0 X 10-4 M Sodium Picrate in
10% THF/90% CHC13 38

7 Decrease in Conductance Due to
Polymier Viscosity 45

8 Conductance of Polymers in CH3CN 46

9 Conductance Studies of Polyethers
with 1.0 X 10-4 M Tetraphenylboride
Salts 54

10 Solvent Effects on Conductance of
P(PHD) Added to 1.0 X 10-4 M NaBPh4
Solutions 55

11 K1 for NaBrh4 PEO 1D000 and 6000 in
Acetonitrile 75









Table Page


12 Binding Constants for Complexation of
Nat to PEO 600J0 in CHl3C 81

13 3.7 X 102 M Buty1 Bromiide + 9.6 X 102
M Sodium Acetate in Acetonitrile with
H20 Added 95

14 3.7 X 102 M1 Buty1 Bromide + 9.6 X 102
I! Sodiumi Acetate Catalyzed by 0.18 M
Polyethers in Monomer in 5% H20/95%
CH3CII 97
15 Polyether Catalyzed Reactions of
3.7 X 10-2 MI Buty1 Bromide + 9.6 X 10-2
M Sodium Acetate in Acetonitrile 101
-2
16 Reaction of 3.7 X 10 M Buty1 Bromnide
with 9.6 X 10-2 P1 Sodiumi Acetate in
CH13CN Catalyzed by Various Polymers 102
-2
17 Catalysis of 3.7 X 10 N Buty1 Bromide
+ 9.6 X 10-2 F1 Sodium and Potassium
Acetate in Acetonitrile 104


V11















LIST OF FIGURES


Figure Page


1 Typical Crown Ethers 4

2 Cation Complexing Agents 16

3 The Antibiotic Tyrocidine 17

4 Viscosity of PEO with K~r Added 21

5 Absorption Maxima Upon Addition of
Tetraglyme to Sodium Picrate 33

6 Benesi-Hildebrande Plot for Initial
Tetraglyme Additions to NaBPh4 35

7 Benesi-Hildebrande Plot for Large
Tetraglyme Additions to NaBPh4 36
-4
8 Conductance of 1.0 X 10 M NaBPh4
+ 18-crown-6 in CH3CN 40
-4
9 Conductance of 1.0 X 10 N~aBPh4
18-crown-6 in Acetone 41

10 Conductance of 1.0 X 10- M Tetra-
phenylboride Salts + Low Amounts P(PHD)
in CH13CN at 25, 48
-4
11 Conductance of 1.0 X 10 11 NaBPh4
Low Amounts P(PHD) in THF at 250 49

12 Conductance of 1.0 X 104 M NaBPh4
+ Low Amounts P(PHD) in Acetone at
250 50

13 Conductance of 1.0 X 10- M Tetra-
phenylboride Salts + Large Amounts
P(PHD) in CH3CNI at 250 51


V111









Figure Pae

14 Conductance of 1.0 X 10- Ml NaB3Ph4
P(PHD) in THF at 250 52
15 Conductance of 1.0 X 10- M NlaBPh +
P(PHD) In Acetone at 250 53
16 Conductance of 1.0 X 10- M1 NaBPh
After Initial PEO Additions in CH3CN
at 250 59
-4
17 Conductance of 1.0 X 10 PEO in CH3CN
at 250 60

18 Conductance of 1.0 X 10 M NaBPh4
PEO 900,000 in CH3CN at 250 65

19 Effects of Alkali Metal Salts on PEO
Viscosity 66

20 A/A(1/R) vs. 1/[P] for PEO 1000 K
Calculation 74

21 A/a(1/R) vs. 1/[P] for PEO 6000 Kq
Calculation 76

22 1/R vs. 1/[P] for PEO 6000 K2 Calcu-
lation 82

23 Buty1 Bromide + Sodium Acetate in 5%
H20/95% CH3CN After 5 Hours 96
24 Buty1 Bromide + Sodium.Acetate in
CH3CN After 3 Days 103
25 Infrared Spectrum of Poly(acryloyl-
benzotriazole) 113

26 Absorption of (A) Tyrocidine-HC1;
(B) Poly(acryloyltyrocidine) in
Methanol 115











Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



ALKALI CATION BINDING TO NONIONIC POLYMERS


By

Ronald E. Cambron

December, 1976


Chairman: Thieo E. Hogen Esch
Major Department: Chemnistry


The alkali cation binding of nonelectrolyte

polyether type polymers in several nonaqueous solvents

was measured by viscometric, spectrophoto~metric, poten-

tiometric and conductance techniques. An attempt was

made to correlate these findings to the ability of those

polymers to catalyze nucleophilic substitution reactions

to acetate salts with n-buty1 bromide.

UV-visible studies indicated that the complexa-

tion of sodium salts to polyethers in chloroform increased

with increasing chain length up to tetraglyme, which con-

tains five oxygen atoms. Further increase in chain length

had no apparent effect on the completing power of the

polyethers. Viscometric and potentiometric studies indi-

cated that in these systems many cations were bound to

a single high molecular weight chain. These studies also









showed that cation binding initially led to chain expan-

sion due to chain bound cation-cation repulsion. Addition

of more salt, however, gave a reduction in viscosity as

more anions within the polymer domain shielded the posi-

tive charges. From conductance results it was shown that

the number of cations bound per chain increased with in-

creasing molecular weight. Further addition of polymer

led to a redistribution of cations from polymer chains

containing more than one cation to chains containing no

cations. Binding of cations was strongest in THF, fol-

lowed by acetonitrile, and acetone. Binding constants

were calculated for the complexation of NaBPh4 to PEO
1000 and PEO 6000 in acetonitrile. Distribution equilibria

were used to study the interaction between alkali cations

and poly(acryloyl~tyrocidine), a polymer containing a

cyclic antibiotic. It was found that the ion selec-

tivity of the polymer was similar to that of tyrocidine

itself.

The effect of poly(ethylene oxide) and similar

polymers in the phase-transfer reaction of sodium and

potassium acetate with n-butyl bromide in acetonitrile

was investigated as well as the corresponding homogeneous

reaction in 5% H20/95% CH3CN (v/v) medium. In both cases

a correlation was found between the completing abilities

and catalytic efficiencies of the lower molecular weight

polyethers, although the effects were more pronounced










in the phase-transfer reactions. Polyethers of molecular

weight 600 and 1000 were found to be the most effective

catalysts, the reaction yields decreasing as the molecu-

lar weight increased to 6000 and beyond. The smaller

differences with the various polymers in homogeneous

media were probably due to the effect of water on anion

nucleophilicities and cation coordination. In hetero-

geneous media the differences may have been due to a

higher efficiency in transport of salt from the crystal

into the solution phase, a less efficient complexation

of salt by higher molecular weight polymers, or a combina-

tion of these effects.















CHAPTER I

INTRODUCTION



The complexation of alkali cations by neutral or

nonionic molecules is a phenomenon which has been ex-

tensively studied. Depending on such factors as cation

size and charge, the size, shape and distance of the

negative counterion, the ability of the solvent to sol-

vate either or both of the ions, the nature of the

binding site, and the temperature, the degree of binding

may vary over a wide range.



M~onomneric Complexing Agenits


A series of polyglycol dimethyl ethers or glymes

of the general formula CH30(CH2CH20)xCH3 was investi-

gated with respect to their behavior in solvating alkali

cations.1, It was found for compounds ranging from X = 1

to X = 6 that coordination complexes were formed with

lithiumn, sodium, and potassium salts of fluorenyl car-

banions in low dielectric constant solvents. The com-

plexation of these salts with glyme led to contact ion

pairs, glyme-separated ion pairs, or to a mixture of

both depending on the size of the cation and the chain









length of the glyme. By adding different quantities of

glyme to fluorenyllithium in dioxane or fluorenylsodium
in tetrahydrofuran (THF), it was possible to measure the

equilibrium constants of the reaction



F M + nG F ,M ,Gn K1


F ,M ,Gn + mG F ,GnlmMi K2


where n and m are the number of glyme molecules partici-

pating in the complexation, F ,M ,Gn is the contact ion

pair, and F ,Gn~mMt is the solvent separated ion pair.
As the number of oxygens in the chain increased, the var-

ious comiplexation constants increased up to an X value

of three for lithium and four for sodium where they

leveled off. Temperature dependence studies indicated

that glymes were more effective than solvents such as

tetrahydrofuran in coordinating alkali ions due to a

smaller loss in entropy. This small entropy loss is

expected because one glyme molecule contains several

oxygen atoms all of which may coordinate with a cation

of appropriate size, whereas several THF molecules would

be needed to occupy the same coordination sites. It has

also been shown that crystalline 1:1 complexes of

NaBPh4 were obtained with glymes having X = 4, 6, and 7
as well as a 2:1 complex with the compound having X = 3.3










The effect of the small entropy loss on binding

observed in the polyglycol dimethyl ethers was even more

pronounced in the macrocyclic polyethers, or crown ethers,

first described by Pedersen in 1967. These compounds

are cyclic in nature composed of repeating (-0-CH2-CH2 n,

units (see Figure 1). In these molecules the coordinating

oxygen atoms are positioned for favorable complexation

of an ion of suitable size. On cation binding, little

entropy loss is experienced. Containing a central hydro-

phillic cavity with a hydrophobic exterior, these crown

ethers were shown to solubilize certain alkali cation

salts in aromatic or chlorinated hydrocarbons. These

complexes as well as those of the polyglycol dimethyl

ethers were believed to be due to ion-dipole interactions
between the cation and ehroyns4,5

Frensdorff further studied the stability constants

of these complexes by potentiomietry using cation-selective

electrodes. Due to water's stronger solvation of the

cation, he found the binding constants in methanol to be

significantly higher than in water. Also, as the ring

size varied, cation selectivity changed so thiat an indi-

vidual crown ether would most efficiently bind the cation

which would just fit into the center cavity. The complexa-

tion of large cations by polyethers having small cavities

was increased in several instances by the formation of a

2:1 crown-cation complex. Replacing the oxygens with











C CO/C0 \C





C~C
\C,0\g;


Dibenzo-18-Crown-6


Poly(vinylbenzo-15-Crown-5)
n=3


Figure 1. Typical Crown Ethers


0--CH 2 CH2


[CH -- CH- 0]n




18-Crown-6
n=4









nitrogen or sulfur weakened the completing ability toward

alkali cations.6 Furthermore, Frensdorff quantitatively

described the ability of crown ethers to extract alkali

salts from an aqueous layer into a nonaqueous layer con-

taining the crown ether.7

Following these initial studies, an extremely

large number of multidentate macrocyclic compounds was

synthesized. These included cyclic polyethers, alky1

and aromatic substituted cyclic ethers, macrocycles con-

taining nitrogen donor atoms, sulfur donor atoms and

mixed donor atoms. In almost every case, these compounds

were shown to complex one or more cations and were usually

specific for a particular cation.8

Crown-complexed salts generated substantial inter-

ests due to their ability to exist as separated ion pairs

in solution.' Takaki, Hogen Esch and Smid studied the

optical spectra of crown ether completed fluorenylsodium

and fluorenylpotassium in tetrahydrofuran and tetrohydro-

pyran (THP). The salts were shown to exist as a mixture

of contact and crown-separated ion pairs, the stability

of the complexes being a function of temperature, solvent,

size of the cation and structure of the crown comipound.10

The cyclic ligand concept was extended in a series

of macroheterobicycles called cryptates, designed by

Lehn.111 As a result of their three-dimiensional cage

structures, which allow cations to be included within the









central molecular cavity, these compounds exhibited bind-

ing constants even greater than the crown ethers. Diffi-

cult synthetic procedures and relative unavailability have

limited their use at this timie.

Another group of compounds which complex with

alkali ions is called the ionophores. These are anti-

biotics which form lipid soluble complexes with alkali

metal salts and then transport these ions across lipid

barriers including artificial biological membranes.

Having a cyclic peptide structure, the ionophores complex

cations by ion-dipole interaction with carbonyl or ether

oxygens. The oxygens replace all or some of the solvent

molecules in the solvation sphere of the aqueous ions.

The complex then assumes a conformation in which the

charged cation is held in the center of the ionophore

while alky1 groups formi a hydrophobic outer surface

allowing the entire molecule to pass through the low

dielectricc environment of lipid membranes.151 This

ability to transport ions through membranes has also been

demonstrated by crown ethers.

Many antibiotics have been shown to complex

cations and transport them through cell mitochondria.

Among these are certain depsipeptides, macrotetralides,

and polypeptides including valinomycin, the macrotetra-

lide actions, the enniatins, and the gramicidins; and

also monocarboxylic polyethers, including nigericin,









dianmycn, he onesins, X-206 and X-537.161 Various

methods have been used to study the binding strength and

selectivity of various antibiotics. Extraction of cations

from an aqueous layer into a nonaqueous layer containing

the antibiotic was used in some of the initial studies.15,19

Other methods used included measuring the change in con-

ductance across lipid membranes due to antibiotic modi-
fied caintasot20,21 ion-selective electrodes,19

circular dichroism,2 and fluorescence spectroscopy.2234

As a result of these and other studies, a striking resem-

blance between antibiotics and certain crown ethers was

noticed in their selectivity toward cations. It was then

realized that crown ethers could serve as model compounds

for the large cyclic antibiotics in examining the processes

involved during cation transport in the miitochondria.2

The property of various ligands, particularly
crown ethers, which enables them to solubilize salts in

nonpolar media and to decrease the cation-anion inter-

action in ion pairs has resulted in their use as catalysts

in synthetic reactions. The process of phase-transfer

catalysis was first investigated by Starks in the dis-

placement reactions of alky1 halides in an organic phase

with inorganic ions in an aqueous phase.26 Although

these reactions are initially inhibited because of phase

separation, they were catalyzed by small amounts of organic

soluble tetraalky1 ammonium or tetraalky1 phosphonium









salts. This process was believed to involve the complexa-

tion of nucleophilic cyanide anions in the aqueous phase

by quaternary cations and the subsequent transfer of these

ions into the organic phase. Here the cyanide anions re-

acted with the alky1 substrates displacing halide ions.

The halide ions were then completed by the quaternary

cations and carried into the aqueous layer. Replacement

of the halide ion by another nucleophilic cyanide anion

led to further phase-transfer catalysis. The reaction

between sodium cyanide and 1-bromnooctane may be described

as follows:



R'Br + R4N CN~ R CN + R4N Br~ (organic phase)


NaBr + R4N~ CN NaCN + R4N Br (aqueous phase)


The nucleophilic displacement reactions have also been

carried out in two phase organic-aqueous systems using

various alky1 mesylates and halides as substrates with

crown ether-complexed alkali halides as nucleophiles.27

The limitation of quaternary ammonium or phosphonium

catalysts was that they must generally be used only with

the transfer of ionls from an aqueous phase into nonpolar

solvents. However, crown ethers have the ability to pro-

mote a direct solid-liquid phase-transfer of salts into

nonpolar solvents. Liotta demonstrated that in benzene









or acetonitrile, 18-crown-6 could complex potassium

fluoride, acetate, cyanide, or azide which in turn would

partially free the anion to act as a strong nucleophile

toward alky1 halides.283 In a similar study, Zubrick,

Dunbar, and Durst examined the reaction of benzyl chloride

with potassium cyanide in acetonitrile containing 18-

crown-6. They found 90-95% reaction and even when 1-2%

water was added, they still achieved an 85-95% yield.32

Other investigations have shown that crown ether-complexed

KMnO4 in benzene could oxidize olefins, alcohols, and
aldehydes,33 and aromatic hydrocarbons containing acidic

hydrogens (e.g., fluorene) could be alkylated by reaction

with aqueous sodium hydroxide in the presence of dibenzo-

18-crown-6.34

The effectiveness of crown ethers as reagents

for solid-liquid phase-transfer reactions can be attribu-

ted to their powerful cation binding ability and the fact

that they are flexible molecules with several polar

sites.3 Their flexibility allows the crown ethers to

interact with the surface of a crystal lattice of a salt

in the most energetically favorable geometry. By assuming

the same shape as the surface of the crystal lattice, the

crown ether can readily transfer a cation from its lattice

site to the crown ether cavity. The anion remains as an

ion pair with the cation complex.










Polymeric Complexing Agents


The complexes which have been described up to this

point involve low molecular weight species where the ligand

usually functions in a 1:1 relationship with the alkali

cation. With polymeric completing agents, however, sev-

eral ions or ion pairs miay be bound to one polymer mole-

cule. In addition to the entropy effects mentioned for

glymes and crown ethers, other effects are also present

in the binding of cations to nonionic polymers. One

effect is polymer conformation which determines the

proximity of binding sites to one another and, therefore,

has a direct effect on the binding process. As ions be-

come bound along the polymer chain, the polymer becomes

in effect a polyelectrolyte. The cation binding may then

be reduced due to charge repulsions by cations already

bound to the chain. The repulsions may be decreased by

counterions in the vicinity of the polymer shielding the

charges.3 Other factors may involve dipole-dipole inter-

actions in the polymer domain between ion pairs on the

chain and ion pairs in solution. These factors may in-

crease or decrease ion binding to the polymer. The

effects listed above are quite complicated and are not

well understood.

These effects were studied with poly(ethylene

oxide), the polymeric analogJ of the previously mentioned

glymes. Viscosity studies by Lundberg, Bailey, and









Callard showed that 0.05 M potassium fluoride added to

2%/ poly(ethylene oxide) in methanol resulted in the

binding of 1 potassium fluoride molecule per 9 ethylene

oxide repeating units. Liu noted cooperative effects

in an NMR study of the binding of potassium iodide to

poly(ethylene oxide) in methanol. The polymeric type

interaction began with the oligomer having seven repeat-

ing units.38 Another approach involved visible spectros-

copy of alkali metal fluorenyls completed to poly(ethylene

oxide). Here again, the strongest complex was formed at

an ethylene oxide unit to metal ion ratio of approximately

5:1 for sodium and 7:1 for potassiumi.39

Further examples of polymieric cooperative effects

were demonstrated by Smid with the syntheses and cation

binding studies of poly(vinyl crown ethers). Using opti-

cal spectroscopy, distribution equilibria, conductance,

viscosity and ion transport through liquid membranes,

he was able to compare the cation binding properties of

polymers containing pendant crown either ligands with

crown ethers themselves. In miost cases, the crown ethers

which formed 1:1 crown-cation complexes demlonstratedd equal

to slightly better binding of cations in the polymeric

form. However, for those crown ethers capable of forming

a 2:1 crown-cation complex the complexation constants for

the polymeric crown ethers were considerably enhanced over

the monomeric species.404









The polymeric cooperative effect is apparent not

only with cation binding but also with increased reaction

rates where polymeric materials serve as catalysts. It

has been shown that the Williamson reaction between

sodium phenoxide and buty1 bromide in dioxane occurred

over 100 times faster in the presence of poly(vinyl

pyrrolidone) as in a solution containing the same concen-
tration by weight of N-methyl prodne45 Hr h

rate enhancement was attributed to the dissociation of

sodium phenoxide based on the tight solvation of sodium

cations by pyrrolidone segments in the polymer coil.

From entropy considerations, it would be a lower energy

process to solvate a sodium cation with pyrrolidone

groups attached to a polymer than with free pyrrolidone

residues. This should lead to higher rates with polymeric

catalysts.



Objectives

Catalysis of synthetic reactions by polymers is

of both practical and theoretical interest. Polymers

may be relatively easily retrieved from homogeneous reac-

tion mixtures by various means which would allow their

reuse as catalysts. Polymiers may also catalyze reactions

in heterogeneous fashion in the form of microporous beads

or gels which would immobilize the polymer in the solid

state. Here again, separation from the reaction mixture





is easily achieved. As mentioned previously, polymers

have the ability to complex cations through a variety of

effects. Frequently such complexations are as strong or

stronger than with the corresponding low molecular weight

analogs. Polymiers, in addition, miay exist in conforma-

tions which would possibly allow completing sites to inter-

act with solid phase reagents in a more energetically

favorable manner than low molecular weight compounds.

In spite of these advantages of polymer catalysts,

there have not at this time been many reported investiga-

tions of polymers in phase-transfer reactions. Phase-

transfer catalysts may serve a dual role. They may not

only transfer a reactive substance, usually a salt, from

a solid or aqueous phase into an organic layer, but they

may also modify this reactant once it is in the organic

phase to increase its reactivity. How this is done, how-

ever, has not been clarified in detail.

It is believed that by studying cation binding

and catalytic properties of a series of phase-transfer

catalysts of varying degrees of polymerization, more

insight may be obtained in the nature of the phase-

transfer process of both low and high molecular weight

compounds. It would also be of interest to study the

correlation of the cation binding and catalytic properties

of various other polymers in order to assess the role of

various cation completing sites. Such studies may be of





help in clarifying the nature of polymer catalysis in

phase-transfer reactions since at least for solid-liquid

phase-transfer processes, the nature of the catalysis

is not well understood.

Therefore, the objectives of this research pro-

ject were as follows:

1. Study the ability of various nonionic poly-

mers, having similar binding sites but dif-

fering in the location of these sites, to

bind alkali cations and to modify the ion

pair structure of their salts. This was

done by conductance and visible spectro-

scopy.

2. Study the phase-transfer ability of these

polymers as a function of molecular weight,

polymer concentration, and solvent composi-

tion.

The polymiers chosen for the phase-transfer cataly-

sis work were ethers. Poly(ethylene oxide) which contains

ether oxygens in its backbone was readily available in

various molecular weights. Various glymes, low molecular

weight analogs of poly(ethylene oxide), were also commer-

cially available. The vinyl glyme 3,6,9,12,15-pentaoxa-

1-heptadecene containing a pendant group with five ether

oxygens was commercially available, easily polymerized,

and the polymer was soluble in most common solvents.










This material in polymer form was investigated and com-

pared to the previously listed ethers. Examples of the

polyethers are given in Figure 2.

Most common cyclic antibiotics which were known

to bind cations had no suitable functional group to aid

in attachment to a polymer or the group present was in-

volved in cyclization of the antibiotic during the com-
plexation poes15,16,18,19,21,46,47,48 h yli ni

biotic tyrocidine contained the ornithine amino acid

function which included a primary amine group, facilitat-

ing attachment to a polymer chain. Also, tyrocidine bound

alkali cations, particularly sodium, and it was also com-

mercially available.49 Therefore, the antibiotic tyro-

cidine met the selected criteria, was attached to a poly-

mer, and was used in the study of alkali cation binding

to various polymers. The antibiotic structure is given

in Figure 3.










CH3-0 -E CH2- CH2 0 J-- CH3

Tetragl yme



{ECH2-- CH2-- OJ-

Poly(ethylene oxide)



-[CH2-- CH}-



OCHj
1H2
CH3



CC2-H~
CHO


Poly(36,9,215-petroxa-1-heptdcee


Figure 2. Cation Complexing Agents









Tyrocidine A

H CH2 H
CH3 PCH3 CH2N CH P~CH3 O \2
CHi2 PH2 CH2 CH2 C2 CH2
H--C-N-C-C-N-C-C-N-C-C-C-N -C-C-N-C-C
IlII Il IlllIIII Ill
HHOHHOHHOHHO HO



l I I 1 I I l I I II I I 11 I I
-C-C-N-C-C-N-C-C-N-C-C-N-C-C-N-
CH2CH2 CH2 CH2 H


OH
X Y




Tyrocidine B X = --CH2


Y = --CH2


Tyrocidine C X = -C
20

Y = --CH2N


Figure 3. The Antibiotic Tyrocidine













CHAPTER II

CATION BINDING


Many different techniques have been used to study

the complexation of cations to nonionic ligands. The

vibrations of alkali cations encaged by crown ethers in

solution have been examined by far infrared spectros-

copy.5 In this region, solvent- and anion-independent

bands were used to determine the effects on the cation

motion frequencies in solution caused by ion binding.

Proton and 2Na magnetic resonance have been used to

study the chemical shifts induced wrhen a cation is coor-

dinated to a cyclic ether.515 Also, calorimetry has

proved useful in determining binding constants as well as

enthalpy and entropy values for cation complexation.535

Another valuable tool for calculating binding constants

was potentiometry with cation-selective electrodes.6

This measured the number of free ions remaining in solu-

tion after a ligand had been added. The use of alkali

cation salts of fluorescent probes or even the fluores-

cent thalliumn cation itself made possible the use of

fluorescence spectroscopy in the study of cation binding

to various liaands, particularly cyclic antibiotics.232

In polymer systems, light scattering and viscosity have









been used to examine the polyelectrolyte effect of chain

expansion upon binding of cations along the polymer

chain.363 Distribution equilibria have been useful in

demonstrating the ability of monomeric or polymeric com-

plexing agents to extract alkali cation salts from an

aqueous phase into an organic layer where the salt would

normally be insoluble.471214 An alternative to

studying parameters directly concerned with the cation

was to study the UV-visible spectra of salts with anions

such as picrate or fluoreny1. These salts were sensitive

to changes in the interionic ion pair distances caused by

complexation of the cation, making them valuable probes

of the binding strength of various ligands.555 Another

method which has been used to determine binding constants

as well as the stoichiometry of the cation complex is

conductance.435 In solvents where the alkali cation

salts existed as free ions, the conductance of the solu-

tion was shown to decrease upon addition of ligands such

as crown ethers as long as the mobility of the completed

ion was lower than that of the noncomplexed ion. This

decrease continued until nearly all the cations were

completed. At this point the conductance began to level

off with the breaking point yielding the stoichiometry

of the complex. In the present study, the following

techniques were used: viscosity, distribution equilibria

on one polymer which had the appropriate solubilities,









cation-selective electrodes, UV-visible spectroscopy,

and conductance.



Results


Viscometry


Lundberg, Bailey, and Callard examined the effects

on viscosity of adding alkali metal halides to poly(ethy-

lene oxide) solutions in methano1.3 They observed high

viscosity with moderately high salt concentration (0.02

M potassium iodide) with a decrease in viscosity upon

further addition of salt. From their results, they calcu-

lated that one molecule of salt associated with approxi-

mately nine ethylene oxide units. The anion was tenta-

tively postulated as the species directly bound to the

polymer. The present study expanded that of Lundberg

et al., particularly to lower salt concentrations and

different cations, to further investigate the polyelectro-

lyte effects caused by binding cations along the polymer

chain, as well as to elucidate the actual ion completed

to the polymer.

To a methanol solution containing 2.5 X 10- M

in monomer of poly(ethylene oxide) of molecular weight

6 X 106 g/mole was added potassium bromide over a concen-

tration range of 1.0 X 104 Ml to 1.0 X 10- M. The

results are shown in Figure 4. Initial additions of KBr




































(d
w
o
u

o

m
r
u

O
O ~
x
~1
Y
uO
w ri
u
rr
LL
O
Q


a,






Oc








O

x


Ni d y
mn ro r
r- o
O O O

dsh










are followed by an increase in viscosity. Here, cations

which we proved to be bound along the polymer backbone

were involved in charge-charge repulsions, expanding the

chain and increasing the viscosity. However, after passing

through a maximum at 2.5 X 103 M KBr, the viscosity began

to decrease. The further addition of salt led to a vis-

cosity decrease caused by counterion shielding. Division

of the concentration of monomer units of poly(ethylene

oxide) by the salt concentration at maximum viscosity

indicated that at that point there was binding of one

cation per 10 monomer units. Similar results were found

by Lundberg, Bailey, and Callard.

In order to be able to evaluate qualitatively the

role of anion shielding, the measurements were repeated

in the presence of varying amounts of the crown ether

dicyclohexy1-18-crown-6. Since cation binding to crown

ethers is very much stronger than to poly(ethylene oxide)

(see Conductance section), the free cation concentration

could be conveniently varied while keeping the anion con-

centration constant. The resulting plot is also shown in

Figure 4. The maximum was decreased and displaced from

S= 1.007 at 2.5 X 103 M KBr with no crown ether to

n = 0.732 at 4.6 X 10- M KBr with crown ether. The value

of 837 sec at 0.0 M free K+ where [KBr] = [CE] was exactly

the same value obtained with the solution of PEO in metha-

no1 with no KBr added.









The contribution of the bromide~anion in these

binding phenomena was further examined by studying the

addition of 2.0 X 10- M tetramethylammonium bromide to

2.5 X 10- M in monomer poly(ethylene oxide) in methanol.

The large tetramethylammonium cation should not bind to

the polymer to a measurable extent.5 The difference

between the viscosity of the poly(ethylene oxide) solu-

tion in methanol and that of poly(ethylene oxide) plus

tetramethylammonium bromide in methanol was negligible.

This indicated that bromide ions did not bind to

poly(ethylene oxide) as had been previously proposed.

Light scattering studies have also been done

elsewhere on these systems. These measurements confirmed

the expansion of poly(ethylene oxide) upon addition of

alkali cation salts in the 1.0 X 103 to 5.0 X 103 M
58
ranae.


Extraction Equilibria


As stated previously, crown ether complexes of

alkali cation salts have been shown to be quite soluble

in certain organic solvents. This provided a method for

extracting salts from aqueous solutions into organic sol-

vents containing cyclic ethers. Extraction has been

efficient only if the anion is large and highly polariz-

able, as for instance picrate, which has the additional

advantage of absorption near 360 nm for easy determination









of the amount transferred. The overall equilibrium between

an aqueous solution containing alkali cation (M ), picrate

ion (A ), and hydroxide ion, and an organic solution con-

taining a cyclic polyether (CE) can be described as fol-

lows:79



M~l + A + CE e~ M ,CE,A
aq aq org org


where M ,CE,A~ designates ion pairs in the organic phase.

If the organic phase is polar enough and the concentration

of the salt low enough, these ion pairs will dissociate

according to


Kd
M+ ,CE, A c M ,CE + A
org org org


If there is partition of uncomplexed polyether between the

two phases, it miust be accounted for in another equilib-

rlum.

Pe
CE~r CEq


The final equilibrium takes into account comnplexation in

the aqueous phase if the concentration of crown ether'in

the aqueous phase is significant.

Ks
++
Maq +Caq MCq










Due to the solubility of all other polymers in

this study in aqueous media, the examination of cation

binding by distribution equilibria was limited to

poly(acryloyltyroci'dine), the polymer containing the anti-

biotic tyrocidine as a pendant ligand. This process, how-

ever, could only be accomplished in a semiquantitative

manner. The antibiotic tyrocidine was supplied as a

mixture of up to five separate components. Various separa-

tion techniques failed to isolate the individual antibiotic

units of the mixture. Since the components contain vary-

ing amounts of tryptophan and tyrosine amino acid residues,

which both absorb at approximately 276 nm in the UV,60

and since no other commonly used technique could identify

the amounts of the various tyrocidine compounds in the

mixture, it was impossible to determine the exact number

of antibiotics attached to a polymer chain. It was assumed,

therefore, that the antibiotic composition in the polymer

was identical with that of the reacting mixture. The UV

absorption of the polymeric antibiotic was compared with

that of the antibiotic mixture to get an estimate of the

number of antibiotic units per polymer. A value of one

tyrocidine molecule was found for every 12 monomer units.

The extraction of sodium, potassium, and cesium

pirates from an aqueous layer into a chloroform layer

containing the polymer was examiined due to the ease with

which this process could be followed by UV-visible










spectroscopy. The picrate salts were dissolved in water at

a concentration of 1.0 X 103 i1. A large quantity of the

corresponding alkali chloride was also added to each aque-

ous solution to increase the number of alkali ions in the

organic layer and also to increase the ionic strength of

the aqueous layer which would prevent the chloroform layer

from becoming dispersed into the water layer. Poly(acryl-

oyltyrocidine) was added to the chloroform layer at an

estimated concentration of 1.3 X 10- M in antibiotic

units. The amount of salt extracted is presented as a

percent of the tyrocidine units present in the organic

phase. The results are shown in Table 1. The sodium

selectivity of the polymer was similar to the 4:1 sodium

over potassium selectivity demonstrated by the antibiotic

itself in cation transport through lipid membranes.49




Table 1. Extraction of Alkali Picrates from Aqueous
Layer into Chloroform Layer Containing
Poly(acryloyltyrocidine) at 25"



Initial Initial Final Chloro-
Aqueous Layer Chloroforml Layer formi Layer
Cation [Picrate] [Chloride] Approx. [Tyrocidine] % Extraction


Na 1. X103 5.0 X 10- 1.3 X 10 -3

K 1.0 X 10- 5.0 X 10- 1.3 X 10- 2

Cs .0X 1-3 5.0 X 102 1.3 X 103









Potentiometry


The use of cation-selective electrodes to determine

the stability constants of complexes of crown ethers with

various univalent cations has proven to be an effective

method with systems having stability constants ranging

over six or more orders of magnitude.6 The electrodes

responded to uncomplexed ions in solution, therefore,

after the addition of a ligand, the drop in the potential

of the solution directly corresponded to the number of

completed ions. In order to verify the experimental pro-

cedures used in this study, binding constants were deter-

mined initially for two typical crown ethers and the re-

sults were compared to literature values. These data

shown in Table 2.


Table 2. Logarithms of Stability Constants of Crown
Ethers Added to NaC1 in Methanol



Experimental Literature
Crown Ether B3inding Constant Binding Constant

Dicyclohexy1-18-crowIn-6 3.95 Isomer A = 4.08
(mixture of isomers) Isomer B = 3.68

18-crown-6 3.89 4.32




The data shown in Table 2 indicate our system gave bind-

ing constants close to published values.









Potentiometry was then used to determine the

cation binding ability of the antibiotic itself. However,

with this system the addition of antibiotic to a salt

solution in methanol produced an increase in solution

potential instead of the expected decrease. It has been

demonstrated that cation-sensitive glass electrodes re-

spond to the presence of various amino acids;61 therefore,

the measured increase in voltage may have been the result

of electrode response to the antibiotic itself. As a

result, potentiometry was discontinued with this system.

The binding of potassium cations to poly(ethylene

oxide) in methanol was a system which could be studied by

potentiometry. After addition of the polymer to a potas-

sium chloride solution in methanol, the solution potential

dropped corresponding to the binding of potassium ions to

the polymer. The new potential was compared to a calibra-

tion curve to determine the concentration of remaining

unbound cations. From this number, the number of bound

cations was calculated and since the number of poly(ethy-

lene oxide) monomer units was known, the number of monomer

units per completed potassium ion could be determined.

The data are given in Table 3. As the concentration of

poly(ethylene oxide) increased, the number of monomer

units per cation increased. This indicated a redistribu-

tion of charges along the polymer chains resulting in

fewer cations bound per polymer. These ratios of monomer
















PEO in Mlonomner mV Uncomplexed Kf Complexed K+ Monomer/K+

36.9 1.0 X 10-3
1.4X1-2 288 61X1-4 3.9 X 10-4 29
1.83 X 10- 23.4 4.3 X 1057X1043
2.9X102 1. 27X1-4 573 X 10-4 33
2.96 X 102 93.8 1.3 X 104 87X1043
3.0X1-2 7. X1 9.1 X 10-4 36
23.4 X 10- 67.6 8.2 X 10- 043

3.58 X 10- 6.1 7.2 X 10- 9.3 X 104 38

3.87 X 10- 5.0 6.0 X 10-5 9.4 X 10 41 3
34.0 X 10-2 45 .2X0-5 95X1-4 4

4.1 X102 .0 3.0 X 10- 9.7 X 10- 43


Table 3. Binding of 1.0 X 10- M KC1
weight 100,000) in Methanol


by PEO (molecular
at 250









units per cation differ from the value of 10 calculated

from conductance studies. These factors will be dis-

cussed later.


UV-visible Spectrometry


It has been observed that salts of picric acid

in low polarity media frequently exhibited pronounced

shifts in their optical spectra upon complexation of the

cation by crown ethers.17 These shifts were most likely

due to a significant increase in the interionic distance

of the tight ion pair as a result of external coordination

of the cation. Further addition of a sufficiently strong

completing ether led to the formation of solvent-separated

ion pairs. This process may be described by the following

equation:



Pi,Na+ + CE Pi ,Na ,CE Pi ,CE,Na+

I II III


Complex I is a tight ion pair which upon coordination by

crown ether becomes the tight ion pair II which has a

slight increase in interionic distance. Further crown

ether addition leads to the loose ion pair III. Examples

of these species are described in Table 4.

The absorption maxima of the sodium and potassium

salts in THF are those of contact or tight ion pairs.









Table 4. Absorption Maxima of Picrate Salts in Chloro-
form in the Presence of Excess Crown Ethers



Salt Crown nax (nm) Type Complex


Na 351a

15-crown-5 356 II

18-crown-6 362 II


K 357a

15-crown-5 378 III

18-crown-6 365 II


aThese values refer to
these salts.


tetrahydrofuran solutions of


The same species would be expected in chloroform if the

solubility of the salts were high enough because of the

relatively low cation solvating power of this solvent.17

The presence of 15-crown-5 in chloroform solubilized the

sodium salt in the form of a crown-complexed tight ion

pair (Amx 356 nm), resulting in a small increase in the

interionic distance. Addition of a large excess of 15-

crown-5 to the potassium salt converted this species to

a loose ion pair (Ama 378 nmn) which existed as a 2:1

complex. With 18-crown-6, both salts formed tight ion

pairs (Amx 362 to 365 nm) with the interionic distance

slightly larger than with the 15-crown-5 sodium complex.









In order to examine the various solvation pro-

cesses involved when a typical ether ligand is added to

sodium picrate in chloroform, a solution of 1.0 X 10-4 M

sodium picrate in 10% THF/90% CHCl3 was titrated with

bis [2-(2-methoxy)ethyl]ether or tetraglyme. The small

amount of THF aided the solubility of sodium picrate in

chloroform but had no effect on the absorption maximum.

The results of these additions are plotted in Figure 5.

As the concentration of tetraglyme increased, the absorp-

tion maximum shifted to longer wavelengths indicating

small increases in the interionic distance of the tight

ion pair. An analogue of the Benesi-Hildebrande equation

for optical absorption allows the calculation of a bind-

ing constant from these shifts in the absorption maxi-

mum.6 The change in absorption on tetraglyme addition

from the X of sodium picrate without tetraglyme, a,
max
can be related to the total tetraglyme concentration

[TG] and the binding constant K by the equation


1 1 1 1 1
n K 80 [TG] 60n


A plot of 1/a against 1/[TG] should be linear with the

intercept at the ordinate yielding a0 and the gradient

yielding the prodnet KA e0. Dividing this product by n0
yields the binding constant K. Benesi-Hildebrande plots

for the addition of tetraglyme to sodium picrate in




























O













L-..
O~
C)







O
Cl

2-


C
O



.0
<1



III 9
mem m-
0 0 0 0
o~~ L


o OD -
U) D In In 0 I
n n m a r, n










10% THF/90% CHC13 are given in Figure 6 and Figure 7.

Results of these plots are given in Table 5. Over the

entire concentration range examined, the Benesi-Hilde-

brande plots were linear and the binding constants were

consistent. This is the first instance where shifts in

the picrate spectra due to added ligand have been used

to calculate a binding constant.

In the graph of absorption maxima vs. tetraglyme

concentration, Figure 5, a gradual increase in interionic

distance was observed up to a Amx of 359 nm. It was

expected that the addition of ligand to picrate salt

could distinguish between the coordinating ability of the

various completing polyethers under analysis. These

materials included 1,2-dimethoxyethane, bis(2-methoxy-

ethyl)ether or diglyme, tetraglyme, poly(ethylene oxide)

of various molecular weights, and the polymeric vinyl

glyme poly(3,6,9,12,15-pentaoxa-1-heptadecene). The

results of the addition of 1.0 X 103 M and 1.0 X 104 M

ligand (in mnonomer units) to 1.0 X 104 M sodium picrate

are given in Table 6. With the exception of DME and

diglyme, there was very little difference between the

coordinating ability of the various completing poly-

ethers.






















Ln










o a

















ca m


o O, m t o I
d o a ci o


a-
o o


















No

a









N




Om
0- E
NX





to o
No

-o

-






NN



or









(r,

-
LI.


comm--------oo
00000000000






























O










OC













EE



r-
OU
L
vU



C




oan0



CJ1
OLL

UQI




--











ha


-lao



,Ise



a
o

v,













-/a"
v

rJ
a

u
L

o



























ar
cn

~1



E
2,

0,

L


O


X

O
C\1


















I
o


r:

m

oo







rr,

o


x

o

N


m
o


x

C3

10













m
I
O


X

ID











Table 6. Absorption Maxima After Addition of Ethers
of Various Molecular Weights to 1.0 X 10-4 M
Sodium Picrate in 10% THF/90%, CHC13




1.0 X 10- M Ether 1.0 X 10-3 M Ether
Ether Miol. Wrt.mx (nm)mx (nm)


DME 90 351 352

Diglyme 134 351 352
Tetraglyme 222 355 358

PEO 600 353 358

PEO 1000 353 359

PEO 4000 352 357

PEO 6000 352 357

PEO 100,000 352 357

PEO 300,000 352 357

PEO 900,000 352 357

P(PHD) 40,000 353 357









Conductance

Pedersen and Frensdorff obtained information on

the stoichiometry of crown ether-alkali cation complexes

by measuring conductance of salt solutions after addition

of crown ethers.5 When the polarity of the solvent was

high enough and the salt concentration low enough to

ensure complete dissociation of the ions, the mobility

of the cation decreased as it became comnplexed to the

crown ether resulting in a decreased conductance of the

salt solution.

Where the binding constant, K, defined by



M+ + A- + CE M ,CE + A


was high, the conductance decreased linearly with an

increase in [crown ether] until comnplexation is virtu-

ally complete. Further addition of crown ether had no

effect. Such plots are shown in Figures 8 and 9. Here it

is demonstrated that a break in the curve occurs at a

crown to salt ratio equal to the stoichiometry of the

complex. Smiid has also shown that addition of a

poly(viny1 crown ether) to a salt solution reduced the

conductance more than the corresponding monomeric crown

ether because of the lower mobility of the polymeric

complex.43 Significantly in the polymer case, the

equivalent conductance decreased somewhat beyond the





40








o








O



o




O =

I C



o

elc a






O




I I---m

I O



o



o

1 1 | | | ) I 1
N o a 0 a N
L1




































O
Ox c
to


co
CO --
iO U
CL



Ox


r-,
ot

C
ou

,oo



0)


(13 O a) rL 4










stoichiometric crown ether to cation ratio. This gradual

decrease beyond the break point is probably due to redis-

tribution of cations upon addition of more polymer,4 a

point which will be discussed later.

Several studies were performed on various molecu-

lar weights of poly(ethylene oxide) and on poly(pentaoxa-

1-heptadecene), or P(PHD), in order to investigate some of

the factors involved in conductance measurements. There

are several complicating factors which can influence con-

ductance determinations: (1) Viscosity effects on ion

mobility co-uld be a problem especially with polymers. As

polymiers are added to a salt solution, the viscosity of

the medium may increase to the point where mobility of

the ions is decreased. This would result in a drop in

conductance without cations being bound to the polymer.

The relationship between conductance and viscosity of a

solution is given by Walden's Rule:63


h0 n = constant


This rule states that the product of limiting equivalent

conductance, A0, and macroscopic viscosity, n, of a sys-
tem is equal to a constant. In order for thb product to

remain a constant, an increase in macroscopic viscosity

requires a corresponding decrease in X0. (2) It is also
possible for the anion of the salt to effect the conductance

if it were completed by the polymer. In this case, a drop in









conductance of a salt solution upon addition of a polymer

would indicate a decrease in anion mobility as well as

cation mobility. (3) Another possible problem is polymer

conductance. If the polymer itself conducts, the conduct-

ance measured upon polymer addition would be higher than

the true value.

In order to examine the possible effects of macro-

scopic viscosity on conductance, a polymer was needed

which does not complex cations. This polymer could then

be added to a salt solution with any resulting decrease

in conductance being attributed to an increase in vis-

cosity. The polymer chosen was poly(styrene). The sol-

vent was THF due to the solubility of poly(styrene) in

this solvent and the fact that NaBPhq exists as free

ions in THF at low concentrations.6 The addition of

poly(styrene) up to 4.0 X 10-2 M in monomer to 1.0 X 10-5

NaBPh4 caused virtually no change in conductance. Since

viscosity studies demonstrated a 20% increase in macro-

scopic viscosity for this polymer concentration, it can

be concluded from Walden's Rule that this increase in

macroscopic viscosity due to the addition of polymer in

this low concentration range was of no consequence in the

conductance measurements.

Another method used for determining the effects of

macroscopic viscosity on conductance was adding polymer

to a solution of tetrabutylammonium tetraphenylboride.









Since the tetrabutylammonium cation does not bind to poly-

ethers (see Viscosity section), any decrease in conduct-

ance should be attributable to the lowering of ion mobili-

ties due to an increase in macroscopic viscosity. The

conductance measured upon binding of alkali cations to the

polymers could then be corrected by this value. Poly(penta

oxa-1-heptadecene) was added over a concentration range of

3.2 X 10-- 4.0 X 10- M in monomer to a 1.0 X 10- M

solution of tetrabutylammonium tetraphenylboride in THF.

No significant change in conductance was measured over

this concentration range of polymer. These results were

consistent with the various poly(ethylene oxides) studied

in this concentration range. For higher polymer concen-

trations, however, significant viscosity effects were

measured. Some examples of the viscosity studies are

shown in Table 7.

The lack of conductance decrease on addition of

low concentrations of poly(ethylene oxide) to tetrabutyl-

ammoniumi tetraphenylboride also indicated the lack of

anion binding. Any anion binding would have resulted in

a measurable decrease in conductance over the polymer

concentration range of 3.2 X 103 to 4.0 X 10- M in

monomer.

The conductance of the polymers themselves was

measured by adding the polymer to pure solvent. These

values could then be subtracted from the measured






45







1N I












E-F-III
u 0


ou<





UO











o
vr


C
L


zz
r-- D o
O 3 L
1. Gk G










conductance of the polymier-salt solutions. Examples of

the conductance of several polymers are given in Table 8.



Table 8. Conductance of Polymlers in CH3CN at 250



Pol ymer
Polymer Mlolecular Weight [Monomer] Conductance, A, in MU


PEO 1,000 5.0 X 103 1.30

1.0 X 10 18.90

PEO 6,000 5.0 X 10-

1.0 X 10-

PEO 100,000 5.0 X 103 0.10

1.0 X 10-1 2.10

PEO 900,000 5.0 X 10- 0.10

1.0 X 10-1 3.00

P(PHD) 40,000 5.0 X 103 1.30
1.0 X 10-1 22.70





An example of the effect of polymer addition to

a salt solution is shown in Figure 12. It may be con-

trasted with the conductance plot obtained by addition

of 18-crown-6, a strong cation completing agent of low

molecular weight, to a similar salt solution in Figure 8.

Both figures are characterized by two separate sections.









Initially there is a region of sharply decreasing conduct-

ance in which cation binding takes place. A comparison of

initial slopes indicates that the crown ether curve is

much steeper in this region than the polymer. Further

addition of ligand causes a gradual decrease in both

slopes. This was also found with all other polymer sys-

tems studied. The second portion of the crown ether

figure is a nearly horizontal line indicating that all

cations have been bound and that further crown ether addi-

tion has no effect on conductance. The second segment of

the polymer curve, however, continues to decrease in con-

ductance upon addition of more polymer. This is due to

a redistribution of cations among the polymer chains.

All the polymiers examined except PEO 1000 gave this con-

ductance decrease with cation redistribution. The results

of the conductance studies are given in Table 9.

Solvent effects. The effects of various solvents

on the comiplexation process were examined by adding

P(PHD) over a concentration range of 3.2 X 103 to

1.0 X 10- M in monomer to 1.0 X 104 M NaBPh4 in acetone,

tetrahydrofuran, and acetonitrile. These additions were

made over two concentration ranges: from 3.2 X 104 M

to 4.8 X 10- 11 and from 5.0 X 10- M to 1.0 X 10 M in

monomer. Examples of these results are given in Figures

10-15. The results are given in Table 10.



















1 m I
me 1 0

t




an U
1 E



O
I I












,z o
I O

I I / I


a ~~ o a =

















O
S(D
rro



O
ro
o
Lo



O -'
o

CO

Oc



O

co


OE

Oc




oi
a
C3

. I I Ci





O








00

O






No
-m e






o*







r-)
O



C1


(1J
1 1 I 1 1 I I I I 1 O
m e e s o ~ o g g g g g g g





51






8 9 o



IZ
IU

I

aa


am I o


Ox





m o


a. c
Om

1( "




Io v




O S

e-'
1111
oooooooc
row-omo





O










o





N c
OB 4


oo-o



cox

CL .c


c

Oo
0-o

GJ
U
C

r,

O
L)r


I -~- I I IX
in n a, r- In n
ct rr) ~ rr) Ic) re)
o,



















O






O









N





OE



Oro






o '


r-

O


O






O


o



lu

OOOOOOOc
acki-Om























0 00 000000 0 000
XXX XXX XXX XXX XX







>) CL 0 *- r ,- 0a U ..- *,-*-a- r-- -








O Y



o c 000000000000000000000
rO
a *axxxxxxxxxxxxxxxxxxxx
L coommomoovmemo~~~mom
-mme-*---mme-*---*- .
u E I ~ s l a i r e I I I I I ~ l i llb 0









000
00 00


O 000 0001100000000
CCOOOD ***~~
UVVL~IIUUU UVUVUVU u
co c c m es m m
zz z z o
-1 1 1 10 0 0 0













O
0,
rJC

r ~C~
OC
a r
m
cr

aru
La,

aa
aE
4~0
u









a,
a
o
--
vr



















La)
a~E
Eo
he

011
a
uC
''












U3
QI~D


O


































so









a

OO



O


m m
o o
--

x x

r? to
o ru

m Io
I I


0 'o 'o



O O O



I I 1


0 0 0




X X X

a .0


01 0



O O



I I


0 0





d m


mLDmLDCLD








03 N c,
m a~ to

~u m










o o

h O
C\1


of
z












C
a


U
Cr
O










For lower polymer concentration range, THF gave the largest

slope followed by acetonitrile and then acetone. For the

higher polymer concentration range, the largest slope was

found in acetonitrile followed by THF and acetone. The

apparent point of complete cation binding is the point at

which the conductance curve approaches a horizontal line.

Beyond this point, there is very little if any change in

conductance upon addition of more polymer. The apparent

point of complete binding was reached first in THF fol-

lowed by acetonitrile and acetone. The competition

between solvent and ligand for solvation of sodium cations

was also examined by adding 18-crown-6 to NaBPh4 in

acetone and acetonitrile, Figures 8 and 9 and Table 9.

As with P(PHD) the highest slope and earliest point of

complexation of all cations occurred in acetonitrile.

This trend does not follow that expected from comparison

of either the Gutmann donor number for the solvents or

the dipole moment. The Gutmnann donor number of a solvent

is a property which expresses the total amount of inter-

action between the solvent and the Lewis acid SbCl5'

including contributions both by dipole-dipole or dipole-

ion interactions and by the binding effect caused by the

availability of the free electron pair, and to some extent

even steric properties of the solvent molecules. Thus

the donor number is considered a semiquantitative measure

of solute-solvent interactions.6 A higher donor number









would mean a stronger solvation of a cation. The dipole

moment gives an indication of the solvent-cation inter-

actions from strictly ion-dipole considerations.66

Cation effects. Comparisons were also made

between sodium and potassium ions with P(PHD) in acetoni-

trile and THF. Salt concentrations of 1.0 X 104 M in

acetonitrile and 1.0 X 105 M in THF ensured the presence

of free ions.646 From Table 9 it may be seen that in

both solvents and for both the lower and higher polymer

concentration ranges, sodium gave a larger slope than

pota ss iumn.

Molecular weight effects. In order to determine

if there was a relationship between molecular weight and

binding as measured by conductance, several polyethers of

various molecular weights were added to 1.0 X 10-4 M

solutions of NaBPh4. The polyethers ranged from tetra-

glyme with a molecular weight of 222 to several poly(ethy-

lene oxide) samples with molecular weights of 1000, 6000,

100,000, and 900,000. Tetraglyme was added over a con-

centration range of 3.2 X 10- M to 4.0 X 102 H to a

1.0 X 10-4 M solution of NaBPh4 in acetone. The change
in conductance was negligible because the mobility of

the complexed ion was essentially the sam~e as that of the

uncomplexed ion.

The poly(ethylene oxide) samples were added to

1.0 X 10- M NaBPh4 in CH3CN. As with the P(PHD) samples









mentioned earlier, the additions were made over a low
concentration range of 3.9 X 10-4 M o48X1-3 Mi

monomer and a high range of 1.1 X 10- M to 1.0 X 10- M

in monomer. The results of PEO 1000, 6000, and 100,000

are given in Figures 16 and 17. The initial slopes were

seen to decrease with molecular weight from -5.21 X 10-

with PEO 1000, -5.16 X 103 with PEO 6000, and -4.19 X 103

with PEO 100,000 to only -4.11 X 103 with PEO 900,000.

However, the final segments of the curves, which began

to approach a horizontal line, were found at higher con-

ductance values with the lower molecular weight polymers.

At 1.0 X 10- M in monomer for each polymer, the conduct-

ance values ranged from 86.00 lil for PEO 1000 and 70.84

pa for PEO 6000, to 60.30 pi for PEO 100,000 and 60.60 vU

for PEO 900,000. Only PEO 1000 and 6000 gave curves

which actually approached a horizontal line at higher

polymer concentrations. The point at which the horizontal

line was approached came at a lower polymer concentration

with PEO 1000. For both PEO 100,000 and 900,000 the

conductance continued to decrease at high polymer concen-

trations. As stated earlier, this decrease was probably

due to a redistribution of cations along the polymer

chains. This will be discussed later.












O
**n








II I o
a ,
/7O t3
,I I ,







co











Il I O





I Ox-o

a
I O




cm


v0 0n O





60




o
a p to





o I







ooo O
000 I I. -



I o


III "
I j o



I .r-


; co
I '+

if a


;* N




I- /

o

I I I9






in N O O O O





Discussion of Glymes and Poly(ethylene oxides)


UV-visible Spectroscopy


From the spectral data in Table 6, it was deter-

mined that the various polyethers in this study coordinate

sodium cations. This was evident from the red shifts due

to an increase in interionic distance upon addition of

ligand. These shifts were almost negligible with DME and

diglyme even at higher ether concentration, but with tetra-

glyme and the polymeric ethers larger red shifts were ob-
served. The absence of distinct red shifts with DME and

diglyme indicated that these ethers have very low binding

constants with sodium picrate in the low polarity THF/CHC13
solvent. When the shifts with tetraglyme and the poly-

meric ethers were compared with the changes in Amx upon

addition of crown ethers, Table 4, it was apparent that

1.0 X 10-4 M sodium picrate in the presence of 1.0 X 10-

M glymes, PEO, and P(PHD) remained as a peripherally sol-

vated tight ion pair absorbing at approximately 358 nm

(see Results).68 In the peripherally solvated ion pair,

the ligands are able to replace most of the solvent mole-

cules from the anion paired cation, but coordination is

not sufficiently strong to cause a large increase in inter-

ionic distance as with crown ethers. This coordination

may be represented as follows.





\O O
Pit Mk + o -+ + n +->





Here solvent molecules, +-) are replaced in the cation

solvation sphere by ether oxygens of a polyether. From

an entropy standpoint, the coordination would be more

efficient if the ethers are contained in one molecule

rather than in several molecules of for example DME or

dig lyme.

Basically the same red shifts were achieved at

high and low ether concentrations with tetraglyme and

the higher molecular weight polyethers. Therefore, there

is not much discernible difference between the completing

ability of these ethers toward tight ion pairs. This

indicates that tetraglyme and the polymeric ethers are

able to formn equivalent complexes with approximately the

same geometry and coordination number. Such a complex

would be less energetically feasible with DM1E or diglyme

due to unfavorable entropy requirements. These results

moreover indicated that cation coordination is not strongly

dependent on whether the ether oxygens are along the poly-

mer backbone or in a pendant ligand.

A slight blue shift was found with the polyethers

of molecular weight above 1000 compared to the largest









Amx values which were obtained with tetraglyme, PEO 600,
and PEO 1000. This blue shift could be attributed to stack-

ing of the picrate anions along the polymer chain as they

remain in the vicinity of the completed cations if the ab-

sorption decreased with ligand addition.69 However, the

absorption remained constant. This blue shift, therefore,

could not be readily explained.

Viscometry

It is well known that the viscosity of polymer

solutions is related to polymer dimension with an increase

in chain dimensions resulting in an increase in viscosity.

Lundberg, Bailey, and Callard interpreted the increase in

viscosity of poly(ethylene oxide) in methanol on addition

of alkali halide salts as due to chain expansion caused by

binding of the anion to the polymier.3 Since there is sub-

stantial evidence for binding of alkali cations to glymes,

PEO, and crown ethers, it was decided to investigate the

system further.2,,,02,55 Using a system of

2.5 X 10-2 M in monomer of PEO having a molecular weight

of 6 X 106 g/mole with addition of 1.0 X 104 M to

1.0 X 10 H KBr in methanol, the results of Lundberg et

al. could be reproduced, particularly the value of one salt

molecule for every ten monomer units at maximum viscosity.

This number, however, is calculated by assuming complete

binding of all cations at maximum viscosity. The addition

of tetramethylammnonium bromide to PEO in methanol gave

different results. Although the tetramnethylammonium cation





should not bind to the polymer,53 it was expected that in-

creased viscosity would still be measured due to anion

binding if Lundberg et al. were correct in their data

interpretation. The negligible change in viscosity upon

addition of tetramethylammonium bromide to PEO indicated

that the cation is the ion bound to poly(ethylene oxide).

There are several processes which are expected to

determiine the dimensions of a poly(ethylene oxide) coil

in the presence of alkali cation salts. These processes

are shown in Figure 19. As alkali cations are bound to

poly(ethylene oxide), positive charge density builds along

the polymer chain. As this charge density increases,

mutual charge-charge repulsions cause the polymer coil to

expand. This expansion, if all other factors are negli-

gible, would continue until the charge density is high

enough to prevent more cations from entering the polymer

domain. At the same time, however, anions are also being

introduced into the vicinity of the polymer. At low salt

concentration, they have little effect but as the salt

concentration increases, the presence of these anions

begins shielding the charge-charge repulsions of the

cations along the chain. A third factor is the increased

shielding at larger positive charge density due to the

requirements of electroneutrality in the polymer dontain.36

As more cations are complexed by the polymer, the large

charge density inside the polymer domain causes anions

to move into the polymer domain reducing the high positive





O








Eo

go

OE
+
-C
0






O

0 0e



a

g a)



o





0 a > 1


0 0 a 0
I N -- O





CD


c;E=
mR$
cc
042T
40



































































___


charge. The net result of these effects is an expansion

of the polymer upon initial salt addition which reaches

a maximum and then begins to decrease as shielding becomes

very effective at high salt concentration.3 This is

demonstrated in Figure 4 with the addition of KBr to PEO

of molecular weight 100,000 in methanol.

The effects of shielding become more evident upon

comparison of the PEO/KBr system to the PEO/KBr system

containing dicyclohexy1-18-crown-6. The 1:1 addition of

crown ether to KBr in the PEO solution in methanol produced

a viscosity reading equal to that of PEO in methanol'with-

out any added salt. This indicated that crown ether was

able to effectively remove all potassium ions from the

polymer domain due to its stronger binding ability.

Therefore, the addition of crown ether to a solution of

PEO and K~r in methanol allowed a viscosity study in which

free K` was varied while the anion concentration remained

constant (Figure 4). The crown ether system produced a

curve in which the maximiumn was decreased and displaced

to higher concentration of free potassium ions. The occur-

rence of a maximum indicates that shielding becomes more

prominent as positive charge density increases along the

chain at constant anion concentration. This is expected

from the behavior of a typical polyelectrolyte system.36

The decreased miaxim~um indicates that the repulsions are

reduced compared to the system without crown ether so that





































































~


the anion concentration appears to have a moderate effect

on polycation shielding. The increase in salt concentra-

tion at maximum viscosity in the presence of crown ether

may be due in part to more cations needed to produce a

particular expansion at the high anion concentration, but

this in inself should increase the number of anions

attracted into the polymer domain. This particular effect

is not well understood.


Potentiometry


A system similar to the one above was studied by

ion-selective electrode potentiometry (Table 3). The

basic differences here are a lower molecular weight PEO

(100,000) and use of the chloride salt instead of bromide.

The concentration of polymer and salt are in the same

range here as with the viscosity studies but this system

differs from the viscosity system because here polymer is

added to the salt solution. The use of chloride anion

was necessary because the Ag/AgC1 reference electrode was

susceptible to contamination by bromide ions. Upon addi-

tion of PEO from 1.14 X 10-2 M to 4.14 X 10- M in monomer

to 1.0 X 103 M KC1 in methanoT, the number of monomer

units per cation bound was found to increase from 29 to

43 as the amount of polymer increased. These values did

not agree with that of ten monomer units per cation found

in viscosity studies. Since both polymers are of high










molecular weight, it may be assumed that they would com-

plex approximately the .same number of cations for a particu-

lar number of monomer units. However, in viscosity studies

of Lundberg et al. it was assumed that all added cations

were bound to the polymer.37 The potentiometry results

indicated that this was not the case.


Conductance


The comiplexation processes involving the binding

of cations to polymer chains miay be described by the follow-

ing equilibria:


+ K1 +
M + P PM~
S1

+l + K2 2+
M PM PM
1 2


M + P i-1 K1 i 1
i-1 + 1



Here PM; represents the polymer containing only one

cation and PM! represents the attachment of the it

cation to this polymer chain. There may be large num-

bers of cations attached to a single chain in the high

molecular weight polymers, but charge repulsions along

the relatively small chains of the low molecular weight

polymers reduce the chances of large numbers of cations

binding to these systems.





The decrease of conductance of a salt solution upon

addition of a polymer indicates coordination of one or

more cations to the polymer resulting in a lowering of the

ion mobilities (see Figure 11). This continues until all

cations have been completed. Addition of miore polymer at

this point should cause redistribution of cations from

polymers containing more than one cation to polymers with

no cations. This would be due to the higher energy of

polymers containing more than one cation resulting from

charge-charge repulsions along the polymer chain. At

very high polymer concentration this redistribution due
to PM.i + i P .i PM1 1s virtually complete and the con-

ductance is not expected to change upon further addition

of polymer. The measured conductance is now the sum of

the limiting conductance of a polymer chain carrying one

cation Xo,1 and the limiting conductance of the anion,


Binding to PEO 1000. The simplest example of the

conductance decrease is given with PEO 1000 which had a

1/R vs. [polymler] conductance curve similar to that for

18-crown-6 in the complexation of N~aBPh4 in acetonitrile
(Figures 8 and 11). In both cases, the initial section

of the curve describing cation binding was nearly linear;

however, the initial slope was much larger with the crown

ether. The transition from this linear segment to a hori-

zontal segment indicating one cation per ligand occurred





much higher in monomer concentration of PEO than with crown

ether but was very rapid with both PEO 1000 and 18-crown-6.

The lack of a promiinent redistribution region with PEO

1000 indicated that the polymer probably completed a very

small number of cations per chain even at low polymer

concentration, which was not unexpected due to the small

size of the chain.

The binding constant K1 for the complexation of one

sodium cation to such a polymer miay be calculated as fol-

lows:



P + M PMl



The equivalent conductance for a system, X, i-s

defined as


1000 K
cR



where K = the cell constant, c = the salt concentration

and R = the solution resistance.7




R 1000 K


Let R be the resistance of the solution without

added polymer, then,










1 c~f*A =c-f (h + h )
R o oo


If a = the fraction of cations bound, then at a

resistance R, the contribution to conductance of the

cations completed by the polymer = c~f~a*/ and the
o,1
contribution of free ions = c~f-(1 a)X+

Therefore,


=c~f [(1 a)h + aA
R ~o o,1


^o



= aA


1 +
R = o~aq


+
- 10,1)


[PM ] + [M ]



K r- 2 c
1 + 0[1 alc;1



Therefore,


so that


-, =h O(1/R) = KIE 1





where




salt limit





salt salt + polymer


so,


S1 1
nETT7 T '[T t 1



Thus a plot of O/6(1/R) vs. 1/[P] is expected

to give a line w~ith slope = 1/K. The plot for PEO 1000

over a free polymer concentration range of 8.81 X 10-

4.84 X 10-5 Mi is given in Figure 20. The line was found

to have a slope of 1.33 X 104 and a value of Kq

7.52 X 10-

Bindng o PO 600.Poly(ethylene oxide) of

molecular weight 6000 had a 1/R vs. [polymer] conductance

plot similar in many respects to that of PEO 1000 (Figure

11). The initial slopes of both curves are very close

with PEO 1000 having a slightly larger slope. Also, the

concentration in monomer units at which PEO 1000 levels

off at limiting conductance is just below the concentra-

tion where PEO 6000 begins to approach its limiting









































O
N







O




CL

o








<1




a

0


0 00 O


ed<3





conductance (Table 9). Unlike PEO 1000, as more PEO 6000

is added in the region following complete complexation,

there is a gradual decrease in conductance. This indi-

cates that all the cations are indeed bound to polymer

chains but some cations are being redistributed (see

below).

The complexation constant for the binding of one

cation to a polymer miay be calculated as with PEO 1000.

The n/a(1/R) vs. 1/[P] plot is shown in Figure 21. The

curvature at low polymer concentration is due to more

than one cation being completed per polymer chain. The

results of this plot are shown in Table 11 where they are

compared to the values for PEO 1000.



Table 11. K1 for NaBPh4 + PEO 1000 and 6000 in Acetoni-
trile



Polymer Slope K1


PEO 1000 1.33 X 10-4 7.52 X 103

PEO 6000 1.35 X 10-5 7.41 X 104




The increased binding constant for the larger

molecular weight polymer is expected for statistical

reasons. This may be explained from the following.





--1





Let K00 = K1 for PEO 1000 and K60 = K1 for

PEO 6000.



K100 -aG100/RT

K6000 -A6000/R



K AG AG
1000 1000 6000
lnK RT RT
6000


-AH -aH as -as
1000 6000 1000 6000
RT R



Since the cation in each polymer is being com-

plexed to a site which is equivalent physically to that

in the other polymer,



H1000 = H6000


S.m k ln W, where W is the possible number of ways to

complex the cation.

Since



as -as
1000 6000
= n W In W
R 1000 6000


1000
= n
6000





therefore,



K00 100
K6000 6000




The total number of possible ways to coordinate the

cation with the polymer is given by W = q![pl(q p)!]

where q represents the number of oxygen atoms per polymer

molecule and p the average number of oxygen atoms coordi-

nated to the cation. However, not all combinations of

oxygen atoms are equally probable. When ethers in the chain

of a polymer are separated by more than two carbon atoms,

the solvating power of the polymer decreases sharply.

Therefore, only those combinations involving consecutive

coordination sites are important. The number of these

is W = q p + 1.7

For sodium, the coordination number p = 5. PEO

1000 has 22 oxygen atoms with W10= 18. PEO 6000 has

136 oxygen atoms giving W60= 132. Therefore, the cal-

culated K60= 7.3 K100 From experimentally determined

binding constants, it was found that K60= 9.85 K100

The small decrease in conductance noted at high

polymer concentration on the 1/R vs. polymere] conduct-

ance plot for PEO 6000 has been described as resulting

from cation distribution from a polymer containing more

than one cation to a polymer containing no cations.





This process may be described as follows:


2 PM1


PM + P


[PM'l ][P]
K,*=
[P 1 2


Assuming the absence of unbound cations, we can



+ 2+
the salt concentration = PM + 2 PM.
1 2

The fraction of salt completed as the dication,



2[PM

[PH1] + 2 [PM2i


write


so that


[PM2+] = *


Niow,


1 +2+
R -C (1 )A ,1 no,2


2+
where h0,2 is the limiiting conductance of a polymer chain
2+
carrying two cations. 10, is dependent on the mobility





2+
of polymeric Pli2 If it is assumed that there is no

change in polymer size upon adding a second cation, the

limiting conductance of a polymer containing two cations,

o2+ should be 2Ao,1

Thus,



1 +
f~c(1 + B)A
R 0,l


Since





2 22

2K*c2 21-0 = B cP



Now if B << 1,


2K*c




Substituting for B and differentiating 1/R with

respect to 1/P, we obtain



~(/=] f.C ho1)= 2fc2 o,1K*



A plot of 1/R vs. 1/[P] over the polymer concen-

tration range of 3.3 X 102 M to 1.0 X 10- 01 in monomer





PEO 6000 where no free ions are present gave a straight
2 +
line with a slope = 2fc K*Ao, (Figure 22). K* was then

calculated and since K* = K2/K1, K2 was also calculated.

It should be noted that the plot was not linear for poly-

mer concentrations below the range listed above. This

deviation is probably due to complexation of more than

two cations per polymer causing 1/R to be lower than ex-

pected. The results for PEO 6000 are given in Table 12.



Table 12. Binding Constants for Complexation of Na+
to PEO 6000 in CH3CN




Slope of
Polymer K1 1/R vs. 1/[P] K* K2


PEO 6000 7.41 X 104 8.18 X 104 0.235 174X0-






Binding to high molecular weight polymers. The

higher molecular weight polymers, PEO 100,000 and PEO

900,000, gave different 1/Rl vs. [PEO] conductance plots

upon addition to salt solutions than the lower molecular

weight polymers (Figures 11 and 12). PEO 100,000 had an

initial slope slightly larger than PEO 900,000 but quite

surprisingly both of these slopes were significantly

smaller than the slopes of PEO 1000 and 6000. Also, in





? c
O o

Op I


4E u


o
OD
O
OO

O


r-


O 0-
N





O ~~ ~ O
= e oa


SO
O





contrast to PEO 1000, the high molecular weight polymers

did not approach a horizontal line at high polymer con-

centration, but instead both showed a continual decrease.

These decreases at high polymer concentration with the

high molecular weight polymers were much larger than with

PEO 6000. Furthermore, the conductance of the PEO 100,000

and 900,000 solutions at high polymers concentration was

much lower than the conductance of the PEO 1000 and 6000

systems.

At low polymer concentration, the slopes decreased

with increasing molecular weight. This can be explained

by considering two factors. First, viscosity measurements

indicate that at these salt concentrations (a 10-4 )

the size of the polymers is not greatly changed on cation

binding so that the conductance per ion increases with

the molecular weight assumiing a constant number of ethylene

oxide units per cation site. Second, the mobility of

polymers in solution decreases with molecular weight but

this decrease is less than linear. For instance, for a

random walk polymer the diffusion coefficient varies as

M-127 Therefore the equivalent conductance of polymer

bound cations is expected to increase with molecular

weight. Thiis leads to lower slopes at low polymer concen-

tration for higher molecular weight polymers.

At high polymer concentration, the miagnitudes of

the slopes were found to increase with increasing molecular





weight. After all cations had been comiplexed, the lower

molecular weight polymers, due to their small size, con-

tained relatively few cations per polymer. As more poly-

mer was added there were relatively few cations transferred

to the uncomplexed polymer chains due to the low concen-
2+
tration of PM2 in the following equilibrium:


2+ +
PM2 + P + 2 PM1



The low number of cations transferred caused little change

in transport of charges or mobility of polymer bound

cations. With the high molecular weight polymers, assum-

ing constant binding of cations per unit weight for all

polymers, there were many cations per polymer chain. As

more high molecular weight polymer is added, there were

many cations redistributed from polymiers containing many

cations to polymers containing fewer numbers of cations.

This led to a greater decrease in conductance than with

the lower molecular weight polymers.

There were several factors which caused diffi-

culty with the calculation of binding constants with

the higher molecular weight polymers. Since the polymers

are so large and so many cations can be bound per polymer

chain at low polymer concentration, the binding and redis-

tribution processes occur simultaneously over most of the

polymer concentration range. This leads to many equilibria





being established which cannot be properly evaluated. The

binding processes themselves, as can be shown from earlier

calculations, would lead to an expression for conductance

which can be described as follows:



1~ + 2 + i +
$ = f-c[(1 B)Ao 81 ol,1l 2 o,2E i o,~i "


where 6. is the fraction of salt completed to the polymers

containing i cations. It is clearly impractical to evalu-

ate the binding constant Ki using this equation.
In addition, complications arise from probable

site binding of counterions to the polycation.7 Such

effects were evident from the work of Lundberg, Bailey,

and Callard, who showed specific anion effects on the vis-

cosity of cation bound poly(ethylene oxide).3 Similar

anion effects have been observed in light scattering ex-

periments on these systems.58 Thus a quantitative or

even semiquantitative description of these high molecular

weight polymers is difficult on the basis of the above

conductance methods.

Mobility of PM1 complexes. For the low molecu-

lar weight PEO 1000, it is possible to calculate ho1

the limiting conductance of the polymer containing one

cation. As was apparent from a comparison between the

18-crown-6 and PEO 1000 1/R vs. [PEO] conductance curves










at high polymer concentration, it

only one sodium ion per polymer.


is reasonable to assume

It has been shown that


1 +
c~f-A = c~f (X + h )
R o oo


where Xo,Na+ = 77.3 and ho,BPh-=
4
At the limiting conductance,


58.1 in acetonitrile.6


Therefore,


1 +
R c jo


+
A )
0,1


so that


(1/Ro) (1/R)
o,1 c~f ao


For PEO 1000, X = 36.6 .
o,1
For PEO 6000, it was shown for the K, calculation
that a plot of 1/R vs. 1/[P] gave a straight line (Figure

22). The intercept of this plot at [PEO] 0 gives the

limiting conductance ho,1 o h which equals 69.6. Sub-

tracting the value of Xo for RPh4 gives a value of 11.5

for Xoq for PEO 6000.









Since the X values are a function of the
o,1
mobility of polymer bound cations, if a randomly coiled

polymer is assumed they should be related by M-/ where
71 +
M is the molecular weight. The ratio of X0,1 for PEO
1000 and 6000 is 3.18 compared to the ratio of their

molecular weights of 2.45. Although this method does not

give the exact relationship between ho,1 values for vari-
ous molecular weight polymers, it can nevertheless be

used to approximate 10, for the hi-gh molecular weight
o~+
polymers. For PEO 100,000 the hoq value is found to be
approximately 3,2 and for PEO 900,000 it is approximately

1.2. Therefore, at high polymer concentration of high

molecular weight polymers 1/R approaches ho'


Discussion of
Poly(3,6,9,12,15-pentaoxa-1-heptadecene_1


From inspection of Table 6, it appears that the

Amax, of sodium picrate in 10% THF/CHC13 in the presence
of P(PHD) was the same as that for equal quantities in

terms of monomer units of poly(ethylene oxide). As stated

earlier for poly(ethylene oxide), the relative shifts

involved led to the conclusions that the polymers are

only peripherally solvating the sodium cation in this

solvent and that ion pair comiplexation constants are about

the same as with PE0.





From the conductance data of P(PHD), Figure 16,

the P(PHD) concentration at which the conductance curve

approaches a "constant level" was approximately 3.0 X 10-2

M in monomer in acetonitrile which was very close to the

value of 2.6 X 10- M in monomer for PEO 1000. This is

further evidence for the similarity of the binding con-

stants of P(PHO) and PEO. The 1/R conductance value

where the curve approaches a "horizontal" line, however,

was much higher with P(PHD) than with any of the PEO poly-

mers. For example, the P(PHD) plot began to level off at

a value of 90.79 pU, whereas PEO 100,000 was still de-

creasing slightly at 60.6 pU. Since the sodium cation may

be effectively comnplexed in a 1:1 stoichiometry by a glyme

containing five oxygens, it may be coordinated to a single

pendant ligand on P(PHD).2 In order for a completed

sodium ion to move along this polymer chain, it would

only be necessary to shift from one pendant glyme to an

adjacent one without involving a major conformational

change in the polymer. With PEO, however, a conforma-

tional change in the polymer chain would be required to

move a sodium ion from a binding site composed of five

oxygen atomns in the polymer backbone to another site

farther along the polymer containing another five oxygen

atoms in the proper conformation. From entropy considera-

tions, transport of a sodium ion along the chain would be

easier with P(PHD) which would lead to a higher limiting

conductance.




Full Text

PAGE 1

ALKALI CATION BINDING TO NONIONIC POLYMERS By RONALD E. CAMBRON A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1976

PAGE 2

Copyright 197 6 by Ronald E. Cambron

PAGE 3

ACKNOWLEDGEMENTS The author wishes to express thanks to his committee, Dr. Thieo Hogen Esch, chairman, Dr. George Butler, Dr. Edmond Gabbay , Dr. Roger Bates, and Dr. Dinesh Shah, for their assistance and support in the completion of this research project. In addition, a debt of gratitude is owed to the secretary, postdoctoral associates, and fellow graduate students of the fourth floor of SSRB for their understanding and moral support. Other people who deserve thanks are the library staff and the personnel of the glass shop, electronics shop, and machine shop for their invaluable assistance with the technical aspects of this project. Special thanks are extended to the author's wife, Nelda, for her encouragement and support during this period of graduate study.

PAGE 4

TABLE OF CONTENTS ACKNOWLEDGEMENTS "Mi LIST OF TABLES vi LIST OF FIGURES viii ABSTRACT x CHAPTER I INTRODUCTION 1 Monomer i c Complexing Agents 1 Polymeric Complexing Agents 10 Objectives 12 II CATION BINDING 18 Results 20 Vi scorn e try 20 Extraction Equilibria 23 Potent iome try 27 UV-visible Spectrometry 30 Conductance 39 Discussion of Glymes and Pol y ( ethyl ene oxides) 61 UV-visible Spectroscopy 61 Viscometry 63 Potent i ometry 68 Conductance 69 Discussion of Poly(3,6,9,12,15-pentaoxa-lheptadecene) 37 Solvent Effects 89 Cation Effects 90 Discussion of Poly(acryloyltyrocidine) 90 i v

PAGE 5

Pa ge III PHASE TRANSFER REACTIONS 91 Homogeneous Reactions 94 Results 94 Discussion 98 Heterogeneous Reactions 99 Results 99 Discussion 105 IV EXPERIMENTAL PROCEDURES 108 Preparation and Purification of Materials 108 Solvents 103 Glymes, PEO, and PVP 108 Vinyl Glyme 109 Crown Ethers 109 Antibiotic Compounds 110 Salts 114 Cation Binding Measurements 116 Viscometry 116 Distribution Equilibria 117 Potent io in etry 118 UV-visible Spectrometry 119 Conductance 120 Reactions 121 Reaction Methods 121 Product Analysis 122 REFERENCES 124 BIOGRAPHICAL SKETCH 129

PAGE 6

LIST OF TABLES Table Page 1 Extraction of Alkali Picrates from Aqueous Layer into Chloroform Layer Containing Poly(acryloyltyrocidine) 2 Logarithms of Stability Constants of Crown Ethers Added to NaCl in Methanol 3 Binding of 1.0 X 10 _3 M KC1 by PEO (molecular weight 100,000) in Methanol 4 Absorption Maxima of Picrate Salts in Chloroform in the Presence of Excess Crown Ethers 5 Binding Constant of Tetraglyme Addition to 1.0 X 10~ 4 M Sodium Picrate in 10?b THF/90/' CHCI3 from BenesiHildebrande Plots 6 Absorption Maxima After Addition of Ethers of Various Molecular Weights to 1.0 X 10-4 M Sodium Picrate in 10% THF/90% CHClg 7 Decrease in Conductance Due to Polymer Viscosity 8 Conductance of Polymers in CHgCN 9 Conductance Studies of Polyethers with 1.0 X 10" 4 M Tetraphenylboride Sal ts 10 Solvent Effects on Conductance of P(PHD) Added to 1.0 X 10" 4 M NaBPh 4 Solutions 11 K] for NaBPh 4 ' PEO 1000 and 6000 in Acetonitrile 26 27 2 9 31 3 J 38 45 46 54 5 5 75 v 1

PAGE 7

Table Page 12 13 14 15 16 17 Binding Constants for Complexation of Na to PEO 6000 in CH 3 CN _? 3.7X10 M Butyl Bromide +9.6 M Sodium Acetate in Acetonitrile H 2 Added X 10 with -2 -2 -2 3.7 X 10" M Butyl Bromide + 9.6 X 10 M Sodium Acetate Catalyzed by 0.18 M Polyethers in Monomer in 5?, H o 0/95% CH 3 CN l Polyether Catalyzed Reactions of 3.7 X 10-2 M Butyl Bromide + 9.6 X 10 M Sodium Acetate in Acetonitrile Reaction of 3.7 X 10" 2 M Butyl Bromide with 9.6 X 10-2 M Sodium Acetate in CH.jCN Catalyzed by Various Polymers Catalysis of 3.7 X 10~ 2 M Butyl Bromide + 9.6 X 10"2 m Sodium and Potassium Aceta te i n Acetoni tr i 1 e 81 95 97 101 102 104 v i i

PAGE 8

LIST OF FIGURES Fi gure Page 1 2 3 4 5 6 7 8 9 10 1 1 12 13 Typical Crown Ethers Cation Complexing Agents The Antibiotic Tyrocidine Viscosity of PEO with KBr Added Absorption Maxima Upon Addition of Tetraglyme to Sodium Picrate Benesi-Hildebrande Plot for Initial Tetraglyme Additions to NaBPh^ Benesi-Hildebrande Plot for Large Tetraglyme Additions to NaBPh^ 4 Conductance of 1.0 X 10 + 18-crown-6 in CH 3 CN Conductance of 1.0 X 10' 18-crown-6 in Acetone M NaBPh NaBPh 4 + ,-4 Conductance of 1.0 X 10 ~" M Tetraphenylboride Salts + Low Amounts P(PHD; in CH-CN at 25° Conductance of 1.0 X 10 -4 M NaBPh Low Amounts P(PHD) in THF at 25° Conductance of 1.0 X 10 -4 M NaBPh 4 + Low Amounts P(PHD) in Acetone at 25° Conductance of 1.0 X 10 -4 M Tetraphenylboride Salts + Large Amounts P(PHD) in CH CN at 25° 4 16 17 21 33 3b 36 40 41 48 4 'J 50 bl VI 1 i

PAGE 9

Figure 14 15 16 17 18 19 20 21 22 23 24 25 25 -4 M NaBPh Conductance of 1.0 X 10 P(PHD) in THF at 25° Conductance of 1.0 X 10 P(PHD) in Acetone at 25° Conductance of 1.0 X 10 After Initial PEO Additions in CH^CN at 25° J M NaBPh, + M NaBPh Conductance of 1.0 X 10 at 25° -4 PEO in CH 3 CN Conductance of 1.0 X 10 M NaBPh. + PEO 900,000 in CH CN at 25° 4 Effects of Alkali Metal Salts on PEO Vis cos i ty A/A(l/R) vs. 1/[P] for PEO 1000 K, Calculation A/A(l/R) vs. 1/[P] for PEO 6000 K, Calculation 1/R vs. 1/[P] for PEO 6000 K 2 Calculation Butyl Bromide + Sodium Acetate in 5% H 2 / 9 5 % CH 3 CN After 5 Hours Butyl Bromide + Sodium. Acetate in CH 3 CN After 3 Days Infrared Spectrum of Poly(acryloylbenzotriazole) Absorption of (A) Tyroc i di ne • HC1 ; (B) Poly (acryl oyl tyroci di ne ) in Me than ol Page 52 53 59 6 65 6b 74 76 82 96 103 113 115 l x

PAGE 10

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ALKALI CATION BINDING TO NONIONIC POLYMERS By Rona Id E . Cambron December, 1976 Chairman: Thieo E. Hogen Esch Major Department: Chemistry The alkali cation binding of nonelectrolyte polyether type polymers in several nonaqueous solvents was measured by viscometric, spec tropho to.me tr i c , potentiometric and conductance techniques. An attempt was made, to correlate these findings to the ability of those polymers to catalyze nucleophilic substitution reactions to acetate salts with n-butyl bromide. UV-visible studies indicated that the complexation of sodium salts to polyethers in chloroform increased with increasing chain length up to tetraglyme, which contains five oxygen atoms. Further increase in chain length had no apparent effect on the complexing power of the polyethers. Viscometric and potent io metric studies indicated that in these systems many cations were bound to a single high molecular weight chain. These studies also

PAGE 11

showed that cation binding initially led to chain expansion due to chain bound cation-cation repulsion. Addition of more salt, however, gave a reduction in viscosity as more anions within the polymer domain shielded the positive charges. From conductance results it was shown that the number of cations bound per chain increased with increasing molecular weight. Further addition of polymer led to a redistribution of cations from polymer chains containing more than one cation to chains containing no cations. Binding of cations was strongest in THF, followed by acetoni tri 1 e , and acetone. Binding constants were calculated for the complexation of NaBPh to PEO 1000 and PEO 6000 in aceton i tr i 1 e . Distribution equilibria were used to study the interaction between alkali cations and poly (aery 1 oyltyroci di ne ) , a polymer containing a cyclic antibiotic. It was found that the ion selectivity of the polymer was similar to that of tyrocidine itself. The effect of po ly ( ethyl ene oxide) and similar polymers in the pha setransfer reaction of sodium and potassium acetate with n-butyl bromide in acetonitrile was investigated as well as the corresponding homogeneous reaction in 5% H 2 / 9 5 ?i CH 3 CN (v/v) medium. In both cases a correlation was found between the complexing abilities and catalytic efficiencies of the lower molecular weight polyethers, although the effects were more pronounced x 1

PAGE 12

in the phasetransfer reactions. Polyethers of molecular weight 600 and 1000 were found to be the most effective catalysts, the reaction yields decreasing as the molecular weight increased to 6000 and beyond. The smaller differences with the various polymers in homogeneous media were probably due to the effect of water on anion nucleoph i 1 i c i t i es and cation coordination. In heterogeneous media the differences may have been due to a higher efficiency in transport of salt from the crystal into the solution phase, a less efficient complexation of salt by higher molecular weight polymers, or a combination of these effects. X 1 1

PAGE 13

CHAPTER I INTRODUCTION The complexation of alkali cations by neutral or non ionic molecules is a phenomenon which has been extensively studied. Depending on such factors as cation size and charge, the size, shape and distance of the negative counter ion, the ability of the solvent to solvate either or both of the ions, the nature of the binding site, and the temperature, the degree of binding may vary over a wide range. M o n o in e r i c C o m p 1 e x i n g Agents A series of polyglycol dimethyl ethers or glymes of the general formula CH 0(CH ? CH„0) CH was investigated with respect to their behavior in solvating alkali 1 2 cations.' It was found for compounds ranging from X = 1 to X = 6 that coordination complexes were formed with lithium, sodium, and potassium salts of fluorenyl carbanions in low dielectric constant solvents. The complexation of these salts with glyme led to contact ion pairs, g 1 ymesepa ra ted ion pairs, or to a mixture of both depending on the size of the cation and the chain 1

PAGE 14

length of the glyme. By adding different quantities of glyme to f 1 uorenyl 1 i thi um in dioxane or f 1 uoreny 1 sod i urn in tetrahydrofuran (THF), it was possible to measure the equilibrium constants of the reaction F"M + + nG t F~,M + ,G F'.M ,G n + mG t F",G , ,M K n n+m 2 where n and m are the number of glyme molecules participating in the complexation, F ~ , M ,G is the contact ion pair, and F ,G,M is the solvent separated ion pair. As the number of oxygens in the chain increased, the various complexation constants increased up to an X value of three for lithium and four for sodium where they leveled off. Temperature dependence studies indicated that glymes were more effective than solvents such as tetrahydrofuran in coordinating alkali ions due to a smaller loss in entropy. This small entropy loss is expected because one glyme molecule contains several oxygen atoms all of which may coordinate with a cation of appropriate size, whereas several THF molecules would be needed to occupy the same coordination sites. It has also been shown that crystalline 1:1 complexes of NaBPh. were obtained with glymes having X = 4, 6, and 7 as well as a 2:1 complex with the compound having X = 3.

PAGE 15

The effect of the small entropy loss on binding observed in the polyglycol dimethyl ethers was even more pronounced in the macrocyclic polyethers, or crown ethers, first described by Pedersen in 1967. These compounds are cyclic in nature composed of repeatinq ( -0-CH„-CH„) 2 2 'n units (see Figure 1). In these molecules the coordinating oxygen atoms are positioned for favorable complexation of an ion of suitable size. On cation binding, little entropy loss is experienced. Containing a central hydrophillic cavity with a hydrophobic exterior, these crown ethers were shown to solubilize certain alkali cation salts in aromatic or chlorinated hydrocarbons. These complexes as well as those of the polyglycol dimethyl ethers were believed to be due to ion-dipole interactions 4 5 between the cation and ether oxygens. ' Frensdorff further studied the stability constants of these complexes by potent iome try using cation-selective electrodes. Due to water's stronger solvation of the cation, he found the binding constants in methanol to be significantly higher than in water. Also, as the ring size varied, cation selectivity changed so that an individual crown ether would most efficiently bind the cation which would just fit into the center cavity. The complexation of large cations by polyethers having small cavities was increased in several instances by the formation of a 2:1 crown-cation complex. Replacing the oxygens with

PAGE 16

c u c 1 I c c O-CH— CH 2 [CH — CH5 — 0] — ' L 2 2 J n Dibenzo-18-Crown-6 18-Crown-6 n = 4 — [CH — CH]— 2 1 x 0— CH~CH 2 [CH— CH — 0] n — ' Poly(vinylbenzo-15-Crown-5) n = 3 Figure 1. Typical Crown Ethers

PAGE 17

nitrogen or sulfur weakened the complexing ability toward alkali cations. Furthermore, Frensdorff quantitatively described the ability of crown ethers to extract alkali salts from an aqueous layer into a nonaqueous layer containing the crown ether. Following these initial studies, an extremely large number of multidentate macrocyclic compounds was synthesized. These included cyclic polyethers, alkyl and aromatic substituted cyclic ethers, macrocycles containing nitrogen donor atoms, sulfur donor atoms and mixed donor atoms. In almost e\/ery case, these compounds were shown to complex one or more cations and were usually Q specific for a particular cation. Crown-compl exed salts generated substantial interests due to their ability to exist as separated ion pairs g in solution. Takaki, Hogen Esch and Smid studied the optical spectra of crown ether complexed f 1 uoreny 1 sod i urn and fl uoreny 1 potass i urn in tetrahydrof u ran and tetrohydropyran (THP). The salts were shown to exist as a mixture of contact and crownsepa rated ion pairs, the stability of the complexes being a function of temperature, solvent, size of the cation and structure of the crown compound. The cyclic ligand concept was extended in a series of macroheterobi cycl es called cryptates, designed by 11-14 Lehn. As a result of their three-dimensional cage structures, which allow cations to be included within the

PAGE 18

central molecular cavity, these compounds exhibited binding constants even greater than the crown ethers. Difficult synthetic procedures and relative unavailability have limited their use at this time. Another group of compounds which complex with alkali ions is called the ionophores. These are antibiotics which form lipid soluble complexes with alkali metal salts and then transport these ions across lipid barriers including artificial biological membranes. Having a cyclic peptide structure, the ionophores complex cations by ion-dipole interaction with carbonyl or ether oxygens. The oxygens replace all or some of the solvent molecules in the solvation sphere of the aqueous ions. The complex then assumes a conformation in which the charged cation is held in the center of the ionophore while alkyl groups form a hydrophobic outer surface allowing the entire molecule to pass through the low .dielectric environment of lipid membranes. ' This ability to transport ions through membranes has also been demonstrated by crown ethers. Many antibiotics have been shown to complex cations and transport them through cell mitochondria. Among these are certain deps i pept i des , macrotet ra 1 i des , and polypeptides including valinomycin, the macrotetralide act ins, the enniatins, and the gramicidins; and also monocarboxyl i c polyethers, including nigericin,

PAGE 19

dianemycin, the monesins, X-206 and X-537. 16 ' 18 Various methods have been used to study the binding strength and selectivity of various antibiotics. Extraction of cations from an aqueous layer into a nonaqueous layer containing the antibiotic was used in some of the initial studies. 15,19 Other methods used included measuring the change in conductance across lipid membranes due to antibiotic modified cation transport, ' ion-selective electrodes, circular dichroism, '" and fluorescence spectroscopy. ' 23 ' 24 As a result of these and other studies, a striking resemblance between antibiotics and certain crown ethers was noticed in their selectivity toward cations. It was then realized that crown ethers could serve as model compounds for the large cyclic antibiotics in examining the processes involved during cation transport in the mitochondria. The property of various ligands, particularly crown ethers, which enables them to solubilize salts in nonpolar media and to decrease the cation-anion interaction in ion pairs has resulted in their use as catalysts in synthetic reactions. The process of phasetrans fer catalysis was first investigated by Starks in the displacement reactions of alkyl halides in an organic phase with inorganic ions in an aqueous phase. Although these reactions are initially inhibited because of phase separation, they were catalyzed by small amounts of organic soluble tetraalkyl ammonium or tetraalkyl phosphonium

PAGE 20

salts. This process was believed to involve the complexation of nucleophilic cyanide anions in the aqueous phase by quaternary cations and the subsequent transfer of these ions into the organic phase. Here the cyanide anions reacted with the alkyl substrates displacing ha 1 i de ions. The halide ions were then complexed by the quaternary cations and carried into the aqueous layer. Replacement of the halide ion by another nucleophilic cyanide anion led to further phasetrans fer catalysis. The reaction between sodium cyanide and 1 -bromooc tane may be described as fo 1 1 ows : R'Br + R N + CN" -> R'CN + R.N + Br" (organic phase) NaBr + R N + CN~ t NaCN + R N + Br~ (aqueous phase) The nucleophilic displacement reactions have also been carried out in two phase organic-aqueous systems using various alkyl mesylates and halides as substrates with ? 1 crown ether-complexed alkali halides as nucl eophi 1 es . The limitation of quaternary ammonium or phosphonium catalysts was that they must generally be used only with the transfer of ions from an aqueous phase into non-polar solvents. However, crown ethers have the ability to pro mote a direct solid-liquid pha set ra n s f er of salts into nonpolar solvents. Liotta demonstrated that in benzene

PAGE 21

or acetoni tri 1 e, 18-crown-6 could complex potassium fluoride, acetate, cyanide, or azide which in turn would partially free the anion to act as a strong nucleophile toward alkyl ha 1 ides . 28-31 In a similar study, Zubrick, Dunbar, and Durst examined the reaction of benzyl chloride with potassium cyanide in acetonitrile containing 18crown-6. They found 90-95% reaction and even when 1-2% water was added, they still achieved an 85-95% yield. 32 Other investigations have shown that crown e ther-compl exed KMn0 4 in benzene could oxidize olefins, alcohols, and 33 aldehydes; and aromatic hydrocarbons containing acidic hydrogens (e.g., fluorene) could be alkylated by reaction with aqueous sodium hydroxide in the presence of dibenzo18-crown-6. The effectiveness of crown ethers as reagents for solid-liquid phaset ra nsfer reactions can be attributed to their powerful cation binding ability and the fact that they are flexible molecules with several polar 35 sites. ' Their flexibility allows the crown ethers to interact with the surface of a crystal lattice of a salt in the most energetically favorable geometry. By assuming the same shape as the surface of the crystal lattice, the crown ether can readily transfer a cation from its lattice site to the crown ether cavity. The anion remains as an ion pair with the cation complex.

PAGE 22

PoJ _ymeric Complexing Aq ents The complexes which have been described up to this point involve low molecular weight species where the ligand usually functions in a 1:1 relationship with the alkali cation. With polymeric complexing agents, however, several ions or ion pairs may be bound to one polymer molecule. In addition to the entropy effects mentioned for glymes and crown ethers, other effects are also present in the binding of cations to non ionic polymers. One effect is polymer conformation which determines the proximity of binding sites to one another and, therefore, has a direct effect on the binding process. As ions become bound along the polymer chain, the polymer becomes in effect a pol ye 1 ect ro 1 y te . The cation binding may then be reduced due to charge repulsions by cations already bound to the chain. The repulsions may be decreased by counterions in the vicinity of the polymer shielding the charges. Other factors may involve d i pol e-di pol e interactions in the polymer domain between ion pairs on the chain and ion pairs in solution. These factors may increase or decrease ion binding to the polymer. The effects listed above are quite complicated and are not we 1 1 under s tood . These effects were studied with poly ( ethyl ene oxide), the polymeric analog of the previously mentioned glymes. Viscosity studies by Lundberg, Bailey, and

PAGE 23

11 Ca 1 lard showed that 0.05 M potassium fluoride added to 2% poly(ethyl ene oxide) in methanol resulted in the binding of 1 potassium fluoride molecule per 9 ethylene 37 oxide repeating units. Liu noted cooperative effects in an NMR study of the binding of potassium iodide to pol y( ethyl ene oxide) in methanol. The polymeric type interaction began with the oligomer having seven repeating units. Another approach involved visible spectroscopy of alkali metal fluorenyls complexed to poly ( ethyl ene oxide). Here again, the strongest complex was formed at an ethylene oxide unit to metal ion ratio of approximately 3 9 5:1 for sodium and 7:1 for potassium. Further examples of polymeric cooperative effects were demonstrated by Smid with the syntheses and cation binding studies of poly (vinyl crown ethers). Using optical spectroscopy, distribution equilibria, conductance, viscosity and ion transport through liquid membranes, he was able to compare the cation binding properties of polymers containing pendant crown ether ligands with crown ethers themselves. In most cases, the crown ethers which formed 1:1 crown-cation complexes demonstrated equal to slightly better binding of cations in the polymeric form. However, for those crown ethers capable of forming a 2:1 crown-cation complex the complexation constants for the polymeric crown ethers were considerably enhanced over 40-44 the monomeric species.

PAGE 24

12 The polymeric cooperative effect is apparent not only with cation binding but also with increased reaction rates where polymeric materials serve as catalysts. It has been shown that the Williamson reaction between sodium phenoxide and butyl bromide in dioxane occurred over 100 times faster in the presence of polyvinyl pyrrolidone) as in a solution containing the same concentration by weight of N-methyl pyrrolidone. Here the rate enhancement was attributed to the dissociation of sodium phenoxide based on the tight solvation of sodium cations by pyrrolidone segments in the polymer coil. From entropy considerations, it would be a lower energy process to solvate a sodium cation with pyrrolidone groups attached to a polymer than with free pyrrolidone residues. This should lead to higher rates with polymeric cata lysts . Objecti ves Catalysis of synthetic reactions by polymers is of both practical and theoretical interest. Polymers may be relatively easily retrieved from homogeneous reaction mixtures by various means which would allow their reuse as catalysts. Polymers may also catalyze reactions in heterogeneous fashion in the form of microporous beads or gels which would immobilize the polymer in the solid state. Here again, separation from the reaction mixture

PAGE 25

13 is easily achieved. As mentioned previously, polymers have the ability to complex cations through a variety of effects. Frequently such compl exa t i ons are as strong or stronger than with the corresponding low molecular weight analogs. Polymers, in addition, may exist in conformations which would possibly allow complexing sites to interact with solid phase reagents in a more energetically favorable manner than low molecular weight compounds. In spite of these advantages of polymer catalysts, there have not at this time been many reported investigations of polymers in phas etran sf er reactions. Phasetransfer catalysts may serve a dual role. They may not only transfer a reactive substance, usually a salt, from a solid or aqueous phase into an organic layer, but they may also modify this reactant once it is in the organic phase to increase its reactivity. How this is done, however, has not been clarified in detail. It is believed that by studying cation binding and catalytic properties of a series of phase-transfer catalysts of varying degrees of polymerization, more insight may be obtained in the nature of the phasetransfer process of both low and high molecular weight compounds. It would also be of interest to study the correlation of the cation binding and catalytic properties of various other polymers in order to assess the role of various cation complexing sites. Such studies may be of

PAGE 26

help in clarifying the nature of polymer catalysis in phase-transfer reactions since at least for solid-liquid phase-transfer processes, the nature of the catalysis i s not we 1 1 unders tood . Therefore, the objectives of this research project were as fol 1 ows : 1. Study the ability of various non ionic polymers, having similar binding sites but differing in the location of these sites, to bind alkali cations and to modify the ion pair structure of their salts. This was done by conductance and visible spectroscopy. 2. Study the phasetran s fer ability of these polymers as a function of molecular weight, polymer concentration, and solvent composition. The polymers chosen for the phasetrans fer catalysis work were ethers. Pol y( ethyl ene oxide) which contains ether oxygens in its backbone was readily available in various molecular weights. Various glymes, low molecular weight analogs of po 1 y ( e t hy 1 ene oxide), were also commercially available. The vinyl glyme 3 , 6 , 9 , 1 2 , 1 5pen taoxa 1-heptadecene containing a pendant group with five ether oxygens was commercially available, easily polymerized, and the polymer was soluble in most common solvents.

PAGE 27

15 This material in polymer form was investigated and compared to the previously listed ethers. Examples of the polyethers are given in Figure 2. Most common cyclic antibiotics which were known to bind cations had no suitable functional group to aid in attachment to a polymer or the group present was involved in cyclization of the antibiotic during the complexation process. 15 ' 16 ' 18 ' 19 ' 21 ' 46 ' 47 ' 48 The cyclic antibiotic ty roc i dine contained the ornithine amino acid function which included a primary amine group, facilitating attachment to a polymer chain. Also, tyrocidine bound alkali cations, particularly sodium, and it was also com49 mercially available. Therefore, the antibiotic tyrocidine met the selected criteria, was attached to a polymer, and was used in the study of alkali cation binding to various polymers. The antibiotic structure is given in Fi gu re 3 .

PAGE 28

16 CH 3 — Of CH 2 ~ CH 2 — JCH Tetraglyme -Ech 2 — ch 2 — oy Poly(ethylene oxide^ -{CH„— CH}2 i * I CH,, I 2 CH„ CH„ I 2 CH 3 Poly(3,6,9,12,15-pentaoxa-l-heptadecene; -[CH„— CH}* I n C=0 I Tyrocidine Poly(acryloyltyrocidine) Figure 2. Cation Complexing Agents

PAGE 29

17 Tyro ci dine A i 2 CH H CHr-C-CH, J l 3 CH„ CHr-C-CH^ I c. 3 | 3 CH„ i iCH CH„ «^ CH„ CH„ CH I 2 \2 i 2 CH r— N— C — C — N — C— C— N — C — C — N — C~ C~ N — C— O I I II I I II I I II I I I! I II HH0HH0HH0HH0 HO 0HH0HHOHHOHH0HH II I I II I I II I I II I I II I I — C— C — N — C — C — N— C — C — N — C — C — N — C — C— N iiii CH 9 CH„ CH„ CH„ CH CH„ C=0 I 2 i C=0 NH, a OH Ni-l o Tyrocidine B X = -«"r® Y = — CH Tyrocidine C X = — CH, Y = — CH, Figure 3. The Antibiotic Tyrocidine

PAGE 30

CHAPTER II CATION BINDING Many different techniques have been used to study the complexation of cations to nonionic ligands. The vibrations of alkali cations encaged by crown ethers in solution have been examined by far infrared spectros50 copy. In this region, solventand an ion-independent bands were used to determine the effects on the cation motion frequencies in solution caused by ion binding. 23 Proton and Na magnetic resonance have been used to study the chemical shifts induced when a cation is coor51 52 dinated to a cyclic ether. ' Also, calorimetry has proved useful in determining binding constants as well as 53 54 enthalpy and entropy values for cation complexation. ' Another valuable tool for calculating binding constants was potent i ometry with ca t i on -se 1 ect i ve electrodes. This measured the number of free ions remaining in solution after a ligand had been added. The use of alkali cation salts of fluorescent probes or even the fluorescent thallium cation itself made possible the use of fluorescence spectroscopy in the study of cation binding 23,24 to various linands, particularly cyclic antibiotics. In polymer systems, light scattering and viscosity have

PAGE 31

19 been used to examine the polyelectrolyte effect of chain expansion upon binding of cations along the polymer chain. ' Distribution equilibria have been useful in demonstrating the ability of monomeric or polymeric complexing agents to extract alkali cation salts from an aqueous phase into an organic layer where the salt would 4 7 18 ? 1 4 ? normally be insoluble. ' ' ' ' An alternative to studying parameters directly concerned with the cation was to study the UV-visible spectra of salts with anions such as picrate or fluorenyl. These salts were sensitive to changes in the interionic ion pair distances caused by complexation of the cation, making them valuable probes 5 5 56 of the binding strength of various ligands. ' Another method which has been used to determine binding constants as well as the s to i ch iomet ry of the cation complex is 43 57 conductance. In solvents where the alkali cation salts existed as free ions, the conductance of the solution was shown to decrease upon addition of ligands such as crown ethers as long as the mobility of the complexed ion was lower than that of the noncomplexed ion. This decrease continued until nearly all the cations were complexed. At this point the conductance began to level off with the breaking point yielding the s to i ch i ometry of the complex. In the present study, the following techniques were used: viscosity, distribution equilibria on one polymer which had the appropriate solubilities,

PAGE 32

20 cation-selective electrodes, UV-visible spectroscopy, and conductance . Results V i s come try Lundberg, Bailey, and Callard examined the effects on viscosity of adding alkali metal halides to poly(ethy37 lene oxide) solutions in methanol. They observed high viscosity with moderately high salt concentration (0.02 M potassium iodide) with a decrease in viscosity upon further addition of salt. From their results, they calculated that one molecule of salt associated with approximately nine ethylene oxide units. The anion was tentatively postulated as the species directly bound to the polymer. The present study expanded that of Lundberg et a! . , particularly to lower salt concentrations and different cations, to further investigate the polyelectrolyte effects caused by binding cations along the polymer chain, as well as to elucidate the actual ion complexed to the polymer. -2 To a methanol solution containing 2.5 X 10 M in monomer of pol y ( ethyl ene oxide) of molecular weight 6 X 10 g/mole was added potassium bromide over a concentration range of 1.0 X 10" 4 M to 1.0 X 10 _1 M. The results are shown in Figure 4. Initial additions of KBr

PAGE 33

21 ds^

PAGE 34

22 are followed by an increase in viscosity. Here, cations which we proved to be bound along the polymer backbone were involved in charge-charge repulsions, expanding the chain and increasing the viscosity. However, after passing through a maximum at 2.5 X 10 M KBr, the viscosity began to decrease. The further addition of salt led to a viscosity decrease caused by counterion shielding. Division of the concentration of monomer units of poly(ethylene oxide) by the salt concentration at maximum viscosity indicated that at that point there was binding of one cation per 10 monomer units. Similar results were found by Lundberg, Bailey, and Callard. In order to be able to evaluate qualitatively the role of anion shielding, the measurements were repeated in the presence of varying amounts of the crown ether di cycl ohexyl 1 8-crown-6 . Since cation binding to crown ethers is very much stronger than to po 1 y ( ethyl ene oxide) (see Conductance section), the free cation concentration could be conveniently varied while keeping the anion concentration constant. The resulting plot is also shown in Figure 4. The maximum was decreased and displaced from n. = 1.007 at 2.5 X 10~ 3 M KBr with no crown ether to n= 0.732 at 4.6 X 10 -3 M KBr with crown ether. The value of 837 sec at 0.0 M free K + where [KBr] = [CE] was exactly the same value obtained with the solution of PEO in methanol with no KBr added.

PAGE 35

2 3 The contribution of the bromideanion in these binding phenomena was further examined by studying the 3 addition of 2.0 X 10 M tetramethy 1 ammon i urn bromide to _2 2.5 X 10 M in monomer poly ( ethyl ene oxide) in methanol. The large tetramethy! ammon i urn cation should not bind to 53 the polymer to a measurable extent. The difference between the viscosity of the poly ( ethyl ene oxide) solution in methanol and that of pol y ( ethyl ene oxide) plus tetramethyl ammon i urn bromide in methanol was negligible. This indicated that bromide ions did not bind to poly(ethylene oxide) as had been previously proposed. Light scattering studies have also been done elsewhere on these systems. These measurements confirmed the expansion of poly(ethylene oxide) upon addition of alkali cation salts in the 1.0 X 10" 3 to 5.0 X 10~ 3 M ra nae . 58 Extraction Equilibria As stated previously, crown ether complexes of alkali cation salts have been shown to be quite soluble in certain organic solvents. This provided a method for extracting salts from aqueous solutions into organic solvents containing cyclic ethers. Extraction has been efficient only if the anion is large and highly polarizable, as for instance pic rate, which has the additional advantage of absorption near 360 nm for easy determination

PAGE 36

2 4 of the amount transferred. The overall equilibrium between an aqueous solution containing alkali cation (M ), picrate ion (A ), and hydroxide ion, and an organic solution containing a cyclic polyether (CE) can be described as fol1 ows 7,59 Ke M* + A" + CE t M + ,CE,A" aq aq org ' ' org where M ,CE,A designates ion pairs in the organic phase. If the organic phase is polar enough and the concentration of the salt low enough, these ion pairs will dissociate according to K d ,CE,A org M + ,CE + A" org org If there is partition of uncomplexed polyether between the two phases, it must be accounted for in another equilibrium. Pe CE 'org CE aq The final equilibrium takes into account complexation in the aqueous phase if the concentration of crown ether'in the aqueous phase is significant. Ks aq + CE aq M + ,CE aq

PAGE 37

75 Due to the solubility of all other polymers in this study in aqueous media, the examination of cation binding by distribution equilibria was limited to pol y ( acryl oyl tyroc i d i ne ) , the polymer containing the antibiotic tyrocidine as a pendant ligand. This process, however, could only be accomplished in a semiquantitative manner. The antibiotic tyrocidine was supplied as a mixture of up to five separate components. Various separation techniques failed to isolate the individual antibiotic units of the mixture. Since the components contain varying amounts of tryptophan and tyrosine amino acid residues, which both absorb at approximately 276 n in in the UV, and since no other commonly used technique could identify the amounts of the various tyrocidine compounds in the mixture, it was impossible to determine the exact number of antibiotics attached to a polymer chain. It was assumed, therefore, that the antibiotic composition in the polymer was identical with that of the reacting mixture. The UV absorption of the polymeric antibiotic was compared with that of the antibiotic mixture to get an estimate of the number of antibiotic units per polymer. A value of one tyrocidine molecule was found for every 12 monomer units. The extraction of sodium, potassium, and cesium picrates from an aqueous layer into a chloroform layer containing the polymer was examined due to the ease with which this process could be followed by UV-visible

PAGE 38

26 spectroscopy. The picrate salts were dissolved in water at _ 3 a concentration of 1.0 X 10 H. A large quantity of the corresponding alkali chloride was also added to each aqueous solution to increase the number of alkali ions in the organic layer and also to increase the ionic strength of the aqueous layer which would prevent the chloroform layer from becoming dispersed into the water layer. Poly(acryloyl ty roci di ne ) was added to the chloroform layer at an -4 . . estimated concentration of 1.3 X 10 fi in antibiotic units. The amount of salt extracted is presented as a percent of the tyrocidine units present in the organic phase. The results are shown in Table 1. The sodium selectivity of the polymer was similar to the 4:1 sodium over potassium selectivity demonstrated by the antibiotic 49 itself in cation transport through lipid membranes. Table 1. Extraction of Alkali Picrates from Aqueous Layer into Chloroform Layer Containing Poly(acryloyl tyrocidine) at 25° Final Chloroform Layer Cation [Picrate] [Chloride] Approx. [Tyrocidine] % Extraction Initial Aqueous Layer Initial Chloroform Layer Na 1.0 X 10" 3 5.0 X 10" 2 K 1.0 X 10~ 3 5.0 X 10~ 2 Cs 1.0 X 10" 3 5.0 X 10" 2 1.3 X 10 1.3 X 10 -3 1.3 X 10 -3 15 2

PAGE 39

27 Poten ti ome try The use of cation-selective electrodes to determine the stability constants of complexes of crown ethers with various univalent cations has proven to be an effective method with systems having stability constants ranging over six or more orders of magnitude. The electrodes responded to uncomplexed ions in solution; therefore, after the addition of a ligand, the drop in the potential of the solution directly corresponded to the number of complexed ions. In order to verify the experimental procedures used in this study, binding constants were determined initially for two typical crown ethers and the results were compared to literature values. These data shown in Table 2 . Table 2. Logarithms of Stability Constants of Crown Ethers Added to NaCl in Methanol Experimental Literature Crown Ether Binding Constant Binding Constant Dicyclohexyl-18-crown-6 3.95 Isomer A = 4.08 (mixture of isomers) Isomer B 3.68 18-crown-6 3.89 4.32 The data shown in Table 2 indicate our system gave bind' ing constants close to published values.

PAGE 40

28 Potent i ometry was then used to determine the cation binding ability of the antibiotic itself. However, with this system the addition of antibiotic to a salt solution in methanol produced an increase in solution potential instead of the expected decrease. It has been demonstrated that cation-sensitive glass electrodes respond to the presence of various amino acids; therefore, the measured increase in voltage may have been the result of electrode response to the antibiotic itself. As a result, potentiometry was discontinued with this system. The binding of potassium cations to poly(ethylene oxide) in methanol was a system which could be studied by potentiometry. After addition of the polymer to a potassium chloride solution in methanol, the solution potential dropped corresponding to the binding of potassium ions to the polymer. The new potential was compared to a calibration curve to determine the concentration of remaining unbound cations. From this number, the number of bound cations was calculated and since the number of polyethylene oxide) monomer units was known, the number of monomer units per complexed potassium ion could be determined. The data are given in Table 3. As the concentration of poly(ethylene oxide) increased, the number of monomer units per cation increased. This indicated a redistribution of charges along the polymer chains resulting in fewer cations bound per polymer. These ratios of monomer

PAGE 41

2 9 Table 3. Binding of 1.0 X 10 3 M KC1 by PEO (molecular weight 100,000) in Methanol at 25° PEO in

PAGE 42

30 units per cation differ from the value of 10 calculated from conductance studies. These factors will be discussed 1 a ter . UV-visible Spectrometry It has been observed that salts of picric acid in low polarity media frequently exhibited pronounced shifts in their optical spectra upon complexation of the cation by crown ethers. These shifts were most likely due to a significant increase in the interionic distance of the tight ion pair as a result of external coordination of the cation. Further addition of a sufficiently strong complexing ether led to the formation of solvent-separated ion pairs. This process may be described by the following equation: Pi,Na + + CE t Pi",Na + ,CE * Pi~,CE,Na + I II III Complex I is a tight ion pair which upon coordination by crown ether becomes the tight ion pair II which has a slight increase in interionic distance. Further crown ether addition leads to the loose ion pair III. Examples of these species are described in Table 4. The absorption maxima of the sodium and potassium salts in THF are those of contact or tight ion pairs.

PAGE 43

Table 4. Absorption Maxintaof Pi crate Salts in C h 1 o r o • form in the Presence of Excess Crown Ethers Salt Crown A (nm) Type Complex max Na + 351 a I 15-crown-5 356 II 18-crown-6 362 II K + 357 a I 15-crown-5 378 III 18-crown-6 365 II These values refer to tetrahydrofuran solutions of these salts. The same species would be expected in chloroform if the solubility of the salts were high enough because of the relatively low cation solvating power of this solvent. The presence of 15-crown-5 in chloroform solubilized the sodium salt in the form of a crown-compl exed tight ion pair (A 356 nm), resulting in a small increase in the 1 v m a x inter ionic distance. Addition of a large excess of 15crown-5 to the potassium salt converted this species to a loose ion pair (A 378 nm) which existed as a 2:1 r m a x complex. With 18-crown-6, both salts formed tight ion pairs (A 362 to 365 nm) with the interionic distance v v max sliqhtly larger than with the 15-crown-5 sodium complex

PAGE 44

32 In order to examine the various solvation processes involved when a typical ether ligand is added to -4 sodium picrate in chloroform, a solution of 1.0 X 10 M sodium picrate in 10% THF/90% C H C 1 ., was titrated with bis [2( 2-methoxy ) ethyl ]ether or tetraglyme. The small amount of THF aided the solubility of sodium picrate in chloroform but had no effect on the absorption maximum. The results of these additions are plotted in Figure 5. As the concentration of tetraglyme increased, the absorption maximum shifted to longer wavelengths indicating small increases in the inter ionic distance of the tight ion pair. An analogue of the Benesi-Hildebrande equation for optical absorption allows the calculation of a binding constant from these shifts in the absorption maximum. The change in absorption on tetraglyme addition from the A of sodium picrate without tetraglyme, A, max can be related to the total tetraglyme concentration [TG] and the binding constant K by the equation K A n [TG] A, A plot of 1/A against 1 / [ TG ] should be linear with the intercept at the ordinate yielding A n and the gradient yielding the prodn't. KA Q . Dividing this product by A Q yields the binding constant K. Benesi-Hildebrande plots for the addition of tetraglyme to sodium picrate in

PAGE 45

33 10 ro

PAGE 46

J 4 sed 10% THF/9f) 0/ run CHCI3 are 9'ven in F lgure 6 and Figure 7 ReSUUS ° f """ P, ° tS '" 9 ive„ in Table 5 . Qver ^ ent,re concentration range e*a„,ined, the Benes i -Hi , de~ Plots „ere,i„ear an, the fcinoung C o„ stants were consistent. This is the first i nstance where shjfts ^ "e picrate spectra due to added ligand have been U! to calculate a binding constant. Ithe graph of absorption „,axi m a vs. tetrag,y„,e concentration, Fioure s » „ j , . Hgure 5, a gradual .ncrease in interionic stance was observed up to a ^ „ f 359 „.. It was ejected that the addition of 1 i gand to picra te sa 1 t -Id distinguish between the coordinating ability of the var.ous co.plexing p„, others under analysis. These "fecials incuded 1 , 2 d i,„e thoxyet hane , b i s ( 2 -„,e thoxy•thyl).th.r .r dig,,... tetrag,y,„e, PPly (ethy, one ox ide ) of various »,ec u ,ar weights, and the p , yme ri c viny , >'W Poly{3.6.9.1 2 .,5-p* nt .o«-l-hepta<(.c.n.). The results of the addition of ,. x lt) -3 Hand , Q , ^.^ 1 n 9 and (in monomer units) to 1 n x in" 4 m ; LU 'u x 10 H sodium picrate are given in Table 6 With f ho b. With the exception of DME and diglyme, there was very IUHp rfi#* ^ry nttle difference between the coordinating abilitv nf h „, • J unity of the various complexing polyethers.

PAGE 47

35 8 o ro

PAGE 48

36 K

PAGE 49

37 — i/i X o O Qr— Ol o c: +-> ra So ai •r"O CD QJ E c +-> O c/i en

PAGE 50

38 Table 6. Absorption Maxima After Addition of Ethers of Various Molecular Weights to 1.0 X 10~4 M Sodium Picrate in 10% THF/90% CHC1 3

PAGE 51

39 Conductance Pedersen and Frensdorff obtained information on the stoi ch iome try of crown ether-alkali cation complexes by -measuring conductance of salt solutions after addition 59 of crown ethers. When the polarity of the solvent was high enough and the salt concentration low enough to ensure complete dissociation of the ions, the mobility of the cation decreased as it became complexed to the crown ether resulting in a decreased conductance of the salt solution. Where the binding constant, K, defined by K M + A + CE M ,CE + A' was high, the conductance decreased linearly with an increase in [crown ether] until complexation is virtually complete. Further addition of crown ether had no effect. Such plots are shown in Figures 8 and 9. Here it is demonstrated that a break in the curve occurs at a crown to salt ratio equal to the sto i ch i ome try of the complex. Smid has also shown that addition of a polyvinyl crown ether) to a salt solution reduced the conductance more than the corresponding monomeric crown ether because of the lower mobility of the polymeric complex. Significantly in the polymer case, the equivalent conductance decreased somewhat beyond the

PAGE 52

40 o CD o a.

PAGE 53

41 3.

PAGE 54

42 stoichiometric crown ether to cation ratio. This gradual decrease beyond the break point is probably due to redis4 2 tribution of cations upon addition of more polymer, a point which will be discussed later. Several studies were performed on various molecular weights of poly (ethyl ene oxide) and on pol y ( pentaoxa1 -heptadecene ) , or P(PHD), in order to investigate some of the factors involved in conductance measurements. There are several complicating factors which can influence conductance determinations: (1) Viscosity effects on ion mobility could be a problem especially with polymers. As polymers are added to a salt solution, the viscosity of the medium may increase to the point where mobility of the ions is decreased. This would result in a drop in conductance without cations being bound to the polymer. The relationship between conductance and viscosity of a solution is given by Walden's Rule: A n n = constant This rule states that the product of limiting equivalent conductance, A „ , and macroscopic viscosity, n » of a system is equal to a constant. In order for the product to remain a constant, an increase in macroscopic viscosity requires a corresponding decrease in \_ . (2) It is also possible for the anion of the salt to effect the conductance if it were complexed by the polymer. In this case, a drop in

PAGE 55

4 3 conductance of a salt solution upon addition of a polymer would indicate a decrease in anion mobility as well as cation mobility. (3) Another possible problem is polymer conductance. If the polymer itself conducts, the conductance measured upon polymer addition would be higher than the t rue value. In order to examine the possible effects of macroscopic viscosity on conductance, a polymer was needed which does not complex cations. This polymer could then be added to a salt solution with any resulting decrease in conductance being attributed to an increase in viscosity. The polymer chosen was pol y ( s ty rene ) . The solvent was THF due to the solubility of pol y ( sty rene ) in this solvent and the fact that NaBPh, exists as free 64 ions in THF at low concentrations. The addition of -2 5 pol y( sty rene ) up to 4.0 X 10 M in monomer to 1.0 X 10 M NaBPh, caused virtually no change in conductance. Since viscosity studies demonstrated a 20% increase in macroscopic viscosity for this polymer concentration, it can be concluded from Walden's Rule that this increase in macroscopic viscosity due to the addition of polymer in this low concentration range was of no consequence in the conductance measurements. Another method used for determining the effects of macroscopic viscosity on conductance was adding polymer to a solution of tetrabu tyl ammon i urn tetraphenyl bor i de.

PAGE 56

4 4 Since the tetrabutyl ammon i urn cation does not bind to polyethers (see Viscosity section), any decrease in conductance should be attributable to the lowering of ion mobilities due to an increase in macroscopic viscosity. The conductance measured upon binding of alkali cations to the polymers could then be corrected by this value. Poly(pentaoxa-1 -heptadecene ) was added over a concentration range of 3.2 X 10~ 3 4.0 X 10" 2 M in monomer to a 1.0 X 10" M solution of tetrabutyl ammon i urn tetrapheny 1 bor i de in THF. No significant change in conductance was measured over this concentration range of polymer. These results were consistent with the various pol y( ethyl ene oxides) studied in this concentration range. For higher polymer concentrations, however, significant viscosity effects were measured. Some examples of the viscosity studies are shown in Table 7. The lack of conductance decrease on addition of low concentrations of poly(ethylene oxide) to tetrabutylammonium tetrapheny 1 bor i de also indicated the lack of anion binding. Any anion binding would have resulted in a measurable decrease in conductance over the polymer concentration range of 3.2 X 10" 3 to 4.0 X 10" M in monomer . The conductance of the polymers themselves was measured by adding the polymer to pure solvent. These values could then be subtracted from the measured

PAGE 57

4G X C 3 c 03 fO "O •!-

PAGE 58

46 conductance of the polymer-salt solutions. Examples of the conductance of several polymers are given in Table i Table 8. Conductance of Polymers in CH 3 CN at 25° Polymer Molecular Weight Polymer [Monomer] Conductance, A,inuc> PEO PEO PEO PEO P(PHD) 1,000 6,000 100,000 900,000 40,000 5.0 X 10 1.0 X 10 5.0 X 10 1.0X10 5.0 X 10 1.0 X 10' 5.0 X 10 1.0 X 10' 5.0 X 10" 1.0 X 10" -3 -3 -3 1.30 18.90 0.10 2.10 0.10 3.00 1.30 22.70 An example of the effect of polymer addition to a salt solution is shown in Figure 12. It may be contrasted with the conductance plot obtained by addition of 18-crown-6, a strong cation complexing agent of low molecular weight, to a similar salt solution in Figure I Both figures are characterized by two separate sections

PAGE 59

4 7 Initially there is a region of sharply decreasing conductance in which cation binding takes place. A comparison of initial slopes indicates that the crown ether curve is much steeper in this region than the polymer. Further addition of ligand causes a gradual decrease in both slopes. This was also found with all other polymer systems studied. The second portion of the crown ether figure is a nearly horizontal line indicating that all cations have been bound and that further crown ether addition has no effect on conductance. The second segment of the polymer curve, however, continues to decrease in conductance upon addition of more polymer. This is due to a redistribution of cations among the polymer chains. All the polymers examined except PEO 1000 gave this conductance decrease with cation redistribution. The results of the conductance studies are given in Table 9. Solvent effects . The effects of various solvents on the compl exat i on process were examined by adding -3 P(PHD) over a concentration range of 3.2 X 10 to 1.0 X 10" 1 M in monomer to 1.0 X 10~ 4 M NaBPh 4 in acetone, tetrahydrofuran , and aceton i t r i 1 e . These additions were -4 made over two concentration ranges: from 3.2 X 10 M to 4.8 X 10" 3 M and from 5.0 X 10 -3 M to 1.0 X 10 _1 M in monomer. Examples of these results are given in Figures 10-15. The results are given in Table 10.

PAGE 60

48 ^r

PAGE 61

49 — «*a.

PAGE 62

50 CD

PAGE 63

51 ^ >> <

PAGE 65

5 3

PAGE 66

5 4 P C <1) -J ro t-u 3. O O O "D c sc o cm•EI Q-OCQ 1 — 1< O O O o I I X I I X X X IX IX O 00 U3 OO O l£> — 1 — OO , — OOOOOOOOOOOOOOOOOOOO xxxxxxxxxxxxxxxxxxxx cm. — i£>i — oonroi — locOLn^in^^tn^-n 1 1 1 1 1 1 1 1 o> m cu o> ajajdjajaiQjcDQjai S-S_S-S_S-S-US-S_ 000 00000000000000 +J -l_> 4-> -l_>4_>4_l+J4_)+j+J-*-'-l->+J-l->-(-'4->-l-> aia)cuLi_Li_Li_Li_ajajQjQja)CLiQjQjQjaja)j ^n. — to. — n, — n, — 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 000000000000000000000 xxxxxxxxxxxxxxxxxxxxx oooudioooooocoo^j-^-ooooooooocoo "^t-^s-. — inirxti — ^r — <3. — m w . — 10 cx> • — en. — en • — cr> . — n^in^^i-roinro! — mi — . — 1 — ro , — oo , — ro — ro. — 0000 <~0 lO OOOO I IOOOOOOOO ccoooo » » * 3 SOOOOOOOO OO « " 'OOOO Si-i — r-u) 1D1 — 1 — cr>cr> ro o t. x 4-> O. 00000000 o.o-O-O-0_a.Q_oI— 0.0.0.0.0.0.0.0.0.0. 1 IOOOOOOOO COOOUJUJLULlJLiJUJLlJUJ r— .— 0.0-0-0-0.0-0-0-

PAGE 67

5 5 C TQJ 4-> O O

PAGE 68

56 For lower polymer concentration range, THF gave the largest slope followed by acetonitrile and then acetone. For the higher polymer concentration range, the largest slope was found in acetonitrile followed by THF and acetone. The apparent point of complete cation binding is the point at which the conductance curve approaches a horizontal line. Beyond this point, there is very little if any change in conductance upon addition of more polymer. The apparent point of complete binding was reached first in THF followed by acetonitrile and acetone. The competition between solvent and ligand for solvation of sodium cations was also examined by adding 18-crown-6 to NaBPh. in acetone and acetonitrile, Figures 8 and 9 and Table 9. As with P(PHD) the highest slope and earliest point of complexation of all cations occurred in acetonitrile. This trend does not follow that expected from comparison of either the Gutmann donor number for the solvents or the dipole moment. The Gutmann donor number of a solvent is a property which expresses the total amount of interaction between the solvent and the Lewis acid S b C 1 ^ , including contributions both by dipole-dipole or dipoleion interactions and by the binding effect caused by the availability of the free electron pair, and to some extent even steric properties of the solvent molecules. Thus the donor number is considered a semiquantitative measure 65 of solute-solvent interactions. A higher donor number

PAGE 69

57 would mean a stronger solvation of a cation. The dipole moment gives an indication of the solvent-cation inter66 actions fro in strictly ion-dipole considerations. Cation effects . Comparisons were also made between sodium and potassium ions with P(PHD) in acetoni-4 trile and THF. Salt concentrations of 1.0 X 10 M in -5 acetoni trile and 1.0 X 10 M in THF ensured the presence of free ions. ' From Table 9 it may be seen that in both solvents and for both the lower and higher polymer concentration ranges, sodium gave a larger slope than potassium. Molecular weight effects . In order to determine if there was a relationship between molecular weight and binding as measured by conductance, several polyethers of -4 various molecular weights were added to 1.0 X 10 M solutions of NaBPh,. The polyethers ranged from tetraglyme with a molecular weight of 222 to several polyethylene oxide) samples with molecular weights of 1000, 6000, 100,000, and 900,000. Tetraglyme was added over a concentration range of 3.2 X lO -3 M to 4.0 X 10~ 2 M to a 1.0 X 10" M solution of NaBPh, in acetone. The change in conductance was negligible because the mobility of the complexed ion was essentially the same as that of the uncomplexed ion. The pol y( ethyl ene oxide) samples were added to 1.0 X 10" 4 M NaBPh. in CHgCN. As with the P(PHD) samples

PAGE 70

58 mentioned earlier, the additions were made over a low concentration range of 3.9 X 10 -4 M to 4.8 X 10~ 3 M in monomer and a high range of 1.1 X 10" 3 M to 1.0 X 10 _1 M in monomer. The results of PEO 1000, 6000, and 100,000 are given in Figures 16 and 17. The initial slopes were 3 seen to decrease with molecular weight from -5.21 X 10 with PEO 1000, -5.16 X 10 -3 with PEO 6000, and -4.19 X 10 3 with PEO 100,000 to only -4.11 X 10 3 with PEO 900,000. However, the final segments of the curves, which began to approach a horizontal line, were found at higher conductance values with the lower molecular weight polymers. At 1.0 X 10" M in monomer for each polymer, the conductance values ranged from 86.00 uU for PEO 1000 and 70.84 uU for PEO 6000, to 60.30 \iV for PEO 100,000 and 60.60 yl5 for PEO 900,000. Only PEO 1000 and 6000 gave curves which actually approached a horizontal line at higher polymer concentrations. The point at which the horizontal line was approached came at a lower polymer concentration with PEO 1000. For both PEO 100,000 and 900,000 the conductance continued to decrease at high polymer concentrations. As stated earlier, this decrease was probably due to a redistribution of cations along the polymer chains. This will be discussed later.

PAGE 71

59

PAGE 72

6

PAGE 73

61 Discussion of Glymes and Po1y(ethylene oxides) UV-visible Spectroscopy From the spectral data in Table 6, it was determined that the various polyethers in this study coordinate sodium cations. This was evident from the red shifts due to an increase in inter ionic distance upon addition of ligand. These shifts were almost negligible with DME and diglyme even at higher ether concentration, but with tetraglyme and the polymeric ethers larger red shifts were observed. The absence of distinct red shifts with DME and diglyme indicated that these ethers have very low binding constants with sodium picrate in the low polarity THF/CHCU solvent. When the shifts with tetraglyme and the polymeric ethers were compared with the changes in A upon max r addition of crown ethers, Table 4, it was apparent that 4 1 1.0 X 10 M sodium picrate in the presence of 1.0 X 10 M glymes, PEO, and P(PHD) remained as a peripherally solvated tight ion pair absorbing at approximately 358 nm CO (see Results). In the peripherally solvated ion pair, the ligands are able to replace most of the solvent molecules from the anion paired cation, but coordination is not sufficiently strong to cause a large increase in interionic distance as with crown ethers. This coordination may be represented as follows.

PAGE 74

6 2 Pi"M + / X <0 Pi M J + n -t-» Here solvent molecules, 4— -> , are replaced in the cation solvation sphere by ether oxygens of a polyether. From an entropy standpoint, the coordination would be more efficient if the ethers are contained in one molecule rather than in several molecules of for example DME or d i g lyme . Basically the same red shifts were achieved at high and low ether concentrations with tetraglyme and the higher molecular weight polyethers. Therefore, there is not much discernible difference between the complexing ability of these ethers toward tight ion pairs. This indicates that tetraglyme and the polymeric ethers are able to for in equivalent complexes with approximately the same geometry and coordination number. Such a complex would be less energetically feasible with DME or diglyme due to unfavorable entropy requirements. These results moreover indicated that cation coordination is not strongly dependent on whether the ether oxygens are along the polymer backbone or in a pendant ligand. A slight blue shift was found with the polyethers of molecular weight above 1000 compared to the largest

PAGE 75

6 3 X values which were obtained with tetraqlyme, PEO 600, max 3 J and PEO 1000. This blue shift could be attributed to stacking of the picrate anions along the polymer chain as they remain in the vicinity of the complexed cations if the ab69 sorption decreased with ligand addition. However, the absorption remained constant. This blue shift, therefore, could not be readily explained. Vi scome try It is well known that the viscosity of polymer solutions is related to polymer dimension with an increase in chain dimensions resulting in an increase in viscosity. Lundberg, Bailey, and Callard interpreted the increase in viscosity of poly(ethylene oxide) in methanol on addition of alkali ha 1 i de salts as due to chain expansion caused by 3 7 binding of the anion to the polymer. Since there is substantial evidence for binding of alkali cations to glymes, PEO, and crown ethers, it was decided to investigate the system further. 2 ' 3 ' 8 ' 10,25,35 ' 59 Using a system of 2.5 X 10 M in monomer of PEO having a molecular weight of 6 X 10 6 g/mole with addition of 1.0 X 10" 4 M to 1.0 X 10" M KBr in methanol, the results of Lundberg e_t a]_. could be reproduced, particularly the value of one salt molecule for every ten monomer units at maximum viscosity. This number, however, is calculated by assuming complete binding of all cations at maximum viscosity. The addition of tetrame thy 1 amnion i urn bromide to PEO in methanol gave different results. Although the te trame thyl ammon i urn cation

PAGE 76

64 53 should not bind to the polymer, ' it was expected that increased viscosity would still be measured due to anion binding if Lundberg e_t aj_. were correct in their data interpretation. The negligible change in viscosity upon addition of te trame thyl ammon i urn bromide to PEO indicated that the cation is the ion bound to poly ( ethyl ene oxide). There are several processes which are expected to determine the dimensions of a pol y ( e thy 1 ene oxide) coil in the presence of alkali cation salts. These processes are shown in Figure 19. As alkali cations are bound to poly(ethylene oxide), positive charge density builds along the polymer chain. As this charge density increases, mutual charge-charge repulsions cause the polymer coil to expand. This expansion, if all other factors are negligible, would continue until the charge density is high enough to prevent more cations from entering the polymer domain. At the same time, however, anions are also being introduced into the vicinity of the polymer. At low salt concentration, they have little effect but as the salt concentration increases, the presence of these anions begins shielding the cha rgecha rge repulsions of the cations along the chain. A third factor is the increased shielding at larger positive charge density due to the requirements of e 1 ectroneu tra 1 i ty in the polymer domain. As more cations are complexed by the polymer, the large charge density inside the polymer domain causes anions to move into the polymer domain reducing the high positive

PAGE 78

(>(, c

PAGE 79

67 charge. The net result of these effects is an expansion of the polymer upon initial salt addition which reaches a maximum and then begins to decrease as shielding becomes very effective at high salt concentration. This is demonstrated in Figure 4 with the addition of KBr to PEO of molecular weight 100,000 in methanol. The effects of shielding become more evident upon comparison of the PEO/KBr system to the PEO/KBr system containing d i eye 1 ohexyl 1 8-crown-6 . The 1:1 addition of crown ether to KBr in the PEO solution in methanol produced a viscosity reading equal to that of PEO in methanol ' wi thou t any added salt. This indicated that crown ether was able to effectively remove all potassium ions from the polymer domain due to its stronger binding ability. Therefore, the addition of crown ether to a solution of PEO and KBr in methanol allowed a viscosity study in which free K was varied while the anion concentration remained constant (Figure 4). The crown ether system produced a curve in which the maximum was decreased and displaced to higher concentration of free potassium ions. The occurrence of a maximum indicates that shielding becomes more prominent as positive charge density increases along the chain at constant anion concentration. This is expected from the behavior of a typical polyel ec trolyte system. The decreased maximum indicates that the repulsions are reduced compared to the system without crown ether so that

PAGE 80

68 the anion concentration appears to have a moderate effect on polycation shielding. The increase in salt concentration at maximum viscosity in the presence of crown ether may be due in part to more cations needed to produce a particular expansion at the high anion concentration, but this in inself should increase the number of anions attracted into the polymer domain. This particular effect is not well understood. Potent i ome try A system similar to the one above was studied by ion-selective electrode potent i ome try (Table 3). The basic differences here are a lower molecular weight PEO (100,000) and use of the chloride salt instead of bromide. The concentration of polymer and salt are in the same range here as with the viscosity studies but this system differs from the viscosity system because here polymer is added to the salt solution. The use of chloride anion was necessary because the Ag/AgCl reference electrode was susceptible to contamination by bromide ions. Upon addition of PEO from 1.14 X 10~ 2 M to 4.14 X 10~ 2 M in monomer -3 to 1.0 X 10 M K C 1 in methanol, the number of monomer units per cation bound was found to increase from 29 to 43 as the amount of polymer increased. These values did not agree with that of ten monomer units per cation found in viscosity studies. Since both polymers are of high

PAGE 81

69 molecular weight, it may be assumed that they would complex approximately the same number of cations for a particular number of monomer units. However, in viscosity studies of Lundberg e_t aj_. it was assumed that all added cations 3 7 were bound to the polymer. The potent i ometry results indicated that this was not the case. Conductance The complexation processes involving the binding of cations to polymer chains may be described by the following equilibria: M + P KPM. M + PM. PM 2 + M + + PM 1 ."] l 1 pm: i + Here PM-, represents the polymer containing only one cation and PM. represents the attachment of the i cation to this polymer chain. There may be large numbers of cations attached to a single chain in the high molecular weight polymers, but charge repulsions along the relatively small chains of the low molecular weight polymers reduce the chances of large numbers of cations binding to these systems.

PAGE 82

7 The decrease of conductance of a salt solution upon addition of a polymer indicates coordination of one or more cations to the polymer resulting in a lowering of the ion mobilities (see Figure 11). This continues until all cations have been complexed. Addition of more polymer at this point should cause redistribution of cations from polymers containing more than one cation to polymers with no cations. This would be due to the higher energy of polymers containing more than one cation resulting from charge-charge repulsions along the polymer chain. At very high polymer concentration this redistribution due to PM i + i P + i PM, is virtually complete and the conductance is not expected to change upon further addition of polymer. The measured conductance is now the sum of the limiting conductance of a polymer chain carrying one cation A , and the limiting conductance of the anion, Binding to PEO 1000 . The simplest example of the conductance decrease is given with PEO 1000 which had a 1/R vs. [polymer] conductance curve similar to that for 18-crown-6 in the complexation of NaBPh. in acetonitrile (Figures 3 and 11). In both cases, the initial section of the curve describing cation binding was nearly linear; however, the initial slope was much larger with the crown ether. The transition from this linear segment to a horizontal segment indicating one cation per ligand occurred

PAGE 83

71 much higher in monomer concentration of PEO than with crown ether but was very rapid with both PEO 1000 and 18-crown-6. The lack of a prominent redistribution region with PEO 1000 indicated that the polymer probably complexed a very small number of cations per chain even at low polymer concentration, which was not unexpected due to the small size of the chain. The binding constant K , for the complexation of one sodium cation to such a polymer may be calculated as fol1 ows : K P + PM The equivalent conductance for a system, A , is defined as 1000 k cR where k = the cell constant, c = the salt concentration 70 and R = the solution resistance. 1000 < cf • A Let R be the resistance of the solution without o added polymer; then,

PAGE 84

72 c-f -A o c-f (A" O If a = the fraction of cations bound, then at a resistance R, the contribution to conductance of the cations complexed by the polymer = cf-a-A + , and the contribution of free ions = c • f • ( 1 a ) A + . o Therefore , c-f L(l a)A + a\* i + A J O 0,1 f -c-a(A A o,l } = aA [PM + ] [PM + ] + [M + ] [P][M ] mil a]c Therefore , K[PJ_ K[P] + 1 so that A(l/R K[PJ+T A

PAGE 85

w h e r e 73 J salt 1 limit (i) R ; R J sal t 1 salt + polymer so , A MITT l i k ' ITT + i Thus a plot of A/A(l/R) vs. 1/[P] is expected to give a line with slope = 1/K. The plot forPEO 1000 over a free polymer concentration range of 8.81 X 10 -5 4.84 X 10 M is given in Figure 20. The line was found to have a slope of 1.33 X 10 and a value of Ki = 7. 52 X 10" 3 . Binding to PEO 6000 . Po ly ( ethyl ene oxide) of molecular weight 6000 had a 1/R vs. [polymer] conductance plot similar in many respects to that of PEO 1000 (Figure 11). The initial slopes of both curves are very close with PEO 1000 having a slightly larger slope. Also, the concentration in monomer units at which PEO 1000 levels off at limiting conductance is just below the concentration where PEO 6000 begins to approach its limiting

PAGE 86

7 4

PAGE 87

7 5 conductance (Table 9). Unlike PEO 1000, as more PEO 6000 is added in the region following complete compl exat i on , there is a gradual decrease in conductance. This indicates that all the cations are indeed bound to polymer chains but some cations are being redistributed (see bel ow) . The complexation constant for the binding of one cation to a polymer may be calculated as with PEO 1000. The A/A(l/R) vs. 1/[P] plot is shown in Figure 21. The curvature at low polymer concentration is due to more than one cation being complexed per polymer chain. The results of this plot are shown in Table 11 where they are compared to the values for PEO 1000. Table 11. K] for NaBPh 4 + PEO 1000 and 6000 in Acetoni tr i 1 e Polymer SI ope PEO 1000 PEO 6000 1 . 33 X 10 1.35 X 10 7. 52 X 10' 7.41 X 10 The increased binding constant for the larger molecular weight polymer is expected for statistical reasons. This may be explained from the following.

PAGE 88

76 -\o<\

PAGE 89

77 Let K 100Q K] for PEO 1000 and K 6000 = K] for PEO 6000 00 AG 1000 /RT K 6000 " AG 6000 /RT e In '1000 AG 1000 6000 RT AG 6000 RT AH 1 000 " AH 6000 + AS 1000 " AS 6000 RT Since the cation in each polymer is being complexed to a site which is equivalent physically to that in the other polymer, AH 1 000 AH 6000 S^k In W, where W is the possible number of ways to complex the cation. Since AS 1 000 " AS 6000 In W 100Q In W g000 = In 1000 6000

PAGE 90

therefore , 7 8 1000 6000 w 1000 6000 The total number of possible ways to coordinate the cation with the polymer is given by W = q![p!(q p)!] , where q represents the number of oxygen atoms per polymer molecule and p the average number of oxygen atoms coordinated to the cation. However, not all combinations of oxygen atoms are equally probable. When ethers in the chain of a polymer are separated by more than two carbon atoms, the solvating power of the polymer decreases sharply. Therefore, only those combinations involving consecutive coordination sites are important. The number of these is W = q p + 1 . 71 For sodium, the coordination number p = 5. PE0 1000 has 22 oxygen atoms with W, 0Q0 = 18. PE0 6000 has 136 oxygen atoms giving W finnn = 132. Therefore, the calculated K crir . n = 7.3 K, nnn . From experimentally determined 6000 1000 r J binding constants, it was found that K rnrvr , = 9.85 K irinn . 6000 1 000 The small decrease in conductance noted at high polymer concentration on the 1/R vs. [polymer] conductance plot for PE0 6000 has been described as resulting from cation distribution from a polymer containing more than one cation to a polymer containing no cations.

PAGE 91

7 9 This process may be described as follows 2 PMj t PM 2 + + p [PM^ + ][Pj K 2 [pm!j 2 k i write Assuming the absence of unbound cations, we can + ? + the salt concentration = P M , + 2 PM^ The fraction of salt complexed as the dication, 2LPM 2+ ] [PM|] + 2 [PM 2 /] so that [PMo + ] Now, f.c[(l B)x; §1 + SA 2 %j 2 + where A ,, ^ s tne limiting conductance of a polymer chain 2 + carrying two cations. A „ is dependent on the mobility

PAGE 92

of polymeric PM 2 + 2 ' 80 If it is assumed that there is no change in polymer size upon adding a second cation, the limiting conductance of a polymer containing two cations, A ^ , should be ^ 2 A ,. o,2 o,l Thus , i f.e(l B)xJ f] Si nee K * = BB/2).c][P] [1 6] 2 c 2 2K*c 2 (l B) 2 =6 cP Now if 8 < < 1 , 2K*c Substituting for 8 and differentiating 1/R with respect to 1 / P , we obtain s ( 1 / R ) Ml/LPT f-c A 0,1 i/Cpj 2fc2A o,l K ' A plot of 1/R vs. 1/[P] over the polymer concen-2 1 tration range of 3.3 X 10 M to 1.0 X 10 M in monomer

PAGE 93

PEO 6000 where no free ions are present gave a straight 2 + line with a slope = 2fc K*X , (Figure 22). K* was then calculated and since K* = K ? /K,, K ? was also calculated. It should be noted that the plot was not linear for polymer concentrations below the range listed above. This deviation is probably due to complexation of more than two cations per polymer causing 1/R to be lower than expected. The results for PEO 6000 are given in Table 12. Table 12. Binding Constants for 'Complexation of Ma to PEO 6000 in CMgCfl Polymer Kl Slope of 1/R vs. 1/[P] K* PEO 6000 7.41 X 10 4 8.18 X 10" 4 0.235 1.74 X 10 -4 Binding to high molecular weight polymers . The higher molecular weight polymers, PEO 100,000 and PEO 900,000, gave different 1/R vs. [PEO] conductance plots upon addition to salt solutions than the lower molecular weight polymers (Figures 11 and 12). PEO 100,000 had an initial slope slightly larger than PEO 900,000 but quite surprisingly both of these slopes were significantly smaller than the slopes of PEO 1000 and 6000. Also, in

PAGE 94

— |q_ u o o

PAGE 95

83 contrast to PEO 1000, the high molecular weight polymers did not approach a horizontal line at high polymer concentration, but instead both showed a continual decrease. These decreases at high polymer concentration with the high molecular weight polymers were much larger than with PEO 6000. Furthermore, the conductance of the PEO 100,000 and 900,000 solutions at high polymers concentration was much lower than the conductance of the PEO 1000 and 6000 systems . At low polymer concentration, the slopes decreased with increasing molecular weight. This can be explained by considering two factors. First, viscosity measurements indicate that at these salt concentrations (% 10 M), the size of the polymers is not greatly changed on cation binding so that the conductance per ion increases with the molecular weight assuming a constant number of ethylene oxide units per cation site. Second, the mobility of polymers in solution decreases with molecular weight but this decrease is less than linear. For instance, for a random walk polymer the diffusion coefficient varies as M . Therefore the equivalent conductance of polymer bound cations is expected to increase with molecular weight. This leads to lower slopes at low polymer concentration for higher molecular weight polymers. At high polymer concentration, the magnitudes of the slopes were found to increase with increasing molecular

PAGE 96

84 weight. After all cations had been complexed, the lower molecular weight polymers, due to their small size, contained relatively few cations per polymer. As more polymer was added there were relatively few cations transferred to the uncomplexed polymer chains due to the low concen2 + tration of P M „ in the following equilibrium: 2 + PM^ + P 2 PM The low number of cations transferred caused little change in transport of charges or mobility of polymer bound cations. With the high molecular weight polymers, assuming constant binding of cations per unit weight for all polymers, there were many cations per polymer chain. As more high molecular weight polymer is added, there were many cations redistributed from polymers containing many cations to polymers containing fewer numbers of cations. This led to a greater decrease in conductance than with the lower molecular weight polymers. There were several factors which caused difficulty with the calculation of binding constants with the higher molecular v/eight polymers. Since the polymers are so large and so many cations can be bound per polymer chain at low polymer concentration, the binding and redistribution processes occur simultaneously over most of the polymer concentration range. This leads to many equilibria

PAGE 97

85 being established which cannot be properly evaluated. The binding processes themselves, as can be shown from earlier calculations, would lead to an expression for conductance which can be described as follows: = f-c[(l + + k 2 + ' A o +6 l A o,l + T A o,2 + T A o + ,i + •J where 6is the fraction of salt complexed to the polymers containing i cations. It is clearly impractical to evaluate the binding constant K . using this equation. In addition, complications arise from probable 72 site binding of counter ions to the polycation. Such effects were evident from the work of Lundberg, Bailey, and Ca 1 lard, who showed specific anion effects on the vis37 cosity of cation bound po 1 y ( ethyl ene oxide). Similar anion effects have been observed in light scattering e x 58 periments on these systems. Thus a quantitative or even semiquantitative description of these high molecular weight polymers is difficult on the basis of the above conductance methods. Mobility of PM , complexes . For the low molecular weight PEO 1000, it is possible to calculate A , , the limiting conductance of the polymer containing one cation. As was apparent from a comparison between the 18-crown-6 and PEO 1000 1/R vs. [PEO] conductance curves

PAGE 98

8 6 « h. 9 h P o, y ,„e r concentration, it „ ^^ ^ ^ on.y one s„o (u ,„ i0 „ per poly| „ e ,,. It has ^^ shom c-f -A C'f (A n + A") O o Where A o Na + = 77 -3 and A~ . » n fi7 °' Na o,BPh 58 ] in acetonitrile. 67 At the limiting conductance, A ° = Vl + A o" Therefore , c-f(A o , 1 ' so that F or PEO 1000, o,l 36.6 For PEO 6000, H was shown foP t(le ^ „,„,.„„„ "at a p„ t of ,/R vs. ./[p] gaue , strajght Hn§ (Fjgure 22)The interc.pt of this p, ot at [PEO] . gives the "'tin, conductance »,_, »; which equal, 69.6. Suhtracting the value of l" f nr nok" + A Q for RPh 4 g 1V es a value of 11. 5 f0r A o,l for PE0 6000.

PAGE 99

Since the X , values are a function of the o , 1 mobility of polymer bound cations, if a randomly coiled -1/2 polymer is assumed they should be related by M where M is the molecular weight. The ratio of A„ , for PEO J o,l 1000 and 6000 is 3.18 compared to the ratio of their molecular weights of 2.45. Although this method does not give the exact relationship between A , values for various molecular weight polymers, it can nevertheless be used to approximate A , for the high molecular weight polymers. For PEO 100,000 the A Q -, value is found to be approximately 3.2 and for PEO 900,000 it is approximately 1.2. Therefore, at high polymer concentration of high molecular weight polymers 1/R approaches A . Discussion of Poly(3,6,9,12,15-pentaoxa-l-heptadecene) From inspection of Table 6, it appears that the A of sodium picrate in 10% THF/CHC1-, in the presence max -i of P(PHD) was the same as that for equal quantities in terms of monomer units of pol y ( ethyl ene oxide). As stated earlier for pol y ( ethyl ene oxide), the relative shifts involved led to the conclusions that the polymers are only peripherally solvating the sodium cation in this solvent and that ion pair complexation constants are about the same as with PEO.

PAGE 100

8 8 From the conductance data of P(PHD), Figure 16, the P(PHD) concentration at which the conductance curve approaches a constant level" was approximately 3.0 X 10 M in monomer in acetonitrile which was very close to the _2 value of 2.6 X 10 M in monomer for PEO 1000. This is further evidence for the similarity of the binding constants of P(PHD) and PEO. The 1/R conductance value where the curve approaches a "horizontal" line, however, was much higher with P(PHD) than with any of the PEO polymers. For example, the P(PHD) plot began to level off at a value of 90.79 \il5, whereas PEO 100,000 was still decreasing slightly at 60.6 u U . Since the sodium cation may be effectively complexed in a 1:1 s to i c h i ome try by a glyme containing five oxygens, it may be coordinated to a single 2 pendant ligand on P(PHD). In order for a complexed sodium ion to move along this polymer chain, it would only be necessary to shift from one pendant glyme to an adjacent one without involving a major conformational change in the polymer. With PEO, however, a conformational change in the polymer chain would be required to move a sodium ion from a binding site composed of five oxygen atoms in the polymer backbone to another site farther along the polymer containing another five oxygen atoms in the proper conformation. From entropy considerations, transport of a sodium ion along the chain would be easier with P(PHD) which would lead to a higher limiting conductance.

PAGE 101

Sol vent Effects From conductance studies in various solvents, P(PHD) demonstrated stronger complexation of sodium ions in THF, followed by acetonitrile and then acetone (Table 10). This trend is not due primarily to any typical polymer effects since the results with acetone and acetoni trile correspond to the results obtained by adding 18crown-6 to NaBPh. in the same solvents (Figures 8 and 9 and Table 9). In the polymer study, the THF sample exhibited the largest slope but had a much smaller A[(l/R ) (1/R)]. This indicates that THF coordinates the sodium cation very strongly, however, since NaBPh. 4 exists primarily as ion pairs at 1.0 X 10 M in THF, complexation of the salt to a polymer would increase the 43 64 ion separation leading to higher conductance. Therefore, complexation of the salt would not be expected to give a large drop in conductance. The above trend for acetonitrile and acetone is hard to justify in terms of commonly used parameters such as the Gutmann donor number or dipole moment since acetone has a higher donor number than acetonitrile but acetonitrile has a higher dipole moment. The overall solvent effect is probably the result of a combination of solvent basicity and dipole moment.

PAGE 102

90 Cation Effects In conductance studies in both acetonitrile and THF, larger slopes were found for complexation of sodium ions by P(PHD) than for potassium ions (Table 9). This indicates sodium selectivity, which agrees with the fact that the monomeric analog of P(PHD), tetraglyme, was found to have a higher equilibrium constant with sodium ions than with potassium ions in the formation of separated ion pairs of fluorenyl salts in THF. The value for sodium 2 was 170 vs. 0.35 for potassium. Discussion of Poly ( aery 1 oyl tyroc i d i ne ) For reasons outlined in the Results section, quantitative analysis of po ly ( aery 1 oyl tyroc i d i ne ) was very difficult. The insolubility of this polymer in aqueous media made possible the use of extraction equilibria to determine the cation binding ability and selectivity. This was the only polymer which could be studied in this manner. Po ly ( aery 1 oyl tyroc i d i ne ) was found to transfer larger amounts of sodium picrate than potassium picrate from an aqueous layer into a chloroform layer containing the polymer and thus appeared to be sodium selective. The cesium cation was found to be too large to be complexed by the cyclic antibiotic.

PAGE 103

CHAPTER III PHASE TRANSFER REACTIONS As mentioned previously, polymers which act as catalysts in organic reactions have many advantages over low molecular weight catalysts. For example, with reactions involving alkali metal salts, polymers may act as catalyzing agents by complexing cations to an extent which in many cases is as strong or stronger than the corresponding low molecular weight analogs or by interacting with solid phase reagents in a manner more energetically favorable than low molecular weight compounds. From the information obtained on binding of cations to nonionic polymers, it may be possible to obtain additional insight into the polymeric catalysis of reactions involving alkali metal salts. Many alkali or alkali earth salts have been used in organic syntheses, but because of their limited solubility in organic solvents, their use was frequently under inefficient conditions such as low concentrations 5 9 or two-phase operations. Following Pedersen s discovery that crown ethers formed complexes with alkali metal salts that were soluble in aprotic solvents due to

PAGE 104

92 this complexation, attention has been focused toward utilizing the anion of the complex for synthetic p u r 29 poses. It was previously found in these solvents that crown ethers could complex the cation strongly enough to reduce the degree of interaction between the cation and 73,74 anion of contact ion pairs. In organic aprotic solvents where they were unencumbered by strong solvation processes, many of these anions have proven to be strong ? 1 3 1 nucl eophi 1 es . The general reaction scheme is shown in the following set of equations for a heterogeneous system : ENa)V . . . + RY — RX + (ENa)V ' solution III NaX sol id NaY ... + E sol id where RY is a typical alkali halide substrate, (ENa) Y~ is a crown ether complexed alkali metal salt, and RX is the product of the nucleophilic attack of X" on RY. Since strong cation coordination played such an essential role in phase-transfer reactions, it was anticipated that the various polyethers would catalyze phasetransfer reactions to the same relative extent to which they coordinated alkali cations. The reaction chosen for

PAGE 105

9 3 studying this process was the nucleophilic displacement of bromide anion from an alkyl bromide by acetate anion in acetonitrile. This reaction involved sodium acetate which was insoluble in acetonitrile, butyl bromide, and the following polyethers all of which were soluble in this solvent: DME, diglyme, tetraglyme, pol y ( ethyl ene oxide) of various molecular weights, poly(3,6,9,12,15pentaoxa-1 -heptadecene ) , 18-crown-6, and poly ( v i nyl benzo15-crown-5). Po ly( v i nyl pyrrol i done ) was also studied in order to compare the coordinating ability of another type complex ing site. The acetate salt was chosen because of the most commonly used nucleophilic anions in ace ton i7 ft trile, acetate is the most effective. Also, in phasetransfer reactions involving the acetate anion in acetonitrile, virtually no alkene elimination product could be detected. Butyl bromide was chosen as the substrate since bromides proved to be more reactive than the corresponding 3 1 tosylates or chlorides in this system. In addition, butyl acetate had a sufficiently low boiling point to facilitate analysis by gas chromatography. The ability of polymers to catalyze this reaction was further studied in homogeneous media by adding 5% water to each system. This allowed comparison of polymer catalysis in which the polymer had to interact with an insoluble salt matrix with catalysis in which the salt was already in solution. The percent yields were measured with gas chromatography

PAGE 106

94 by comparing the peak areas of butyl acetate and butyl bromi de . Homogeneous Reactions Resul ts The simplest case of polymer-catalyzed reactions would involve only step II of the scheme presented above. This would be a homogeneous reaction involving complexation of salt molecules already in solution. It was found that upon addition of 5% water to the mixture of 0.037 M butyl bromide, 0.096 M sodium acetate and polyether catalyst in acetoni tri le , sodium acetate dissolved completely giving homogeneous solutions. Here, the catalyst did not have to interact with the surface of a salt matrix, but possible complications due to water catalysis had been introduced. In order to examine this problem, the reaction of sodium acetate and butyl bromide was carried out in acetonitrile containing various amounts of water. No additional catalyst was added. The results are given in Table 13. Water is shown to catalyze the reaction to a significant extent. Apparently the effect levels off above 10« water. Also, the reaction does not proceed farther after two days. These results indicate that an equilibrium may be present which prevents the reaction from going to completion.

PAGE 107

95 Table 13. 3.7 X 10" 2 M Butyl Bromide + 9.6 X 10~ 2 M Sodium Acetate in Acetonitrile with H ? Added Percent H 2 5 hrs Percent Reaction (+ 3%) 8 hrs 1 day 2 days 3 days 15 16 32 54 58 10 20 47 48 76 80 84 85 83 85 The dependence on polyether concentrations was examined with P(PHD) and also tetraglyme in the homogeneous reaction media. In both cases, there were virtually no changes in product yield over the concentration range of 1.6 X 10 to 2.5 M in monomer P(PHD) and from 1.8 X 10" 1 to 2.7 M tetraglyme with 3.7 X 10" M butyl bromide and 9.6 X 10 M sodium acetate. The reaction of butyl bromide and sodium acetate catalyzed by the various polyethers was carried out in acetonitrile containing 5% water. Here the catalytic ability of the polyethers was expected to depend primarily on their ability to effectively coordinate the sodium cation. The results of these reactions are given in Table 14 and Figure 23.

PAGE 108

% I • I UJ Kg. i— • — i i — • — i Ui9 CL S££ D oo ° LUO u a. vd u — © — I I— • — I l— • — " IT) ro CM m cr

PAGE 109

97 Table 14. 3.7 X 10~ 2 M Butyl Bromide + 9.6 X lO -2 M Sodium Acetate Catalyzed by 0.18 M Polyethers in Monomer in 5% H 2 0/95% CH 3 CN Polyethe r Percent Reaction (+ 3% 5 hrs day 3 days DME 23 D i g 1 y m e 2 3 Tetraglyme 2 5 PEO 600 30 PEO 1000 34 PEO 4000 31 PEO 6000 32 PEO 10 5 31 PEO 3 X 10 5 PEO 9 X 10 5 P(PHD) 40,000 (1.0 M) 69 6 7 67 7 3 86 76 7 8 7 7 77 76 75 80 82 80 84 83 81 8 6 8 7 70

PAGE 110

98 Since the results for one and three days are relatively close together, the reactions must be approaching equilibrium after one day. Therefore, the results after five hours should give a better indication of relative catalytic abilities. The maximum catalytic ability was achieved with PEO 1000 with the other polymers all giving slightly lower amounts of product. Discussion It is apparent from the results that an equilibrium exists in the homogeneous system which prevents the reactions from going to completion. This equilibrium condition is being approached after one day due to the similarity of results for one and three days. The fact that an equilibrium does exist in the presence of water indicates that water may have a "leveling effect" on the 75 nucl eophi 1 i c i t ies of the acetate and bromide ions. Hydrogen bonding by water may cause the nuc 1 eophi 1 i c i ti es of the anions to become similar, allowing the reverse reaction to take place. The data indicate that the low molecular weight glymes, DME and diglyme, catalyze the reaction to a lesser extent than the polymeric ethers. This correlates well with the UV-visible and conductance studies on cation binding. As shown earlier, DME and diglyme have a weaker complexing ability toward sodium cations than PEO due to a more negative entropy of binding.

PAGE 111

99 There is very little difference among the various high molecular weight polyethers in their catalytic ability in the presence of water. This may be the result of efficient coordination of the cation by water molecules. This strong solvation by water may interfere with the binding process in a manner that would equalize the binding ability of the various molecular weight polymers resulting in approximately the same yields for each molecular weight. Heterogeneous Reactions Resul ts A heterogeneous reaction involves removal of salt from a solid matrix by the catalyst, sufficient coordination of the salt by the catalyst to activate the anion, and then depositing of the byproduct salt out of the solution phase. The same reaction examined in the homogeneous reaction section but without water present was found to proceed through these steps. It has been found by using flame photometry that the solubility of potassium acetate in pure acetonitrile at 25°C is 5.0 X 10~ 4 M. 3 Although in the present study sodium acetate is used and the acetonitrile solution is maintained at reflux temperature, visual observation indicates that the concentration of sodium acetate which

PAGE 112

100 dissolved without compl exation was not significant when compared to the overall salt concentration of 9.6 X 10 M used in these reactions. _2 In the reaction of 3.7 X 10 M butyl bromide with 9.6 X 10" 2 M sodium acetate catalyzed by 1.0 M P(PHD) or 1.8 X 10" 1 PE0 600, both polymers gave linear plots of log [butyl bromide] vs. time, indicating the reactions were first order in butyl bromide. The reaction order in salt could not be determined because of the low solubility of sodium acetate in acetonitrile. No significant concentration dependence was found for the polyethers P(PHD) and tetraglyme over the range 1.6 X 10~ M to 2.0 M in monomer, The relative efficiency of the various polyethers in catalyzing the reaction between butyl bromide and sodium acetate was measured in acetonitrile. The results are presented in Table 15. At this polyether concentration the higher molecular weight polymer solutions were too viscous for gas chromatographic analysis. For the glymes and polyethylene oxide) polymers, maxima are apparent for the 600 and 1000 molecular weight polymers. Poly ( pen taoxa -1 heptadecene ) , in which the oxygens are on ligands pendant to the polymer backbone, gave catalytic results comparable to the higher percentage PE0 polymers. In order to compare the efficiencies of the higher molecular weight poly ( ethyl ene oxides), the glymes and

PAGE 113

Table 15. Polyether Catalyzed Reactions of 3.7 X 10 M Butyl Bromide + 9.6 X 10" 2 M Sodium Acetate in Acetonitrile Polyether (1.0 M in monomer' Percent Reaction (+ 3% 3 days 4 days No catalyst DME a D i g 1 y m e Tetraglyme PEO 600 PEO 1000 PEO 4000 PEO 6000 P(PHD) 1 1 4 3 2 8 39 6 4 59 5 2 49 55 58 46 6 5 78 78 a The values for DME are proportionally higher than values measured later at lower concentration. Experimental error may be involved here.

PAGE 114

102 PEO samples were run at lower polymer concentration, 1.8 X 10" M in monomer units. The results are given in Figure 24. Here again the highest conversion was achieved with PEO 1000 followed by a gradual decrease in product as the molecular weight increased. The catalytic abilities of various other polymers were examined in this heterogeneous media. The results are gi ven in Table 16. Table 16. Reaction of 3.7 X 10 M Butyl Bromide with 9.6 X 10' 2 M Sodium Acetate in CH CN Catalyzed by Various Polymers Polymer (1.0 M in monomer] Reaction Time Percent Yield (+ 3%) P(PHD) P(PHD) under dry N„ Poly( vinyl pyrrol idone' PEO 600 PEO 1000 4 days 4 days 4 days 4 days 4 days 49 52 77 78 78 Catalysis by P(PHD) was not significantly changed when traces of water were removed from the system. Poly( vi nyl pyrrol i done) apparently had a catalytic ability similar to PEO 600 and 1000 which were the most effective PEO catalysts.

PAGE 115

103 " > H I • 1 I • 1 I • 1 " • 1 l « 1 I L O m ro O ro m CVJ o >o CM — no

PAGE 116

104 In order to determine the relative catalytic ability of the various molecular weight PEO polymers and P(PHD) with respect to crown ethers, the reaction was run in the presence of 18-crown-6 and the polymeric crown ether, po ly ( v i ny 1 benzo1 5crown-5 ) . Due to the ability of these materials to selectively complex potas42 sium ions over sodium ions, the reactions were also run with the potassium salt. The results are given in Table 17. Table 17. Catalysis of 3.7 X 10 M Butyl Bromide + 9.6 X 1 " 2 m Sodium and Potassium Acetate i n Aceton i tri 1 e Ligand

PAGE 117

5 derived from sodium and potassium acetate. The results are i ncl uded in Table 17. Even with much lower ligand concentration, the powerful cation complexing agents catalyzed the reaction to a larger extent than did PEO or P(PHD) after three days. With both crown ethers, the potassium salt gave higher percent reactions. Discussion The variation in percent reaction with the various polyether catalysts indicates that there is an absence in leveling effects when no water is present. Also, the increase in reaction between three and four days from Table 15 indicated that no equilibrium was being approached after that length of time. Since these reactions are much slower than those in homogeneous media, failure of the crown ether solutions to go to completion may have been due to insufficient reaction time rather than an equilibrium being established. Once again, the relatively low reactivities with the low molecular weight glymes correlate well with the argument of less favorable entropy of complexation with these polyethers which was brought out in the UV-visible and conductance studies. The variation in the catalytic ability of the poly(ethylene oxide) samples could possibly be the result of two factors. The lower molecular

PAGE 118

6 weight polymers may be more efficient in their transport of salt from the crystal into the solution phase. Also, the high molecular weight polymers may complex the salt less efficiently once it is in solution. Either or both of these effects would lead to higher reactivities with the lower molecular weight polymers. From Table 16, it may be seen that P(PHD) produces approximately the same percent yield as the low molecular weights of PEO. This agrees with the cation complexing abilities of these polymers as determined by UV-visible and conductance studies. The reaction between butyl bromide and sodium acetate in the presence of crown ethers gave results which could be predicted on the basis of the cation complexing ability of the cyclic ethers. Even small quantities of the powerful cation coordinating crown ethers gave very high yield reactions. Polyvinylpyrrolidone), which has been shown to catalyze the William45 son reaction between sodium phenoxide and butyl bromide, gave a high yield equal to the most efficient PEO polymers. Apparently, the pyrrol idon-e units are quite efficient in coordinating the sodium cation due to their high dipole moment . The higher yield reactions demonstrated by potassium salts as compared to sodium salts with P(PHD) were directly opposite to the relative degree of binding of these salts determined by conductivity. This may be

PAGE 119

107 compared to the higher yields with potassium salts in the presence of crown ethers which have demonstrated potassium selectivity with respect to cation binding. Although more sodium ions are bound to P(PHD), there are apparently other factors involved which reduce the interaction between the acetate and potassium ions more than with sodium ions. There should be less electrostatic interaction between potassium and acetate ions compared to the interaction between sodium and acetate ions so that the apparent higher reactivity of the complexed potassium salt is reasonable.

PAGE 120

CHAPTER IV EXPERIMENTAL PROCEDURES Preparation and Purification of Materials Sol vents In most cases, the solvents used were reagent analytical grade and were not further purified. Those solvents used in UV-visible spectrometry or conductance were measured before use to ensure that they would not interfere with sample measurements. Acetonitrile used in the polymer catalysis reactions was distilled from C a H „ and stored over molecular sieves to reduce the moisture content. Glymes, PEO, and PVP Dimethoxyethane , b i s ( 2-methoxyethyl ) ether or diglyme, and b i s [2( 2-methoxye thoxy )e thyl ]e ther or tetraglyme were all commercially available and were distilled under vacuum prior to use. The low molecular weight po 1 y ( e thy 1 ene oxide) samples including PEO 600, PEO 1000, PEO 4000, and PEO 6000 were obtained from Union Carbide under the trade name Carbowax. The higher 108

PAGE 121

09 molecular weight po ly ( ethyl ene oxide) samples including PEO 10 5 , PEO 3 X 10 5 , PEO 9 X 10 5 , and PEO 6 X 10 6 were also available from Union Carbide under the name Polyox. All of the poly (ethyl ene oxide) samples were used without further purification. Pol y ( v i nyl pyrro 1 i done ) , molecular weight 40,000, was obtained from Poly sc i ences , Inc. Vinyl Glyme The vinyl glyme 3 , 6 , 9 , 1 2 , 1 5pen taoxa 1 -hepta decene was obtained from Aldrich. Approximately 100 ml of monomer were polymer ized by adding a small drop of distilled BF 3 -0Et 2 to the monomer in a 250 ml round-bottom flask. The system was kept under nitrogen and periodically shaken for approximately one week. The resulting polymer was soluble in virtually all common solvents. The sample was placed under vacuum (0.05 mm) at approximately 80°C to remove any unreacted monomer. The absence of monomer was confirmed by IR analysis. A slight yellow color was present which was removed by dissolving the polymer in CH C 1 3 and passing this solution through a column of 1:1 diatomaceous earth and charcoal. The molecular weight was measured on a Mechrolab Model 502 High Speed Membrane Osmometer and was found to be 40,000. Crown Ethers Dicyclohexyl-18-crown-6 was obtained from Dr. H. K. Frensdorff of E. I. duPont de Nemours Elastomers

PAGE 122

Department and rec ry s ta 1 1 i zed from petroleum ether. The nonsubstituted crown ether, 18-crown-6, was purchased from PCR, Inc. Purification was accomplished according to the method of Gokel et al . Poly ( v i ny 1 benzo1 5-crown-5 ) was obtained from Dr. J. Smid, SUNY, Syracuse, and was used without further purification. Antibiotic Compounds The antibiotic tyrocidine was obtained in the HC1 form from ICN Pharmaceuticals, Inc. as a mixture of at least five separate compounds . These components varied in the amino acids present at three of the ten positions of the cylic peptide. None of the available separation techniques was successful in separating these components. As a result, the antibiotic was used as the mixture. Poly(acryl oyl tyrocid i ne) was synthesized by the following reaction: -[CH 2 — CH}+ Tyroc — NH. C = ®: \ {CH 2 CH3c=o HN-Tyroc where Tyr-NH„ is the ornithine amine group

PAGE 123

This synthesis first required the synthesis of poly(acryl oyl benzotr i azol e ) . Practical grade benzotriazole was recrystallized by melting 95 g of crude material over a hot flame and adding the molten material into 3 00 ml of benzene. The solution was stirred until crystallization began. The solution was chilled for two hours and filtered. A solution of 48 g benzotr i azol e , 60 ml of dry tri ethyl ami ne , and 600 ml of dry benzene was added dropwise to a well-stirred solution of 32 ml acryloyl chloride in 200 ml of dry toluene. The reaction mixture was kept between -5° and 0°C by external cooling with a dry iceisopropanol bath. The reaction mixture was continuously stirred while it was allowed to warm to room temperature over a two-hour period. The solution was filtered, and the filtrate was extracted with cold water and with a saturated solution of sodium chloride. The organic layer was dried over anhydrous Na ? S0. and evaporated to dryness under vacuum after adding 0.5 g of t-butyl catechol. The temperature was never allowed to rise above 30°C during evaporation. The residue was a yellow, flaky material. Some losses, due to polymerization, were often encountered during recrystallizations, and the preferred procedure was to treat the crude product with n-heptane preheated to 80°C, filter, and cool the filtrate rapidly in a dry ice-isopropanol bath. The precipitate was 14 g

PAGE 124

1 2 of a white, powdery solid, m.p. 65-67°C. The literature value was 67-68°C. 78 Poly( acryloyl benzotriazol e) was made by placing 10 g of monomer in 30 ml of dry benzene into a 50 ml round-bottom flask, adding 50 mg of azob i s i sobu tyron i trile, and purging with nitrogen. The flask was then kept at 65-70°C for approximately 10 hours. The polymer precipitated, was collected, and dissolved in methylene chloride. The polymer was then reprec i pi ta ted in an excess of anhydrous ether and dried under vacuum. The yield was 23%. The IR spectrum of po 1 y ( aery 1 oy 1 benzotriazole) is given in Figure 25. It matched the infrared spectrum given by Ferruti e_t a 1 . The attachment of the antibiotic tyrocidine to poly( acryl oyl benzotri azol e ) presented possible crosslinking problems due to two reactive sites on the antibiotic, the ornithine primary amine and the tyrosine phenolic group. The tyrosine function was acylated by dissolving ty roc i d i ne • HC 1 in an equal volume mixture of acetic acid and acetic anhydride and stirring for four hours. The acetic acid and unreacted acetic anhydride were then removed by vacuum distillation. The product was a light brown, powdery solid. The HC1 on the ornithine amino acid blocked this position. The HC1 could then be removed with Et 3 N in order to provide a nucleop h i 1 i c site.

PAGE 125

113 1

PAGE 126

114 Po ly ( aery 1 oy 1 tyroc i d i ne ) was prepared by adding 2.0 g poly(acryloylbenzotriazole) to 0.4 q of acylated tyrocidine and 3.0 ml tri ethyl ami ne in 35 ml of chloroform. The mixture was stirred for two days at 60°C. At this time 8 ml of propylamine was added to cleave unreacted benzotriazole units in order to prevent hydrolysis and possible polyelectrolyte formation. The mixture was stirred for an additional day at 60°C. Chloroform and tri ethyl ami ne were removed by rotary evaporation. The remaining solid was added to a small amount of chloroform and filtered to remove unreacted tyrocidine and the byproduct E t ^ N • H C 1 . The polymer was then precipitated in ether. The product was 0.6 g of flaky, light brown material. The UV spectrum indicated a tyrocidine absorption at 280 nm with no indication of benzotriazole at 251 nm (Figure 26). Assuming that the components of the antibiotic reacted in the same percentage in which they were present in the initial mixture, the absorption of the polymer was compared to that of tyrocidine itself and it was determined that the product contained one antibiotic molecule for every 12 monomer units. Salts Most of the salts studied, such as NaCl, KC1, KBr, NaBPh., and NaOAc, were all available in analytica

PAGE 127

1 5 28 r x> < 220 250 280 310 A(nm) 340 220 250 280 310 A(nm) 340 Figure 26. Absorption of (A) Tyrocidine-HCl ; (B) Poly(acryloyl tyrocidine) in Methanol

PAGE 128

reagent grade. They were dried overnight at 0.5 mm prior to use . KBPh^ was made by treating a dilute aqueous solution of sodium tetrapheny 1 bor i de with a dilute aqueous solution of KC1. The insoluble KBPh. precipitated, was filtered, washed, and rec rysta 1 1 i zed from 1:3 water79 acetone. Bu^NBPh. was made in a similar manner by adding a Z% aqueous solution of NaBPh. to a 2% aqueous solution of B u 4 N I . The precipitate was washed with water after filtering, recrys ta 1 1 i zed from a 1:3 water-acetone mixture and dried for three days under vacuum at room temperature. The pi crate salts of Na, K, and Cs ions were all prepared by neutralizing picric acid with the appropriate alkali hydroxide in ethanol. The products were recrystal 1 i zed from absolute ethanol and dried at room temperature for two days at 0.01 mm. Cation Binding Measurements V i scome try The polymer solution was made by adding PEO 6 X 10 to methanol to form 100 ml of a saturated solution. This solution was allowed to stand for two days, was decanted, centrifuged, decanted again and diluted to 1 liter. This method was used because vigorous stirring

PAGE 129

1 7 of the polymer solutions would lead to mechanical degradation. The concentration was determined by evaporating the solvent from 10 ml samples and weighing the remaining polymer. The concentration was found to be 1.10 g/1. Appropriate quantities of salts were added to the polymer prior to viscosity determinations. The viscosity measurements were made with a Cannon Viscometer Number 25-K632 in a constant temperature water bath at 25.0°C. Distribution Equilibria Distribution equilibria studies involving the extraction of alkali picrate salts from an aqueous layer into an organic layer containing po 1 y ( ac ryl oy 1 tyroc i d i ne ) were followed by UV-visible spectroscopy due to the absorption of the picrate anion at % 360 nm in CHC1-. The reported extinction coefficient of this absorption is 1.65 X 10 cm M . The spectra were taken on a Beckman ACTA V Spectrophotometer. The aqueous layers consisted of 1.0 X 10 M picrate salt and 5.0 X 10 M of the corresponding alkali chloride. The presence of the chloride served to increase the ionic strength of the aqueous layer to prevent intermixing of the aqueous and chloroform layers and to increase the transfer of metal salt into the organic layer to increase compl exat ion . The chloroform layer contained

PAGE 130

1 18 sufficient poly ( acryToyl tyrocidi ne ) for the solution to have an estimated tyrocidine concentration of 1.3 X 10" 3 M. Twenty ml of each solution were placed in a stoppered 50 ml Erlenmeyer flask and shaken periodically for approximately one day. The pi crate absorption in the CH C 1 3 layer was then measured and compared to a blank sample which contained no polymer. The concentration of salt extracted was calculated using the reported extinction coefficient. Po tent i ome t ry Potentiometry involving the use of ion-selective electrodes was used to follow the process of cation complexation by PEO. The electrodes used were Corning Monovalent Cation Electrode (Catalog Number 476220) for potassium ions, Corning Sodium Ion Electrode (Catalog Number 476210) for sodium ions, and an Ag/AgCl reference electrode made by Dr. M. Mohan from this department. This reference electrode necessitated the use of alkali chloride salts due to possible electrode contamination with other anions. The potential was measured with a Corning Digital 112 Research pH meter. In order to measure solution potentials in methanol, it was necessary to condition the glass electrodes stepwise in aqueous methanol solutions containing greater percentages of methanol until the electrode was finally in pure methanol.

PAGE 131

A closed glass cell with two openings for the elec trodes, which were sealed by serum caps, was used for the measurements. The 25 ml solutions were stirred by a small magnetic stirring bar after addition of solid ligand until the ligand had dissolved. Stirring was then discontinued, and readings were taken at one minute intervals until the readings differed by no more than 0.2 mV. Calibration curves of [MCI] vs. emf were measured for [MCI] from 1.0 X 10" 5 to 1.0 X 10" 2 . This gave the potential for free ions in solution. Then when PEO was added to a salt solution, the emf value was compared to the calibration curve to obtain the concentration of free ions remaining in solution. This procedure was checked for NaCl plus d i eye 1 ohexy 1 1 8-c rown -6 and 18-crown-6 and was found to reproduce literature values. UV-visible Spectrometry The change in the absorption spectra of sodium pi crate in 10% THF/90% CHCK upon addition of polyethers was measured by the Beckman ACTA V Spectrometer. The addition of THF was necessary in order to dissolve sufficient concentration of sodium picrate in C H C 1 ., but this had no effect on the absorption spectra. The change in A of sodium picrate with respect ma x to tetraglyme concentration was measured by titrating 100 -4 ? ml of 1.0 X 10 M sodium picrate with 1.0 X 10 M

PAGE 132

120 tetraglyme plus 1.0 X 10" M sodium picrate in 10% THF/ 90% CHC1 . This kept the picrate concentration constant while increasing the tetraglyme concentration. The absorption was measured after each tetraglyme addition. -4 The absorption spectra of 1.0 X 10 M sodium picrate in the presence of 1.0 X 10" 4 M and 1.0 X 10~ 3 M polyethers were measured as follows. A 1.0 liter stock solution of 1.0 X 10 M sodium picrate was prepared. _ 3 From this solution, 50 ml solutions of 1.0 X 10 M 4 polyether plus 1.0 X 10 M sodium picrate were prepared. Diluting these solutions by a factor of 10 with the -4 1.0 X 10 M stock solution of sodium picrate gave solutions of 1.0 X 10~ 4 M sodium picrate plus 1.0 X 10" 4 M polyether . Conductance The conductivity experiments were carried out by adding various polyethers to 15 ml of 1.0 X 10 M tetraphenylboride salts in acetone, tetrahydrofuran, and acetonitrile at 25°C. The cell constant for the particular cell used was 0.1159. For the low concentrations of -4 -3 polyether, approximately 4.0 X 10 M to 5 X 10 M in monomer, the additions were made by titration in which the titrant contained the polyether and the same concentration of tetraphenylboride salt that was being titrated. Therefore, the solution in the conductance cell always

PAGE 133

21 contained the same concentration of tetra phenyl bori de salt but varied in polyether concentration. The higher polyether concentrations, 1.1 X 10 M to 1.0 X 10 _1 M in monomer, involved the direct addition of solid polymer to the NaBPh. solution. After each addition of polyether, the conductance was determined by a General Radio 1673-A Automatic Capacitance Bridge operating at 1 KHz. This was connected to a General Radio 1672-A Digital Control Unit also operating at 1 KHz . React ions Reaction Methods The catalytic reactions were all carried out in 5 ml round-bottom flasks with a side arm stoppered with a septum cap. The side arm allowed withdrawal of small amounts of solution for gas chromatographic analysis without interrupting the reaction. The flasks were connected to condensers and drying tubes. These systems were grouped in series of five, supported by a metal rack and immersed in an oil bath. The oil baths were maintained at % 100°C to ensure sufficient reflux i n g of the acetonitrile solutions. For each reaction, 0.02 g of alkali acetate salt, the polyether, and 2.5 ml acetonitrile were added to the

PAGE 134

122 round-bottom flask and refluxed for 30 minutes. This was followed by the addition of 10 ul of butyl bromide with a syringe. For homogeneous reactions, the salt was added first followed by 0.13 ml H„0 and the flask was then swirled to dissolve the salt. The sequence was continued as above. On many of the homogeneous reactions, 10 pi of ch 1 orobenzene were added with the butyl bromide as an internal standard because of interference of water with the butyl bromide peak in the GC analysis. The reactions were allowed to proceed for time intervals varying from five hours to five days. Product Analys i s Product analysis was made on a Hewlett Packard 700 Laboratory Chroma tog ra ph . The column was packed with 15% UC0N 50 LB 550 X on Chromosorb G. The instrument conditions were attenuation, 1; column temperature, 90 °C; injection port, 160°; detector, 182°; helium pressure, -2 20 lbs in . The injection port was packed with glass wool in order to prevent polymers from the solution from becoming pyrolyzed in the end of the column. For analysis of the reaction mixtures, 20-30 ;> 1 of solution were injected onto the column. Peak areas were measured by a K and E 62 0000 Compensating Polar Plan i meter.

PAGE 135

23 The instrument was calibrated by injecting equal molar quantities of butyl bromide and butyl acetate into the column. The relative size of the peaks gave a correction which, when applied to acetate peaks on the product traces, gave the relative amounts of butyl bromide and butyl acetate present in the mixture. Division of the size of the corrected acetate peak by the total of the two peaks gave the percent yield. For the homogeneous system, the c hi orobenzene peak was used to determine how large the bromide peak would have been had it not been obscured by water.

PAGE 136

REFERENCES 10 1 1 , 12. 13. 14. L. L. Chan and J. Smid, J. Amer. Chem. Soc, 89, 4547 (1967). — L. L. Chan, K. H. Wong and J. Smid, J. Amer. Chem. Soc. , 92, 1955 (1970) . J. Smid and A. M. Grotens, J. Phys. Chem., 77, 2377 (1973). — C. J. Pedersen, J. Amer. Chem. Soc, 89, 7017 (1967). C. J. Pedersen, Fed. Proc, Fed. Amer. Soc., Exp. Biol . , 27, 1305 (1968) . H. K. Frensdorff, J. Amer. Chem. Soc, 93_, 600 (1971). H. K. Frensdorff, J. Amer. Chem. Soc, 9^_, 4684 (1971) J. J. Christensen, D. J. Eatough and R. M . Izatt, Chem. Rev. , 7_4, 351 ( 1974) . a) J. N. Roitman and D. J. Cram, J. Amer. Chem. Soc, _9_3, 2231 (1971 ) . b) J. Zavada, M. Svoboda and M. Pankova, Tetrahedron Lett. , 71 1 (1972) . U. Takaki, T. E. Hogen Esch and d. Smid, J. Amer. Chem. Soc. , 93 , 6760 ( 1971 ) . B. Dietrich, J. M. Lehn and J. P. Sauvage, Tetrahedron Lett. , 2889 (1969). B. Dietrich, J. M. Lehn and J. P. Sauvage, J. Chem. Soc . Chem. Comm. , 440 (1971). B. Dietrich, J. M. Lehn and J. P. Sauvage, Tetrahedron, 29, 1647 (1973). J. M. Lehn, Structure and Bonding, J_6, 1 (1973). 1 24

PAGE 137

121 15. B. C. Pressman, Ant imi cr.ob. Agents Chemother., 28 (1969). 16. B. C. Pressman, Fed. Proc., Fed. Amer. Soc. Exp. Biol., 32, 1698 (1973) . 17. K. H. Wong, K. Yagi and J. Smid, J. Membrane Biol., 1_8, 379 (1974). 18. H. Lardy, Fed. Proc, Fed. Amer. Soc. Exp. Biol., 27, 1278 (1968). 19. B. C. Pressman, Fed. Proc., Fed. Amer. Soc. Exp. Biol., 27_, 1283 (1968). 20. P. Mueller and D. 0. Rudin, Biochem. Biophys. Res. Comm. , 26, 398 (1967) . 21. D. C. Tosteson, Fed. Proc., Fed. Amer. Soc. Exp. Biol . , 27_, 1 269 (1968) . 22. II. Degani and H. L. Friedman, Biochemistry, 13, 5022 (1974) . 23. M. B. Feinstein and H. Felsenfeld, Proc. Nat. Acad. Sci . , U.S.A. , 68, 2037 (1971 ) . 24. G. Cornelius, W. Gartner and D. H. Haynes, Biochemistry, J_3, 3052 (1974). 25. J. J. Christensen, J. 0. Hill and R. M. Izatt, Science, 174 , 459 (1971 ). 26. C. M. Starks, J. Amer. Chem. Soc, 9_3, 195 (1971). 27. D. Landini, F.'Montanari and F. Pirisi, J. Chem. Soc. Chem. Comm. , 879 ( 1 975) . 28. C. L. Liotta and E. E. Grisdale, Tetrahedon Lett., 4205 (1975). 29. C. L. Liotta and F. L. Cook, J. Org. Chem., 39, 3416 ( 1974) . 30. C. L. Liotta and H. P. Harris, J. Amer. Chem. Soc., 96, 2250 ( 1974) . 31. C. L. Liotta, H. P. Harris, M. McDermott, R. Gonzalez and K. Smith, Tetrahedron Lett., 2417 (1974). 32. J. W. Zubrick, B. I. Dunbar and H. D. Durst, Tetrahedron Lett. , 71 (1975) .

PAGE 138

2 6 33. D. J. Sam and H. E. Simmons, J. Amer. Chem. Soc., 9_4 4024 (1972). 34. M. Makosza and M. Ludwikow, Angew. Chem. Int. Ed., U, 665 (1974). 35. G. W. Gokel and H. D. Durst, Aldrichimica Acta, 9, 3 (1976). 36. Herbert Morawetz, Macr o molecules in Solution , Interscience Div., John Wiley and Sons, New York, 1965, Chapter VII . 37. R. D. Lundberg, F. E. Bailey and R. W. Callard, J. Polym. Sci . , Al , 4, 1563 (1966) . 38. Kang-Jen Liu, Macromol ecu 1 es , 1_, 308 (1968). 39. I. M. Panayotov, C. B. Tsvetanov and D. K. Dimov, Makromol. Chem., V7_L> 279 (1976). 40. S. Kopolow, T. E. Hogen Esch and J. Smid, Macromolecules, 4, 359 (1971)'. 41. J. Smid, S. Shah, L. Wong and J. Hurley, J. Amer. Chem. Soc. , 97, 5932 (1975) . 42. S. Kopolow, T. E. Hogen Esch and J. Smid, Macromolecules, 6, 1 33 (1973) . 43. S. Kopolow, Z. Machacek, U. Takaki and J. Smid, J. Macromol . Sci . , A7 , 1015 (1973). 44. K. H. Wong, K. Yagi and J. Smid, J. Membrane Biol., J_8, 379 (1974). 45. N. Yamazaki, A. Hirao and S. Nakahama, Polymer J., 7, 402 (1975). 46. B. T. Kilbourn, J. D. Dunitz, L. Pioda and W. Simon, J. Mol . Biol . , J_8, 379 (1974) . 47. J. F. Blount and J. W. Westley, J. Chem. Soc. Chem. Comm. , 927 ( 1971 ) . 48. E. C Bissell and I. C. Paul, J. Chem. Soc. Chem. Comm. , 967 ( 1972) . 49. M. C. Goodall, Biochim. Biophys. Acta, 2J_9, 471 (1970).

PAGE 139

27 50. A. T. Tsatsas, R. W. Stearns and W. M. Risen, Jr., J. Amer. Chem. Soc, 9_4, 5247 (1972). 51. K. H. Wong, G. Konizer and J. Smid, J. Amer. Chem. Soc. , 9_2, 666 (1970) . 52. E . Shchori, J. Jagur-Grodzinski and M . Shporer, J. Amer. Chem. Soc, 9_5, 3842 (1973). 53. R. M. Izatt, D. P. Nelson, J. H. Rytting, B. L. Haymore and J. J. Christensen, J. Amer. Chem. Soc., 9_3, 1619 (1971 ). 54. E. M. Arnett and T. C. Moriarity, J. Amer. Chem. Soc. , 93_, 4908 (1971 ) . 55. M . Bourgoin, K. H. Wong, J. Y. Hui and J. Smid, J. Amer. Chem. Soc, 9_7, 3462 (1975). 56. J. Smid, Angew. Chem. Int. Ed., 1J_, 112 (1972). 57. D. E. Evans, S. L. Wellington, J. A. Nadis and E. L. Cussler, J. Sol. Chem., 1_, 499 (1972). 58. T. E. Hogen Esch, unpublished results. 59. C. J. Pedersen and H. K . Frensdorff, Angew Chem. Int. Ed. , 11, 16 (1972). 60. R. Barker, Organic C h e in i s t ry of Biological Compounds , Prentice-Hall, Inc., Lnglewood Cliffs, New Jersey, 1971 . 61. G. Eisenman, Advan. Anal. Chem. Instr., 1 , 213 (1965). 62. R. Foster, Organic Charge Transfer Complexes , Academic Press, New York, N . Y., 1969, Chapter 6. 63. A. K . Covington and P. Jones, Hyd rogen Bonded Solvent System s , Taylor and Francis, Ltd., London, 1968, Section III. 65. M . S z w a r c , C ar b anions, Livi n g Polymers, and Electron Trans fer Pr oce sses, Wiley-Interscience, New York, N. Y. , 1968, p. 264. V. Gutmann, Coordination Chemi s try in Non-Aqueous Solutions , Springer-Verlag, New York, 1968.

PAGE 140

28 66. A. J. Gordon and R. A. Ford, The Chemi s t's Companion John Wiley and Sons, Inc., New York, 1972. 67. R. L. Kay, B. J. Hales and G. P. Cunningham, J. Phys. Chem. , 7_1_, 3925 (1967) . 68. M . S z w a r c , Ions and Ion Pairs in Organic Reactions , Vol. 1, Wiley -Inter science, New York, 1972. 69. E. G. McRae and M. Kasha, J. Chem. Phys., 28, 721 (1958). 70. F. Daniels and R. Alberty, Physical Chemistry , John Wiley and Sons, Inc., New York, 1967, p. 387. 71. Herbert Morawetz, Macromol ec ul es in Solution , Interscience Div., John Wiley and Sons, New York, 1975, Chapter VI . 72. U. P. Strauss and Y. P. Leung, J. Amer. Chem. Soc., 87_, 1476 (1965) . 73. K. H. Wong, K. Konizer and J. Smid, J. Amer. Chem. Soc . , 92, 666 (1970) . 74. T. E. Hogen Esch and J. Smid, J. Amer. Chem. Soc., 9J_, 4580 (1969). 75. E. R. Thornton, Solvolysis Mechanisms , Ronald Press, New York, 1964, Chapter 4. 76. G. W. Gokel, D. J. Cram, C. L. Liotta, H. P. Harris and F. L. Cook, J. Org. Chem., 39, 2445 (1974). 77. Organ ic Syntheses , Collective Volume 3, John Wiley and Sons, New York, 1955, p. 106. 78. P. Ferruti, A. Fere" and G. Cottica, J. Polym. Sci., Polym. Chem. Ed., j_2, 553 (1974). 79. D. N. Bhattacharyya , C. L. Lee, J. Smid, and M. Szwarc, J. Phys. Chem., 69, 608 (1965). 80. F. Accascina, S. Petrucci and R. M. Fuoss, J. Amer. Chem. Soc. , 81 , 1301 ( 1959) .

PAGE 141

BIOGRAPHICAL SKETCH Ronald E. Cambron was born July 6, 1948 in Jackson, Tennessee. . After graduation from Jackson High School he entered the University of Tennessee at Knoxville, where he received a Bachelor of Science degree in chemistry in 1971 . While at the University of Tennessee, Mr. Cambron was enrolled in the Cooperative Education program with assignment to the Tennessee Valley Authority in Muscle Shoals, Alabama. He also served as an analytical chemist for the Water Qualify Control Branch of the Tennessee Public Health Department for two summers. During his junior year, he received the Analytical Chemistry Award from the Department of Chemistry. In September 1971, he entered the Graduate School of the University of Florida. During his first year, he was the recipient of a graduate school fellowship. Mr. Cambron is married to Dr. Nelda H. Cambron, who is currently an assistant professor in Educational Administration at the University of Florida. 129

PAGE 142

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. / ±£t u Thieo E. Hogen Esch, Chairman Associate Professor of Chem i stry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Roger G \ Bates Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. n/i.KutLs George B . Butler Professor of Chemistry

PAGE 143

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Edmond — cHr" (j a b b a y Professor of Chemistjry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. IV ^ k.v L Dinesh 0. Shah Professor of Chemical Engineering This dissertation was submitted to the Graduate Faculty of the Department of Chemistry in the. College of Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December , 1976 Dean, Graduate School

PAGE 144

UNIVERSITY OF FLORIDA 3 1262 08553 3114