¢ - aw 7‘ A u I
. V“. 1 w. t
- n u .u - It .. u 1 v
u ‘ v1
- .- : - z- I .. u. 9...
n V: o
D ‘ - 14' ' F
H.- 2. ‘.5 «I '4. Jo-o. hr. 1‘1 I. §
- ‘ u . It
a . A I u o .
- l. I '1 v 'I
.. .. _... .z .3 .
'- -.. .
JIV

.Ul

. suwtot I ‘
:4 ..l .
.1. v1.
4 '0 g‘ .54
45: .
$7.. .....

0:.
.54.:-

r‘. .
LI 0

o
. (It ( I .. 037‘»...
v I; I l. . I I-.. 4 or .
. . n V a I v: Q n 5 1| 9. : I. 33..
- I J as 1 a l .0 4: . x I‘luh’ll‘yuo‘
.o -- +1 .u v. : .Ifl; .. 1‘ I .11, .3:
.n . . . K - 131:. c -11.. ' a: 9
. t .n. 15-4.1.1“?! 1. . L. .1...
. h. .r... Infinawllu ( .2 1..-
'2' 4ng: Hurli . D.
.fu.urlu..wfi.fl.~mx :0- I. P; vb...uk.f‘....uu.
. I 2’1"..." 4
. . Alzhiax I.
, 1'. 0
N. (If. . .v
fill I

It!
{ ID
I 1

.. ill.
R4

. .1
M 11',t4f .

. . . y . If. m. i. («v
‘ l. .. ‘3'; I.
. .. 1.1.31.1: \ .tho A

l ‘ - '
www.mwdufluu 1 .; a. , ‘

.0 D‘JD ‘I’ a it! IJ“-C"IMfilfl

. . A! . ‘ ‘ .5.
y

.
.f’.
|I. If} 7 In
It , a...’
.

b m5
U" .

137‘“ '

l. ‘4“

 

  
    

L [B R A I{ Y .‘
Michigan State j
Universi _. f
V, 5

T‘th...

6“,

This is to certify that the

thesis entitled

The Molar Free Energies of Transfer
from Water to Several Nonaqueous
Solvents for DBlBCG and BlSCS and

Proton and Carbon-l3 NMR Studies of
Some 18C6 ConPl fiffiy

pre en e
Davette Jones Whitaker

has been accepted towards fulfillment
of the requirements for

Masters degree in Chemistry

Major professor

Alexander I. Popov
Date November 28, 1979

0-7639

 

claw:
25¢ per day per item
RETURNING LIBRARY MATERIALS:

Place in book return to remove
charge from circulation records

 

 

 

THE MOLAR FREE ENERGIES OF TRANSFER FROM WATER
TO SEVERAL NONAQUEOUS SOLVENTS FOR DB18C6
AND BISCS AND PROTON AND CARBON-l3 NMR

STUDIES OF SOME 18C6 COMPLEXES

By

Davette Jones Whitaker

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

MASTER OF SCIENCE

Department of Chemistry

1979

ABSTRACT

THE MOLAR FREE ENERGIES OF TRANSFER FROM WATER
TO SEVERAL NONAQUEOUS SOLVENTS FOR DBl8C6
AND BISCS AND PROTON AND CARBON—l3 NMR

STUDIES OF SOME 18C6 COMPLEXES
By
Davette Jones Whitaker

The molar free energies of transfer, AGE, from water
to several nonaqueous solvents for DB18C6 and B1505 were
determined by solubility measurements. AG: of D818C6 for
methanol, acetonitrile, and 1,2-Dichloroethane are
-2.1 a 0.5, —u.3 i 0.5, and —u.8 i 0.5 Kcal moi'l,
respectively and —l.2 i 0.2 Kcal mol"l for AGE of 81505
for methanol. These values are explained in terms of the
solvating ability of the solvent and the solvent struc-
ture.

Proton and carbon—13 NMR studies of several alkali
metal cations and tetraethylammonium cation complexes

of 1806 were performed. Determination of the formation

constants for these complexes was attempted. The results

Davette Jones Whitaker

indicate the insensitivity of proton and carbon-l3 NMR
as probes in the study of complexation reactions of non-

benzo—substituted crown ethers such as 1806 at low field

strengths.

To Clayton and Nicole

11

ACKNOWLEDGMENTS

The author would like to thank Professor Alexander I.
Popov for his patience, guidance, and encouraging words.
Thank you Professor Popov.

The help, friendship, and laughter of my brothers and
sisters of the group has been deeply appreciated. You
will all be missed.

The author acknowledges the financial assistance of
the Department of Chemistry, Michigan State University,
and the National Science Foundation during the course of
this study.

Without the love and constant moral support of my
family this study would not be. Therefore, deep and
humble appreciation is extended to my Grandmother, brother,
and especially to my sister who has indeed been a friend
to a friend in need. Finally, I thank my children, Clayton
and Nicole for their unquestioning love, to them I dedicate

this thesis.

TABLE OF CONTENTS

Chapter

LIST OF TABLES. . . . . .

LIST OF FIGURES .

LIST OF ABBREVIATIONS

CHAPTER 1. HISTORICAL REVIEW .
A. Introduction.
B. Macrocyclic Polyethers.

1. Introduction .
2. Properties of Polyether Crowns

Solubility of Macrocyclic
Ethers . . . . .

Heats of Solution.

Conformations of Crown
Ethers . . . . .

Basicity of Crown
Polyethers .

Complex Formation.
Toxicity of Cyclic
Polyethers
0. Transfer Activity Coefficients (Free
Energies of Transfer) . . . . .
1. Introduction

2. Determination of Transfer
Activity Coefficients.

Solubility Studies

Determination by Measure-

ment of the Standard Potentials
of a Galvanic Cell Reversible
to its Ions.

iv

Page

vi

xi

m>~4 -q

15
l6

19

21
22

23

26
26

29
31

3A

Chapter

CHAPTER

CHAPTER

A.

C.
LIST OF

Vapor Pressure Measure-
ments.

3. Ligand—Solvent Interactions.
Proton and Carbon-l3 NMR Study of

1806 with Some Potassium and
Rubidium Salts. . . . .

1. Introduction

2. Proton and Carbon-l3 NMR of
Polyether Complexes of Alkali
and Alkaline Earth Metals.

2. EXPERIMENTAL MATERIALS
AND METHODS . .

Materials
1. Ligands. . . . . . . .
2. Salts.

3. Solvents

Methods . . . . . . . . .
l. Spectroscopy
2. Other Analyses
3. NMR Data Handling.
A. Solubility Measurements.
5. Validity of Method

3. RESULTS AND DISCUSSION.

Molar Free Energies of Transfer From
Water to Several Nonaqueous Solvents
for DB1806 and B1505. . . . . .

Proton and Carbon-l3 NMR of 1806
Complexes of Some Alkali Metal and
Tetraethylammonium Salts. .

l. Proton NMR
2. Carbon-13 NMR.

Future Work
REFERENCES.

Page

35
36

AA
AA

A6

52
53

53
53
5A
60

6O

61
62
62
6A

68

69

77

77
80

101
103

Table

LIST OF TABLES

Page

Molecular Weights and Melting

Points of Some Polyether

Crowns . . . . . . . . . . . . . . . . . ll
Diameters of Cavities in

Some Cyclic Polyethers . . . . . . . . . 12
UV Spectrometric Characteris-

tics of Some Polyether

Crowns . . . . . . . . . . . . . . . . . 1A
Solubilities of Some

Crown Ethers . . . . . . . . . . . . . . l7
Ionic Diameters of Alkali

Metal Cations. . . . . . . . . . . . . . 2A
Properties of Solvents Used in

the Study of the Free Energy

of Transfer of Some Polyether

Crowns . . . . . . . . . . . . . . . . . 58
Pertinent Properties of Solvents

Used in the Proton and Carbon-l3

NMR Study of Potassium and

Rubidium Complexes of 1806 and

Magnetic Susceptibility Cor-

rection Factors. . . . . . . . . . . . . 66

vi

Table Page

8 Solubility of Benzoic Acid as

Determined by UV Spectroscopy,

Titration, and Gravimetric

Methods. . . . . . . . . . . . . . . . 66

9 Ultraviolet Characteristics

and Solubilities of DBl8C6 and

B1505 in Various Solvents at

25 i O.5°C . . . . . . . . . . . . . . 7O
10 Transfer Activity Coefficients

and Free Energies of Transfer

of DB1806 and B1505 from

Water to Various Solvents. . . . . . . 71
11 Comparison of Solvent Properties

and Molar Free Energies of

'Transfer, AGE from Water to

the Various Solvents . . . . . . . . . 73
12 Comparison of Solubilities

and ACE (H20 + Solvents) of

DBl8C6 with the e/u Ratio. . . . . . . 7A
13 Proton nmr Chemical Shift-Mole

Ratio Data for 1806 Complexation

Studies (018c6=0.05 M; 60 Hz=l ppm). . 78
1A Proton NMR Chemical Shift Data

Obtained at 180 MHz for 1806 and

18C6-K+ in MeOH and DMSO . . . . . . . 79

vii

Table Page

15 Carbon-l3 NMR Chemical Shift-

Mole Ratio Data for 1806 Com-

plexation Studies. . . . . .'. . . . 81
16 1806 Concentration Study Data

in DMSO Obtained by Carbon-13

NMR. . . . . . . . . . . . . . . . . BA
17 Carbon-l3 NMR Limiting Chemical

Shifts and Log K of 1806-K+ and

f

1806'Rb+ Complexes in Various

Solvents . . . . . . . . . . . . . . 87
18 Comparison of Carbon-l3 NMR

Limiting Chemical Shifts of

Various Salt Complexes of

1806 in DMSO with Ionic Diameter

of the Cations . . . . . . . . . . . 93
19 Comparison of Carbon-l3 NMR

Limiting Chemical Shifts of

18C6°K+ Complexes and Solvent

Properties . . . . . . . . . . . . . 96
20 Comparison of Carbon—l3 NMR

Limiting Chemical Shifts of

the 1806-Rb+ Complex and

Solvent Properties . . . . . . . . . 97
21 Formation Constants of

1806°K+ and 18C6-Rb+ Complexes

in Various Solvents. . . . . . . . . 98
viii

Table

22

Page

Comparison of Formation

Constants Determined from
13c and 39K NMR for the

1806°K+ Complex in Various

Solvents . . . . . . . . . . . . . . 99

ix

Figure

LIST OF FIGURES

Synthetic macrocyclic polyether
ligands.

Carbon-13 chemical shifts XE
K+/1806 mole ratio in various
solvents

Carbon-l3 chemical shifts Xi
Rb+/1806 mole ratio in various
solvents . . . .

Carbon-l3 chemical shifts XE

cation/18C6 mole ratio in DMSO

Page

88

89

95

DB1806
B1505
1806
MeOH

An
1,2-DOE
Ac

DMSO
DMF

TMS

DC18C6

LIST OF ABBREVIATIONS

Dibenzo—l8-crown—6
Benzo-lS-crown-5
18-crown-6
methanol
acetonitrile
1,2-Dichloroethane
acetone
dimethylsulfoxide
dimethylformamide

tetramethylsilane

Dicyclo-lB-crown-6

xi

CHAPTER 1

HISTORICAL REVIEW

A. INTRODUCTION

 

An important and interesting group of macrocyclic
compounds is known as crown ethers. The crown ethers
were first synthesized and reported by Charles Pedersen
in 1967 (l). Pedersen used a trivial nomenclature for
these compounds because of their long and complicated IUPAC
names. Shown in Figure l are some examples of polyether
crowns which are identified by their IUPAC and trivial
names. The IUPAC names are shortened by using the general
scheme of naming the groups attached to the crown ether
ring followed by the word "crown" and then the number of
oxygen atoms in the crown ether (1,2). According to this
scheme ligand I of Figure 1, whose chemical name is
2,3,11,12-dibenzo-1,A,7,lO,13,l6-hexaoxacyclooctadiene,
can be called dibenzo—lB-crown-6, abbreviated in this
study to DB1806.

As seen in Figure l, the crown ethers are uncharged
molecules which contain holes of varying size depending
on the number of atoms in the ring. The ring consists
of donor oxygen atoms usually separated from each other
by two methylene groups, —CH2-CH2-. Pedersen observed
that crown ethers displayed complexing properties with
some metal cations and especially the alkali and alkaline

earth metal cations, by encapsulizing them inside the

 

 

IUPAC name: 2,3,11,12-
Dibenzo-l,u,7,10,l3,16—
hexaoxacyclooctadeca-2,ll-
diene

Crown name: Dibenzo—18-

crown-6 (DB1806)

IUPAC name: 1,“,7,10,l3,l6-
hexaoxacyclooctadecane

Crown name: 18—crown-6 (18C6)

IUPAC name: 2,3-Benzo-1,A,-
7,lO,l3-oentaoxacyclopenta-
deca—2-ene

Crown name: Benzo—lS-crown-

5 (815C5)

a=b=O, c=1 C211
Cryptand a=b=l, c=O C221

a=b=c=l C222

Figure 1. Synthetic macrocyclic polyether ligands.

hole. The crown ethers' ability to solubilize inorganic
salts in polar and nonpolar solvents was also demonstrated
(1,3-7). Pedersen relates in his account of the discovery
of the crown ethers (8) that his and his coworkers' ex-
citement "turned into elation when it was realized that,
at long last, a neutral compound had been found capable

of forming stable complexes with salts of the alkali
metals". The excitement was justified because previously
the alkali and alkaline earth metal ions were thought to
be inert and unreactive (9).

Since the initial synthesis of the first crown ethers,
hundreds of similar macrocyclic compounds have been pre-
pared (10), including various substituted crown ethers
and heterocyclic crown ethers in which either nitrogen
or sulfur atoms replace the oxygen atoms. Reinhoudt and
de Jong (11) and Bradshaw (10) have written excellent re-
views with extensive references on the synthesis of these
compounds.

Many of the crown ether compounds resemble naturally
occurring macrocyclic antibiotics such as valinomycin and
nonactin, both in structure and in their ability to form
stable complexes with alkali cations. Therefore, the crown
ethers have proven to be excellent models for these anti-
biotics and have even exhibited similar ion transport
properties across biological membranes (8,12-15). The

importance of macrocyclic ligands in membrane transport

has been summarized by Cockrell gt _1. (16), Fenton (l7),
and Morf gt a1. (18). The use of dicyclohexyl-lB-crown-6
in membrane transport is discussed by Grimali and Lehn (19).

Cryptands are another group of macrocyclic compounds
that were first synthesized by Lehn and coworkers (20).
They are three dimensional macrocyclic polyethers contain-
ing three polyether strands Joined together by nitrogen
bridgeheads (See Figure 1). The polyether strands form a
cavity which varies in size depending on the number of
ether oxygen groups contained in the strands. The cryp-
tands form very stable three dimensional complexes called
cryptates with alkali and alkaline earth metal cations by
encapsulizing the cation inside the cryptand cavity.
Cryptand compounds are named by stating the word cryptand
followed by the number of oxygen atoms in each polyether
strand. An example is cryptand 222 which is pictured in
Figure l. The name cryptand is usually shortened to Just
C222.

Therehmye been many interesting applications developed
for the use of macrocyclic ethers based on their complex-
ing ability for and selectivity of different alkali and
alkaline earth metal salts. A few examples are: 1) use
as anchor groups in ion exchangers developed for the sepa-
ration of cations with a common anion, or anions with a

common cation (21); 2) use as extracting reagents for uni-

valent and bivalent cations (22,23); 3) use in the formation

of transition metal complexes by incorporation of donor
atoms or groups into the polyether crowns to improve their
complexing ability with the transition metals (2“); A) use
in ion-selective electrodes, in which they act as neutral
macrocyclic carriers for cations such as K+ (18,25);

5) use in the determination of K+ and Na+ ions in serum (26);
6) use as catalyst for reactions such as nitrogen fixa-
tion and organic reactions in which alkali and alkaline
earth salts are used with the macrocyclic ligand serving
to improve the salts efficiency by increasing their solu-
bility (27); and 7) use of crown ethers in photographic
film in which metal complexed crown ethers containing four
or more oxygen atoms per molecule are incorporated into
pressure sensitive imaging materials and act as color
formers in the film (28).

Another interesting application is the use of macro—
cyclic compounds in the preparation of salts of alkali
metal anions (29). The first alkali metal anion salt was
synthesized by Dye and coworkers in 197“ (30). This com-
pound is important because it is believed to represent the
first example of a new class of compounds. Solvent-free
salts of the alkali metal anions are produced by stabiliz-
ing the cation by incorporation into a crown ether or
cryptand complex.

The above examples demonstrate the great importance

and flexibility of macrocyclic compounds. The chemistry

of crown ether complexes in various solvents has been
investigated extensively, however, there is little informa-
tion available concerning the ligands themselves. A study
of the relative solvation of the crown ethers in various
solvents would be very beneficial because it would provide
information concerning the behavior of the crown ethers
in the various solvents and may also provide an insight on
the structure of the solvents. It is to the end of obtain-
ing information on the solvation of a few selected crown
ethers that part of this study was conducted.

The use of nuclear magnetic resonance spectroscopy as
an analytical tool for the study of the complexation prop-
erties of macrocyclic compounds with alkali and alkaline
earth metals is well documented in the literature (31-39).
However, most of these studies use alkali and alkaline
metal nmr. Therefore, an investigation of the senSitivity
of proton and carbon-13 nmr in the study of the stability of
the complexes formed by 1806 with potassium iodide, potassium

hexafluorophosphate, and rubidium iodide was conducted.
B. MACROCYCLIC POLYETHERS

1. Introduction

 

A true understanding of the solvation phenomenon re-
quires some knowledge concerning the nature of the entity

which is being solvated. The chemistry of the crown

ether metal complexes has been investigated in water and
in various nonaqueous solvents using such methods as proton,
carbon-13, and alkali metal nmr, potentiometry, conductance,
spectroscopy, and extraction studies. The major interaction
between the crown ether and cation is electrOStatic in
nature. The major factors effecting the cation selectivity
and complex stability are 1) number, type, and arrangement
of donor atoms, 2) type and charge of the cation, 3) sub-
stitution on the macrocyclic ring, A) type of anion or
counterion, and 5) solvent properties. Many reviews exist
(l,2,7,27,AO-A3) that extensively cover this subject.
In addition, investigators from this laboratory (AA-51)
have added significantly to the existing information.
Therefore, it is not the intention here to repeat this
information. However, as stated earlier, information con—
cerning the ligands themselves is sparse. Following is an
attempt to bring together data concerning the general
prOperties of polyether crowns.

Recently there have been three books published about
macrocyclic compounds (27,A1—A2), containing a multi-
tude of references. In no way can this review duplicate
the quantity and depth of these works, consequently the
reader is directed to these references for any additional

information desired.

2. Prgperties of Polyether Crowns .'

Macrocyclic crown ethers were first synthesized by
the reaction of catechol and d,w—dihalides or a,w-ditosylates
in the presence of a base with the macrOcyclic polyether
being obtained by formation of a carbon-oxygen bond in the
cyclization step. An example of crown ether preparation

is shown below (1)

 

%
Z d“ + 4 NdOH + ’2. CH” CH 067404 CZ
CR‘LLChOL A;S zeta/oroc'tijneflrer

(I)

 

Find an 01¢)

 

+ 4NaOH + 4 HZO

SDBIBCé

The polyethers are thermally stable (dibenzo-lB-crown-6
can be distilled at 380°C at atmospheric pressure), but

must be protected from oxygen at high temperatures. The

10

polyether ring is destroyed by reagents which cause the
scission of aromatic and aliphatic ethers. Polyether
crowns can undergo many substitution reactions to form
polymeric products (52).

Macrocyclic polyethers with aromatic side rings are
colorless crystalline compounds while the saturated poly-
ethers are colorless, viscous liquids or solids with low
melting points. DB2A08 is known to exist as at least three
polymorphs, a glass and two crystalline forms having
identical melting points (19).

The melting point for an individual polyether rises
with the number of benzo groups attached to the ring.

(See References l,2,7,27,A0-A3.) Values of molecular
weights and melting points of a few selected polyether
compounds are shown in Table 1. Diameters of the cavities
for a few representative cyclic polyethers are shown in
Table 2.

Spectral data obtained from nmr, uv, and ir spectros-
copy are used to identify crown ethers. The protons on the
carbon attached to the ether oxygen and those on the aro-
matic moiety give characteristic nmr spectra with observed
intensities that are close to the theoretical number of
protons (1). Nmr spectra of crowns in deuteriochloroform
confirm the absence of terminal groups such as hydroxyl
or alkoxy which may be suspected to be present considering

the synthesis reactants.

11

Table 1. Molecular Weights and Melting Points of Some

Polyether Crowns.a

 

 

Molecular

Ligand (Ea/hag?) Me It ingclioints
B1505 268 79 - 79.5
DB1505 316 113.5 - 115
1806 26A 39 - U0
B1806 312 H3 — NA
DB1806 360 16“
DCl8C6 372 68.5 - 69.5
DB2UC8 ““8 113 — 11A
DB30010 536 106 - lOU.5

 

 

aReference (l).

12

Table 2. Diameters of Cavities in Some Cyclic Polyethers.

 

 

Polyether Cavity Size (A)
INCA 1.2 - 1.5 (a)
1505 1.7 - 2.2 (a)
DB1505 2.7 (b)
1806 2.6 — 3.2 (a)
DB1806 u.o (b)
2107 3.A - A.3 (a)
DB2A08 >u.o (b)
DB30010 >u.o (b)

 

(a) Reference (1).

(b) Reference (7).

13

All cyclic polyethers containing benzo groups have a
characteristic uv absorption maximum in methanol at 275 nm.
This absorption band is characteristic of compounds derived
from catechol. In general, molar absorptivities, E, of
crown ethers containing benzo groups depend on the number
of benzo groups attached to the crown ether. Crown ethers
with one, two, three, and four benzo groups and derived
from catechol have molar absorptivities of approximately
2100 - 2300, “MOO - 5200, 6300 - 7200, and 8&00 respectively.
Those crowns derived from A-t-butylcatechol show molar
absorptivities of 2700 for one attached benzo group and
5000 — 5200 for those containing two groups (1). Saturated
crown ethers such as 1806 and D01806 do not show any ab—
sorption above 200 nm, and consequently, cannot be studied
by uv spectroscopy since absorption below 200 nm is out of
the range of most commercial uv spectrometers. Table 3
shows characteristic features of uv spectra of some crown
ethers.

Ultraviolet spectra can be used as a qualitative
detection of the complexation of the crown ethers with
metal cations. For example, DB1806 in methanol shows one
peak at about 275 nm. On complexation with some metal
cations such as Na+, Li+, K+, Ca+2, and Ba+2, a shoulder
forms on this peak. In some cases the resolution of the

shoulder shows a dependence on the concentration of the

cation and becomes more resolved with increasing cation

1A

Table 3. UV Spectrometric Characteristics of Some Poly-
ether Crowns.

 

Ligand Solvent Peak Wavelengths (nm)

 

031806 Cyclohexane 278 (e A700) (a)
27A (6 AAOO) shoulder
283 (e 3600) shoulder
Methanol 27A (8 5200) (a)
Water 273.5 (a)
279 shoulder
Water 273 (e mSAOO) (b)
278 shoulder

B1505 Methanol 275 (a)

 

(a) Reference (1).

(b) Reference (53).

15

concentration. It is suggested that the increase in
resolution of the shoulder is an indication of the stability
of the complex formed such that the more resolved the peak
the more stable the complex (1). However, if the shoulder
does not appear on addition of a metal cation to a crown
ether solution, it cannot be definitely concluded that
complexation does not occur.

Infrared spectra has been used to confirm the absence
of a hydroxyl group. The ether linkage is confirmed by
strong, broad bands centering near 1235 cm-1 for an aromatic
- o - aliphatic linkage and near 1130 cm"1 for an ali-
phatic - o — aliphatic ether linkage (1). Therefore, ir
spectra for saturated crowns have a characteristic band

1 and aromatic crown ethers have characteristic

1

near 1130 cm-
bands near 1130 and 1235 cm- The ir spectrum will change
on complexation displaying band shifts corresponding to

the change in the bonding character of the ligand.

Solubility of Macrocyclic Ethers

 

The solubility of macrocyclic crown ethers tends to
be lower the higher the melting point. In general, com-
plex formation of the crown ethers with metal cations in-
creases their solubility in solvents of high dielectric
constants and decreases it in solvents of low dielectric
constants (l). Benzo groups attached to the ring de-

crease the solubility in water, alcohols, and many other

l6

solvents at room temperature, but, the benzo—substituted
crown ethers are readily soluble in methylene chloride and
chloroform. Saturated cyclic polyethers are much more
soluble in all solvents than the corresponding benzo com-
pounds. Table A lists the available solubilities from the
literature. DB1806 has been studied quite extensively, but

there are little data on the other crown ethers.

Heats of Solution

 

The heat of solution gives information concerning the
behavior of a solute in a particular solvent. The dif-
ference in the heats of solution of a solute in its stan-
dard states in two solvents, where one solvent is desig—
nated the reference solvent, gives the enthalpy of transfer,
AH; from the reference solvent to the other solvent.

This relationship can be expressed as follows:

AHgO1n (solvent,s) - AHgO1n (ref. solvent) =
(2)
AHg (ref. + solvent,s)

Arnett and Moriarity (52) have measured the heats of

solution at 25°C for D01806 as a mixture of isomers in

l7

 

 

: 00.H : Ufiom OHELOM

= 0000.0 : Hocmnpo

= 0000.0 : mCOpoow

z NH.0 = ocfiwfihmm

: H00.0 = Hocwpsmna

: mm0.0 = cwgsmopvmswppou

: H0.0 = 0006000 Hmnpo

: Hm.0 : EpomomOHco

= 0H0.0 : wcomcon

= 50000.0 oHLpoEH>mgm ocmxonoaomo

Amv 00m.0 H mm 0 000.0 ofippoefi>mgw ovflgoano

Iwgpop conpwo

ADV 00H H mm 0 m00.0 >5 opwgoazo

Iwgpop conpmo

Amv Doom 0 HHoo.o canpmefi>mcm Hocmnpme

ADV 00H H mm @ mmfioooo.o >3 topaz
Amv Oom.o H mm a moooo.o aficpm5a>anw topaz mowamo

A00 00mm 0 0H0.0 pmumz

Aav comm @ mmo.o nope:
Adv Doom 0 mmo.o Loam: momaom
Apopfia\moaoev mpHHHQSHom vogue: pco>aom ogmmfiq

 

.mponpm czoao oEom mo mmeHHHQSHom

.3 mfinme

18

.Azmv mocopmmmm A00
.Ammv mocohomom ADV
.AHV monopommm Amv

 

 

: nmm.0 = Hocoamxlzam

= 0H>.0 = ocmnpo

nocoHcOAceum.m.m

= 00m.0 : Ufiow owpmp59080gmlm

= m0H.0 : ogfiaficmOLOHQOIN

: nma.0 = mconchthHc

= m~0.0 : wvmnovfimucon

= mm0.0 : Honooam Hancocono

: 0H.0 : HomoLOIE

: m>0.0 : Hocooaw HSNcon

= NH.0 : oaflmpficoncoo

: mmm.0 = ocHHHcm

on 00mm 0 50.0 = ococonooumom

: 0w0.0 = oHHLquouoom

= 030.0 = mefixocfismflgnmeAU

: w:0.0 = ocmnpoEoppfic
Amv oom.0 H mm 0 0m0.0 afipposa>mmm oUHEmEpomamcuoEHp mowama
AhmpHH\mOHOEv mpfififindaom Uonpmz pcm>aom Ucmwfiq

 

.coscfipcoo

.: magma

19

THF, DMSO, and AC and for the individual isomers (cis-syn-
cis and cis-anti-cis isomers) in AC by calorimetric titra-
tion. The heats of solution, heats of complexation, and
the enthalpy of transfer for the crown ethers complexed
with various alkali metal and ammonium salts in the above
solvents were also determined. Values for the ligands
were 6.30 s 0.19, 7.58 i 0.A2, and 7.32 i 0.2M kcal mol"1
for the isomer mixture in THF, DMSO, and AC respectively.
The isomer labelled A (melting point 5A.5 — 61.5°C) had a

1

heat of solution of 8.57 i 0.27 kcal mol- where isomer B

(melting point 81 - 83°C) had a heat of solution of 10.0
i 0.27 kcal mol_1. Based on the positive values of the

heats of solution, the solution process is endothermic.

It is interesting that the order of the dielectric constants
for the solvents, A5, 20.7, and 7.58 for DMSO, AC, and THF
respectively, follow the order of the heats of solution
values. This suggests that the lower the dielectric constant

of the solvent the less energy required to dissolve the

ligand.

Conformations of Crown Ethers

 

Crystallographic data (55-58) show that the angles and
bond lengths of the cyclic ethers do not change signifi-
cantly on complexation, but that there are conformational
changes. For example, in the uncomplexed DB1806 ligand,

the center of the ligand is at the crystallographic center

20

of symmetry and the six oxygen atoms are not coplanar,
while in the ligand, complexed with rubidium thiocyanate,
the six oxygen atoms are almost coplanar and the molecule
has 02 symmetry (55). Bush and Truter (57) determined the
crystal structures of DB1806 complexed with NaBr, B1505
complexed with NaI, and DB30010 complexed with KI. The
structures show that in the cases of the DB1806 and B1505
complexes, the cations are coordinated not only to the
ligand, but to water molecules and/or the respective anions
as well.

However, in the case of DB30010, the ligand is able
to enclose the cation by wrapping itself around it and to
completely replace the hydration sphere of the alkali metal
cation. It was concluded that while large ring systems
can completely replace the hydration sphere of the alkali
metal, the small rings can only partially replace this
sphere allowing additional interactions between the solvent,
or anions, or other ligands and the cation. They further
conclude that the accessibility of the complexed cation to
additional ligands probably accounts for the variation of
selectivity with the anion and solvent used.

Shamsipur (50) also showed the ability of large crown
ethers to form three dimensional "wrap around" complexes.
He found that the ligands' ability to form such complexes,
with the same cation, decreases with the size of the

ligand such as DB30010 > DB2AC8 > DB2107.

21

Basicity of Crown Polyethers

 

Lockhart et_al. (59) used a lanthanide shift reagent
as a Lewis acid to study the basicity of the oxygen atoms
in the asymmetrical ligands 31505 and B2107. A clear dif-
ferentiation in the behavior for sets of hydrogen nuclei
of the crown ethers was observed. This behavior was as-
cribed to differential basicity of the individual oxygen
atoms and their selective coordination to the lanthanoid
shift reagent. The oxygens nearest (on a time-averaged
basis) the site of the strongest lanthanoid interaction are
the most basic. These oxygens are located directly op-
posite the benzo group, oxygen number three for B1505 and
oxygen number four for B2107 if the first oxygen next to

the benzo ring is considered as oxygen number one.

Dale and Kristiansen (60) found that all the oxygen atoms
in more symmetrical crown ethers are equivalent.

Shchori et_al. (33) concluded that the basicity of
four oxygens in the macrocyclic ring of DB1806 may be

affected by the introduction of electron-withdrawing or

22

electron-donating substituents into its aromatic rings.
Introduction of the electrophilic NO2 group into the benzo
ring of DB1806 caused a five fold decrease of the formation
constant of the sodium ion complex in DMF solution.

In summary, 1) the asymmetrical ligands show differen—
tial basicity in the individual oxygen atoms where the more
symmetrical ligands display equivalent basicity in their
oxygen atoms, and 2) ring substitution can influence the

basicity of the oxygen atoms of crown ethers.

Complex Formation

 

Crown ethers form mostly one to one (1:1), cation to
ligand complexes, but they can also form complexes of dif—
ferent cation to ligand ratios. Shamsipur (50) found the
presence of three sodium DB30010 complexes, Na2(DB30010),
Na(DB30010)2, and Na(DB30010) in nitromethane and aceto-
nitrile solutions. The capability of the crown polyethers
of forming more than just 1:1 complexes results when the
diameter of the cation to be complexed with the ligand is
either smaller or larger than the diameter of the ligand
(See Table 2). The 2:1 complexes are called "sandwich"
compounds where the cation is located between two molecules
of crown ether; and the 3:2 complexes are called "club
sandwich" complexes where the two cations are layered
between three ligands (ligand:cation:ligand:cationzligand)

(3A,A5,6l). The existence of the 2:1 complex has been

23

confirmed by Mallenson and Truter (61) by x-ray crystallog-
raphy. The existence of the 3:2 complex is still un-
confirmed. The conformational rearrangement of the ligand
during complexation occurs so as to allow for the greatest
interaction between the oxygens of the crown ether and

the cation. This interaction usually results in the cation
being equi-distance from each oxygen atom in the ligand
cavity. The cation may lie either in the plane of the
cavity, slightly above or slightly below the cavity.
Therefore, the incomplete enclosure of the cation by the
ligand due to size differences of the ligand and cation
(See Tables 2 and 5) or the inability of the ligand to
completely "wrap around" the cation results in possible
interactions between the cation and solvent molecules,
counterions, and other ligands and may result in the

formation of the "sandwich" or "club sandwich" complexes.

Toxicity of Cyclic Polyethers

 

All new compounds must be treated with respect. Little
is known about the toxicity of cyclic polyethers, but from
the information available these compounds should definitely
be handled with the upmost care to guard against immediate
as well as any possible long-delayed harmful action.

The strong complexing ability of the crown ethers with

the alkali, alkaline earth, and some transition metal

2A

Table 5. Ionic Diameters of Alkali Metal Cations.

 

 

Cation Diametera
Li+ 1.86
Na+ 2.3A
K+ 2.98
Rb+ 3.28
05+ 3.66

 

(a) Reference (62).

25

cations, makes these compounds suspect to possible inter-
ferences in the vital processes in which the cations
participate (8). It should be remembered that some of
the polyether crowns are being used to determine K+ and
Na+ in serum (26).

An approximate lethal dose for dicyclohexyl-lB-crown-6
for ingestion by rats is 300 mg/kg and causes death in 11
minutes. However, in a lO-day subacute oral test of
60 mg/kg/day, the compound did not exhibit any cumulative
oral toxicity to male rats and a dose of 200 mg/kg was
not lethal in lA days. Dicyclohexyl-l8—crown-6 also
produced some generalized corneal injury, some iritic
injury, and conjunctivitis when introduced in propylene
glycol. It is readily absorbed through the skin of test
animals and was fatal when absorption reached 130 mg/kg
level (8).

Leong (63-6A) reported that vapors of 120A caused
testicular atrophy when inhaled by rats. Gad (65) reported
that rats and mice given 1806 i.p. at 20 - 160 mg/kg/day
showed aggression, tremors, muscle weakness, and degrada-
tion of some reflexes. Rabbits given i.v. 6.0 mg/kg/day
1806 displayed tremors, hyperactivity, unsteady gait, and
stereotypic behavior. For all cases, acclimation was
observed when the dose was maintained constant and the
symptoms disappeared when the treatment was discontinued.
The possible dangers of the polyether crowns is therefore

emphasized.

26

C. TRANSFER ACTIVITY COEFFICIENTS (FREE ENERGY OF

TRANSFER)

1. Introduction

 

The transfer activity coefficient, designated by Yt’
has been called the medium effect, medium activity coef—
ficient, primary medium effect, partition coefficient, and
the distribution coefficient (66). Regardless of how it
is labelled this quantity represents a measure of the effect
of changing the medium or solvent of a solute. It is de-
fined as a measure of the difference between the partial
molal free energies of the solute in its standard states
in a solvent 3, and in a reference solvent, r. This
difference is known as the molal free energy of trans-

° for the solute (67-69). This relationship

t
is shown in Equation (3)

fer, AG

r i = RT fin Yt (3)

The chemical activity of a solute is defined by Lewis (70)
in terms of the chemical potential. If the reference sol-
vent is chosen as water, the chemical potential or partial
molal free energy for the solute in its aqueous and non-
aqueous standard states can be written as

27

and

— = _0 *
Gi sGi + RT in A1 (5)

where

01 = partial molal free energy of solute i;
fiG: and SC; = standard free energy of solute i in
water and in solvent 5, respectively.

A1 and A; - activity of solute i in water and in

solvent 5, respectively

The activities of solute i in water and in the nonaqueous

solvent can be represented as

A1 = mi in (6)
and
AI = mi sYi (7)
where
mi = molality of solute i;
in = activity coefficient referred to an aqueous

standard state; and

28

sYi = activity coefficient referred to a standard

state in solvent 5.

In order to compare the activity of a solute in different
solvents, a single reference state for the solute must be
chosen. If henryan standard states are assumed, then in
the limit as the molality of solute i approaches zero, the
activity coefficient of solute i in solvent s, sYi’ ap-
proaches unity. However, the activity coefficient of
solute i, in’ in water approaches unity in water only.

In any other solvent 8, as the molality of solute 1 ap—
proaches zero, the activity coefficient referred to the
aqueous standard state, in’ will approach the transfer

activity coefficient, Yt'

lim sYi + l in solvent 5 g (8)
m+0

lim in + l in water only (9)
m+0

lim

m+0 in + Yt in any other solvent (10)

Therefore, when referred to an aqueous standard state

the activity coefficient of solute i in a nonaqueous

29

solvent can be represented as the product of the salt
activity coefficient, 8Y1, which represents the effect

of electrostatic ion-ion interactions and the transfer
activity coefficient, Yt’ which represents the difference
in ion-solvent or interparticle interactions in the two

solvents (66).

in = sYi Yt (11)

Multiplying both sides of Equation (11) by the molality

of solute, i, mi, Equation (12) can be obtained
A - * 2
i — Ai 7t (1 )

Finally, combining Equations (A) and (5) with Equation

(12) results in the defining Equation (3).

- 0 =
AGt _ 501 — w 3 RT in vt (3)

The transfer activity coefficient, Yt’ will therefore equal

unity by definition in the reference solvent.

2. Determination of Transfer Activity Coefficients

Determination of transfer activity coefficients has
been reviewed in detail by Popovych (68-69) and by

Bates (67). It is possible to measure accurately the

3O

transfer coefficients of nonelectrolytes and electrically
neutral electrolytes. However, transfer activity coef-
ficients for single ions are not experimentally measurable
since it would be necessary to measure some process which
transfers the single charged species into or out of the
solution (70). This constitutes the transfer of ionic
charge Z, across the interface of two solvents where a
potential difference, 0, exists. This process would result

in an additional term in Equation (3) of wZF.

D
Q
0
II
0|
I
S
0|
Ho
u

RT in Yt + WZF (l3)

Ionic transfer activity coefficients presented a
challenge to chemical ingenuity that has been dealt with
by the age old process of estimation, namely the use of
"extra thermodynamic" assumptions which have met with a
great degree of success. The literature which amply covers
this subject includes reviews (67-69,7l-77) which discusses
the various assumptions, their merits, and the importance
of single ion transfer activity coefficients. Consequently,
there will be no further discussion here since this work
deals with the transfer activity coefficients of non-
electrolytes.

The three methods that are most often used for the
determination of transfer activity coefficients of solutes

are solubility measurements, calculation from standard

31

potentials of a galvanic cell reversible to its ions, and
vapor pressure measurements. The solubility method is

most often used as it can be applied to nonelectrolytes and
electrolytes alike. These three methods are discussed

below.

a. Solubility Studies

 

When a saturated solution of solute i, in water (or
in other reference solvents) and in another solvent, S is
in equilibrium with solid solute i, the partial molar free
energies of solution are equal (See Equations (3-5)).
Therefore, the value of the free energy of transfer from

water to solvent S is given by

 

 

0°- ‘09 (A) -
AG; = s i w 1 = in i saturated (1“)
RT (Ai)saturated

and the transfer activity coefficient is given by

(A )
in i saturated

(A

 

in 7t = (15)

*>
i saturated

Solubility measurements can be used to determine Yt

for electrolytes and nonelectrolytes and is given by

32

 

sKSp
yt = (for electrolytes) (l6)
s sp
Si '
Y = E——- (for nonelectrolytes) (17)
t 881

where wKSp and sKSp are the solubility products of an
electrolyte in water and in solvent S, and WSi and S81
are the solubilities of a nonelectrolyte in water and in
solvent S. In the case of electrolytes the salt activity
coefficients are usually treated in one of two ways:

1) ignored or assumed to be unity in the case of low solu-
bilities, or 2) corrected for by the use of the Debye—
Huckel or a similar equation (66).

The methods used in the determination of solubilities
are classified as "synthetic" or "analytic". Synthetic
refers to methods applied to a system of solute and solvent
in which the temperature or the pressure or both are varied
until the solute just dissolves. The analytical methods,
in general, consist of obtaining a saturated solution and
determining the concentration of the resulting solution by
a suitable method (78).

Many methods are described in the literature that
utilize a variety of simple and complex apparatus (78-82).

It seems though,that in all systems, regardless of the

33

method employed there are three important factors that

affect all solubility measurements, purity of reagents,
temperature regulation, and especially establishment of
equilibrium.

It is generally assumed that equilibrium conditions
are established when repeated analyses of a solution give
constant results. The time required to reach equilibrium
depends on the solute and the solvent. It can range from
minutes to several weeks. Other factors to be considered
are the possibility of hydrolysis in aqueous systems, the
stability of the solution, the inertness of the solubility
vessel toward the components of the solution, the tendency
of some solutes toward colloidal formation at concentrations
at, or slightly above, saturation, and particle size.

In some systems investigations of the above factors
could be very difficult if not impossible. Atomic ab-
sorption or emission can be used to verify the inertness
of the solubility vessel by testing for metals such as
sodium. Particle size can be controlled by grinding.

Measurement of the resulting saturated solution depends
on the solute and solvent involved. Some of the major
techniques suggested and used in various studies are
gravimetric or residual weight, chemical analysis; elec-
trical methods such as conductometry, electromotive force,
polarography, and pH determinations; optical methods in-

cluding colorimetry, spectrophotometry, turbidimetry,

3A

nephetometry, refractometry, interferometry, polarimetry,

and microscopy; radioactive tracer methods and some special

application methods.

b. Determination by Measurement of the Standard
Potentials of a Galvanic Cell Reversible to its

Ion

Measurement of the standard potential of an electrolyte
in a galvanic cell reversible to its ions, in a reference sol-
vent and another solvent 5, and application of the following
equations to the resulting potentials, can yield the trans-

fer activity coefficients for the electrolyte (68).

 

o - _ o
AGt - nFAEcell (18)
(SE; - WEI)

K = RT in lO/F

When problems such as poor solubility of a salt arise
in the standard potential measurements, polarographic
half-wave potentials are substituted in Equation (19).
This causes errors of its own because the half-wave poten-

tial values are measured in the presence of supporting

35

electrolytes in cells with liquid junctions. Even with
reproducible conditions the half-wave potential value is
only an approximation of the corresponding standard po—

tential (68).

c. Vapor Pressure Measurements

 

In this method either the partial pressures of a
volatile solute or volatile solvent can be measured.
Measurements for a volatile solute are based on the same
principle as that for solubility measurements. If two
phases are in equilibrium with the same third phase they
have equal chemical potentials. Therefore, as with solu-

bility measurements, equation (20) can be derived.

A
Y = *1 (20)
t
A
A = activity of the solute referred to infinite dilu-
tion in water,
*
A = activity of solute referred to infinite dilution

in a nonaqueous solvent.

If the partial pressures of a volatile solute above aqueous
and nonaqueous solutions are measured the transfer activity

coefficients may be obtained from Equation (21) (68).

Yt = —? (21)

where P and P* are the partial pressures of a volatile
solute above the aqueous and the nonaqueous solutions and
A and A* have been defined above. For equal activities

in both solutions, Equation (21) becomes
P*
Yt = IT (22)

This method at present is the least popular of the
three methods discussed for the determination of transfer
activity coefficients. However, with the advancements in
the technology of gas chromatography, more reliable

quantitative information may be obtained.

3. Ligand-Solvent Interactions

 

Differences in the transfer activity coefficients are
caused by the nonspecific effects due to the differences
in the dielectric constants of the solvents and by the
specific interactions between the solute and the solvents
(66). These two factors are interrelated and the values
of the transfer activity coefficients must be interpreted
with this interrelationship in mind.

A positive value of the logarithnl of the transfer

activity coefficient for the transfer of a solute from

37

some reference solvent to another solvent indicates that
the solute is more solvated by the reference solvent than
by the other solvent. Hydrogen-bond accepting solutes
would be more solvated in water than ina nonaqueous sol-
vent, but a hydrogen-bond donor would be more strongly
solvated by the more basic solvents such as dimethylforma—
mide and dimethylsulfoxide.

Thermodynamic quantities such as the free energy,
enthalpy, and entropy of transfer for the transfer of
inert solutes from water to other solvents can be inter—
preted in terms of the perturbation by the solute of the
solvent structure. In the solvation process of a non-
polar molecule there are several major contributions to
the solvation energy: 1) formation of a cavity for the
solute in the solvent; 2) solute-solute dispersion inter—
actions; 3) the energy change for the transfer of a mole-
cule from its standard state in the gas phase to its
standard state in a solution; and A) the possible struc-
ture making or structure breaking energies in highly struc-
tural solvents like water (68). Solvation of nonpolar
species is energetically unfavorable in water as compared
to the less structured solvents because of a large entropy
loss due to "icebergs" or "cages" of water structure
around the solutes (68).

Jolicoeur and Lacroix (82) investigated nonpolar or

hydrophobic solutes in light and heavy water. Their

38

objective was to establish whether the geometry of a series
of isomeric ketones, which included saturated, unsaturated,
and polycyclic varieties, some with varying degrees of
branching, had a significant influence on their thermo-
dynamic properties, especially on those properties which
are commonly assigned to the structural effects. They
proposed that the free energy of transfer can be broken

down into several contributions:

0
A03 = A0g(0AV) + AGE(SOL) + AGg(STR) + AGEX (H1) (23)
These terms represent the solvent isotope effect of light

and heavy water on the free energy that is due to:

CAV : process of cavity formation in the solvent;

SOL : solvation of the polar group of the molecule
(dipole-dipole interactions, hydrogen bonding);

STR : structural rearrangement of the solvent around
the cavity enclosing the solute; and

HI : solute-solute interactions (hydrophobic inter-

actions).

Millen (83) calculated a solvent structure parameter
for protic and dipolar aprotic solvents, by considering
the work of creating a liquid surface equal to the surface

of a spherical solute of radius r. The free energy is

39

given by,
AGfi = Aflr20 (2“)
where
r = radius of spherical cavity; and

Q
ll

free energy per square centimeter of liquid

surface.

Hermann (8A) found a correlation between the solubility
of hydrocarbons and their surface area which has been
used as evidence for the importance of the G°(CAV) term
in the free energy of solvation. If Equation (2A) and
the above correlation are assumed, an equation for esti-
mating G°(CAV) is given by,

G°(CAV) = An r2 0° (25)

where o° is the bulk surface tension and r the radius of
a spherical cavity (82).

It has been postulated that non-polar or hydrophobic
solutes in water can promote some hydrogen-bonded struc-
ture in the surrounding solvent. This structure-making
effect has been supported in part by large entropy losses

upon dissolution of inert gases in water. Jolicoeur and

A0

Lacroix came to the following conclusions: 1) the differ-
ences in the observed values for AGE for the different
ketones originates in the structural part of AG°; 2) the
magnitudes of AGE and ARE of large hydrophobic solutes are
strongly influenced by the structure of these solutes; and
3) the solvent isotope effect on the thermodynamic quanti-
ties for the solvent appears to be an important contribu-
tion to both AGE and ARE.
Lucas (85) studied the effect of the size of some non-
polar solutes on their transfer from one solvent to another
solvent. The free energies of transfer were computed by
use of the scaled particle theory (SPT) advanced by Reiss
_E._l-a Pierotti, and Wilhelm and Battimo (86). This theory
is also used to predict solubilities, heats of solution,
and partial molar volumes of simple gases in nonpolar
solvents. In the SPT the solvent molecules are considered
to be hard spheres which are related to the real solvent
molecules in that the hard-sphere diameter is as close as
possible to the true solvent molecular diameter of the
solvent. Plots of the free energy of solution from the
gaseous state for a nonpolar solute in the various solvents
versus solute diameter shows that a nonpolar solute with
a diameter of approximately A A, is more soluble in a
nonpolar solvent than in water. Calculations of AC?

from SPT shows that the negative values of AGE for the

transfer from water to a nonpolar solvent results mainly

Al

from the fact that the nonpolar solvent molecules used

in the study have a larger size than the water molecules.
It is also shown that the importance of dispersion forces
relative to cavity energy effects increases with solute
size and further that solvent dimensions are important in
determining the sign of the free energy of transfer for

a nonpolar solute from one solvent to another.

Vesala (87) determined AGE for some nonelectrolytes
such as sulfur dioxide, variously substituted benzenes and
amines, and phenanthrene from light to heavy water by
solubility measurements. The main criterion for solute
selection was that they did not contain any exchangeable
hydrogen atoms. Two linear correlations were found, one
between the free energy of transfer of the corresponding
nonelectrolytes from the gas phase to a hypothetical one
aquamolal solution in water, and another between the free
energy of the nonelectrolytes from light to heavy water
and the heats of melting, AHm, of the various compounds.
Values found for AGE for the nonelectrolytes were small
and varying in sign, ranging from +192 cal mol-1 and +158

1 for phenanthrene and methoxybenzene, respectively,

1

cal mol-
to -50 cal mol- and -A5 cal mol'l for argon and tributyl—
amine, respectively.

Dahlberg (88) determined the free energy and enthalpy

of transfers from light to heavy water of several ketones

and alcohols that were representative of aliphatic, cyclic,

A2

and aromatic compounds such as acetone, cyclohexanone,
benzene, naphthalene, toluene, and benzyl alcohol. The
free energies of transfer were determined by solubility
measurements. Free energies of transfer were zero for

most of the nonelectrolytes studied. Exceptions occurred
for some cyclic compounds with side groups such as toluene
with a AGE value of +25 i 15 cal mol-l, while benzene has a
AGg value of O i 15 cal mol—l. The results were explained
in terms of solvent structure-breaking and solvent struc-
ture-making spheres of influence with polar groups being
structure breakers, while nonpolar groups are structure
makers. If the polar and nonpolar groups of structural
influence overlap, their ability to alter solvent struc-
ture decreases as they tend to cancel one another. This
effect is termed the concept of overlapping spheres.
Values of AGE for benzene and acetone were near zero
while values for toluene, nitrobenzene, and cyclohexanone

1, +157 cal mol-l, and +52 cal mol-l

were +25 cal mol-
respectively. This difference was explained to be caused
by addition of side chains to the ring compounds. Compound
shape itself was also shown to be an important influence
on the values of AGg.

Cox (89) used vapor pressure measurements as well as
solubility measurements to obtain the free energies, en-

thalpies, and entrOpies of transfer of nonelectrolytes from

water to mixtures of water with dimethyl sulfoxide,

A3

acetonitrile, and dioxan. Large increases in both the
enthalpy and entropy with correspondingly small changes
in the free energy were found.

Ahrland (90) points out that the ratio of the di-
electric constant to the dipole moment, e/u can reflect
the relative degree of solvent structure. Water,whose
high degree of structural order due to hydrogen bonding,
has an exceptionally high e/u ratio. Methanol which dis-
plays weaker hydrogen bonding has a lower e/u ratio than
water, but a higher ratio than for other aprotic solvents.
A comparison of these ratios with values found for the
free energy of transfer of a solute for the same solvent
would prove interesting if the order of the magnitudes
were the same. This relationship could lead perhaps to the
use of E/U ratios as an indication of the relative values
of AGg(STR) for the various solvents.

These selected illustrations demonstrate the importance
and use of the magnitudes and signs of the free energies
of transfer as a diagnostic tool for solvent structure
elucidation, solute influence on solvent structure, and
solute-solute interactions. The separation of the free
energy of transfer into individual contributions can aid
in at least differentiating if not isolating the important

factors influencing the solvation process.

AA

D. PROTON AND CARBON-13 NMR STUDY OF 1806 WITH SOME

 

POTASSIUM AND RUBIDIUM SALTS

 

1. Introduction

 

About half of the known nuclei possess spin or angular
momentum which generates a magnetic moment along the axis
of spin. If the spinning nuclei are placed in an external
magnetic field Ho’ their magnetic moments align either with
the field or against the field. (This is a simplestic
View. For further detail see References 91—93). Energy
must be absorbed in order to "flip" the spinning nuclei
into a higher energy level or excited state - one anti-
parallel to the field H0 in the case of protons and
carbon-13 (92). A radio frequency (rf) field H1, can
be applied to the system to provide the energy necessary
to "flip" the nuclei. Either a constant magnetic field

H can be applied and the rf field H1, is varied as

0’

described above or a constant rf field H is applied

1’

and the magnetic field H is varied. In either case

0’
the energy is varied until it matches the energy required
to cause transition of the spinning nuclei which produces
the observed absorption signal.

The frequency at which the nuclei absorbs depends on
the magnetic field strength that the nuclei experiences.
This field strength is different from the applied field

strength due to the chemical environment of the nuclei

“5

and is called the effective field strength. The effective

field strength can be expressed by
H = HO(1 — o) _ (26)

where o is the screening or shielding constant. The shiled—
ing constant c, is generally the sum of two terms, a para-
magnetic term (shielding) Up’ and a diamagnetic term (de-

shielding) od.
0 = 0d + o (27)

The diamagnetic term Od’ depends on the circulation of
local electrons induced by the applied field about the
nucleus. The paramagnetic term op, depends on several
factors such as the energy difference between the ground
and excited states, electron density in the outer p
orbitals of the nucleus, and the distance between the
nuclei and the outer p orbital is inversely proportional
to the excitation energy and directly proportional to
atomic number (9A). In proton nmr Cd is the important
factor in the shielding constant while Up is the important
factor in carbon—l3 nmr (9A) and in multinuclear nmr in
general.

Separate absorption bands occur for nuclei of the same

molecule that are in different chemical environments.

A6

"Chemical shift" is the term used to describe the dif-
ferences in resonance conditions required for the same
isotope in different chemical environments. The absorp-
tion bands found for the nuclei are referred to as the
chemical shifts for the nuclei.

The chemical shifts are expressed with reference to an ar-
bitrary standard. Tetramethylsilane (TMS) is the most common
reference chosen for proton and carbon-l3 nmr because its'
nmr signal appears near one of the extreme ends of the
carbon-13 shielding range for neutral molecules. The
chemical shift of TMS is referred to as upfield (0 Hz or
0 ppm). The nmr signal of a proton or carbon-l3 nuclei
that is less shielded than the nuclei of TMS will appear

downfield from the TMS nmr signal.

2. Proton and Carbon-l3 NMR of Polyether Complexes of

 

Alkali and Alkaline Earth Metals

 

Proton NMR has been used to identify macrocyclic
polyethers (1). Lockhart (95) used proton NMR as one method
of characterizing the structures of some newly synthesized
crown ethers. Both proton and carbon-l3 NMR shieldings
are very sensitive to variations in molecular structure
(9A). When a change in molecular structure occurs changes
in the chemical shifts and vicinal coupling constants,

important in conformational determination,may also be

17

observed depending on the resolution power,and therefore
on the operating field strength of the NMR instrument used.
Complexation of a compound causes a change in molecular
structure and consequently causes spectral changes that

can be analyzed to give important conformational informa-
tion about the compound. This is the case in the complexa—
tion of crown ethers with metal cations. The addition of a
metal cation to the crown ethers induces two major spectral
changes: 1) a change in the vicinal coupling constants,
and 2) a change in the chemical shifts. By appropriate
choice of crown ethers and the use of a high field NMR
spectrometer, the assignment of protons in the ether region
is possible and the vicinal coupling constants can be ob-
tained.

The use of proton and carbon-l3 NMR techniques have been
fundamental in the elucidation of the structure of some
crown ethers in solution. Live and Chan (38) analyzed
proton and carbon-l3 NMR spectra to ascertain the solution
structures of B1806, DB1806, and DB30010 complexes with

+
Na , K+, 05+, and Ba2+

in water, water-acetone mixture,
and chloroform. Benzo-substituted crown ethers were
chosen because the benzene groups, due to their ring cur-
rent magnetic anisotropy, causes the separation of the
ether resonances and deduction of the spatial relationship

between the ether protons and the aromatic group leading

to less complex spectra. Also with the increased spectral

A8

resolution the ether vicinal coupling constants were readily
extracted. A high field NMR spectrometer was used to take
thepumnxnispectra operating at 220 MHz and carbon-l3 spectra
operating at 55.3 MHz. Carbon—l3 and proton chemical shifts
were referred to internal tetramethylsilane. DB1806 with
four nonequivalent hydrogens gave the simplest proton
spectrum. B1806 and DB30010 with seven and six nonequiva-
lent hydrogens respectively,produced spectra that were

more complex. Chemical shifts were small. For example,
proton NMR shifts were all less than 0.5 i 0.002 ppm.

From a detailed analysis of the proton and carbon-l3 NMR
spectra, the proton-proton vicinal coupling constants, and
the salt-induced chemical shifts Live and Chan concluded
that the complexes of B1806, 051806, and DB30010 in the
various solvents are in syn- and anti-gauche conformations

with rapid rotamer interconversion. X-ray crystallographic

(aw-:7

glue/I: Conformation,

data of DB1806 show the presence of both trans and gauche

2+ complexes of DB30010 have

rotamers. The K+, 05+, and Ba
conformations in solution that are consistent with KI-

+
DB30010 complex in the crystal. However, the Na -DB30010

complex is different from the others which demonstrates

A9

the ability of DB30010 to adapt itself to different-sized
cations.

Formation constants derived for the complexes were

consistent to those derived by other workers (A3,50).
For example, log Kf (formation constant) for DB3OClOCS+
complex in AC was A.23, Shamsipur (50) found an average
for log Kf of A.0 for DB3001008+ complexes with various
anions.

Lockhart, gt_§l. (39) used proton NMR to study the con-
formations of B1505 and B2107 in solution as free ligands
and as complexed ligands with alkali-metal iodides or thio-
cyanates. Deuterated methanol (CD3OD) was used as the sol-
vent. Most of the spectra were obtained at 90 MHz on a
Bruker Spectrospin HFK—6, but some spectra were obtained at
60 MHz and spectra for free B1505 and its NaI complex were
obtained at 220 MHz. The results indicate that the gross
conformations of the B1505 and B2107 complexes in solution
are similar to the conformations found in the crystal forms
of the complexes. The possibility of the use of proton NMR
spectra as a diagnostic tool in the investigation of complex
formation stoichiometry was indicated as a result of the
following observations. The proton signals for ligand:
cation combinations expected to give l:l complexes were
downfield from those of the free crown at all crown:cation
ratios investigated, while proton signals for ligand:cation

combinations expected to give 1:1 and 2:1 complexes, the

50

chemical shifts were upfield from free crown for crown:
cation ratio of 2:1 or more and downfield at low crown
cation ratios. Chemical shifts observed in the study were
small as expected, with all shifts being less than 0.2 ppm
and a majority of them less than 0.1 ppm with a precision
of 10.002 ppm.

Alkali metal NMR has been shown to be a powerful tool
for the investigation of the complexation of macrocyclic
polyethers with alkali metal cations because it is very
sensitive to the immediate chemical environment of the
alkali metal cations (31-37). As shown above,proton and
carbon-l3 NMR are powerful tools for conformational elucida-
tion of the crown ether complexes in solution. However,
proton and carbon-13 NMR have been considered unsuitable
for complexation studies, especially in the determination
of formation constants, of the macrocyclic polyethers
because of the small chemical shifts observed between
complexed and uncomplexed forms of the ligands (32) and
because the spectra of cyclic polyethers can be quite com-
plex (38). As mentioned above, Live and Chan (38) reported
some formation constants obtained in their study from carbon
NMR data. It is to be remembered however, the chemical
shifts involved were obtained with a high field NMR spec-
trometer with the chemical shifts good to within 10.5 Hz
(mi0.002 ppm) for protons and :1 Hz (mi0.02 ppm) for carbon

1
and also that benzo-substituted crowns which lead to better

51

resolution of the NMR data was used.

It is the purpose of this part of the study to investi-
gate the sensitivity of proton and carbon-l3 NMR instruments
of low operating fields (proton-60 MHz, carbon-l3-20 MHz)
in the complexation of the saturated unsubstituted crown
ether 1806, with potassium iodide, potassium hexafluoro-

phosphate, and rubidium iodide salts.

CHAPTER II

EXPERIMENTAL MATERIALS AND METHODS

52

A. MATERIALS

 

1. Ligands

B1505 was purified by recrystallizing at least three
times from reagent grade n-heptane (Fisher); vacuum dried
at least 2A hours at room temperature prior to use; melting

point 80.5°0; literature 79 - 79.5°C (l).

DB1806 was purified by recrystallizing at least three
times from reagent grade benzene (Fisher); vacuum dried at
least 2A hours at room temperature prior to use; melting

point l6A.5 - 165.5°C, literature 16A°C (l).

1806 was purified by first forming the solid aceto-
nitrile 1806 complex. The adduct was precipitated from
an 1806 solution in acetonitrile by cooling it in an ice—
acetone bath. The solution was filtered rapidly and aceto-
nitrile was removed under vacuum; melting point 37—38°C,

literature 39°C (96).

2. Salts

 

Benzoic acid (Mallinkrodt) was purified and dried by

 

John Hoogerhide (A6).

Potassium iodide (Mallinckrodt) was dried in the vacuum

 

oven for 72 hours at 60°C (A8).

53

5A

Potassium hexafluorophosphate (Pflaltz and Bauer)
was dried in the vacuum oven for A8 hours at room tempera-

ture prior to use.

Rubidium iodide (Fisher) was purified by recrystalliz-

 

ing from methanol and dried for A8 hours in the oven at

110°C.

Sodium_perchlorate (G. F. Smith) was oven dried for

 

several days at 150°C.

Sodium tetraphenylborate (J. T. Baker) was vacuum

 

dried at 50°C for several days.

Tetraethylammonium perchlorate (Eastman) was vacuum

 

dried several days at room temperature.

3. Solvents

Methanol (Mallinckrodt, Fisher) was refluxed over
magnesium turnings (approximately 20 g/l) and iodine
(0.1 g/t) for m2A hours and then distilled under nitrogen
atmosphere. Distillate was allowed to stand over activated
molecular sieves (heated at approximately 500°C for at
least 2A hours) for at least 2A hours and redistilled from

the sieves under nitrogen atmosphere.

Water was used as received from the laboratory of Dr.

M. Weaver. The water, labelled Milli-Q water, was

55

purified by a system of Milli-pore filters which produces
8

water of conductance of 5.88 x 10' mhos/cm.

Acetonitrile (Matheson, Coleman and Bell or Fisher)

 

Method 1. Acetonitrile was refluxed over calcium hydride
for m2A hours; distilled onto activated molecular sieves;
allowed to stand over the molecular sieves for at least

2A hours; and redistilled from the molecular sieves.

Method 2. (97) Acetonitrile was refluxed over anhydrous
aluminum chloride (15 g/t) for one hour followed by rapid
distillation; refluxed over potassium permanganate (10 g/A)
and lithium carbonate (10 g/t) for fifteen minutes followed
by rapid distillation; refluxed over potassium bisulfate
(15 g/t) for one hour followed by rapid distillation; re-
fluxed over calcium hydride (2 g/t) for one hour followed
by careful fractionation onto activated molecular sieves;
redistilled from the molecular sieves. The middle 80%
fraction was retained. Both methods resulted in a solvent
with a uv cutoff of approximately 210 nm and a water content
of less than 100 ppm. Therefore, Method 1, which is con-

siderably shorter, was subsequently used.

1,2-Dichloroethane (Fisher) was washed with 0.1 N NaOH

 

solution and distilled water; refluxed over calcium chloride
for at least 2A hours (98); distilled onto activated mol-

ecular sieves; allowed to stand over the molecular sieves

56

for at least 2A hours and then redistilled from the

molecular sieves.

Acetone (Drake Brothers) was refluxed over calcium
sulfate for at least 2A hours; distilled onto activated
molecular sieves; allowed to stand over the molecular
sieves for at least 2A hours and then redistilled from

the molecular sieves.

Dimethylsulfoxide (Fisher) was refluxed over calcium

 

hydride for approximately 2A hours under reduced pressure
and distilled onto activated molecular sieves; allowed to
stand over the molecular sieves for at least 2A hours and

then redistilled from the molecular sieves.

Dimethylformamide (Fisher) was vacuum distilled over

 

phosphorus pentoxide onto activated molecular sieves;
allowed to stand over the molecular sieves for at least

2A hours and then redistilled from the molecular sieves.

Acetone—d6, acetonitrile-d3, dimethylsulfoxide-d6
(Stohler Isotope Chemicals) deuterated solvents were used
as received. Live and Chan (38) used acetone—d6 in their
proton nmr studies and found that using this solvent as
it came directly from the manufacturer's sealed vials gave
the same results as when the solvent was carefully dried
with calcium sulfate or over activated molecular sieves.

Molecular sieves (Davison) were 3 A pore size, 8-12

57

mesh, and were activated by heating at approximately 500°C
for at least 2A hours under a flow of dry nitrogen after
being washed with distilled water and dried in an oven at
110°C overnight.

The above drying methods for the solvents produced
solvents with less than 100 ppm water content as measured
by Karl Fisher analysis. Purity of the solvents were
checked by gas chromatography.

Table 6 lists some pertinent properties of the solvents
used in the molar free energy of transfer study. The sol-
vents were chosen because they represent: 1) a wide range
of dielectric constants; 2) solvents whose uv cutoff value
does not interfere with the crown's absorption spectra; and
3) both protic (strong and weak H-bonding characteristics
for water and methanol respectively) and aprotic with
acetonitrile being an anoxic solvent. No solvent represent-
ing an aprotic oxic solvent such as dimethylsulfoxide could
be found which was suitable for study by uv spectroscopy.

Table 7 lists properties of solvents used in the nmr
study of potassium and rubidium complexes of 1806.

The ligands used in the molar free energy of transfer
study were chosen because they contained benzo groups and
could therefore be studied by uv spectroscopy. Melting
points and IR spectra were obtained periodically on the
ligands remaining in the flasks after the solubility

determinations to check for contamination of the ligands

58

.Ammv mocmpowom ADV

.Ammv mocohomom Amv

 

 

 

Es 00H mm.m: m0.H 0.mm 3.05 topaz
EC 0am :m.mH 05.H 5.mm 5.mm Hocmnpoz
E: 0mm H5.0 0H.H 0m.0H 0:03pm
nosoaeoaeum.fl
E: 0am 00.0 mm.m H.3H 0.0m mafigpfiCOpmo¢
newcoao>m3 psoEoz oaoafim pcmEoz ambassz ucmpmcoo pco>aom
mo use >3 pgmpmcoo oHLpooHon maoafio Locoo oagpomaofia

 

.mczopo pospmmaom @500 no

LoamcmgB mo awgocm comm one no 50500 on» CH womb mpcm>aom mo mmfipmoaogm .0 camps

59

Table 7. Pertinent Properties of Solvents Used in the
Proton and Carbon-l3 NMR Study of Potassium and
Rubidium Complexes of 1806 and Magnetic Suscept-
ibility Correction Factors.

 

 

Magnetic
Dielectric Donor Susceptibility
Solvent Constant Numbera Corrections
Acetone 20.7 17.0 -0.5A5
Acetonitrile 38.8 lA.l -0.390
Dimethylformamide 36.7 26.6 -0.308
Dimethylsulfoxide A6.7 29.8 -0.2Al
Methanol 32.7 25.7 -0.A29
Water 78.A 33.0 0

 

(a) Reference (99).
(b) Reference (100).

60

and hydrate formation.

B. METHODS

l. Spectroscopy
Ultraviolet SpectrOSCOpy

All UV measurements were carried out on a Cary 17-909
Spectrophotometer which was converted to a Cary 17 D by
the addition of a digital readout system. Spectra were
obtained in the range of 310 - 200 nm using 0.1, 1.0, and
2.0 cm quartz cells as necessary for Optimum absorbance

measurements.

Nuclear Magnetic Resonance

Proton.NMR spectra were obtained on a Varian T-60
NMR Spectrometer. All chemical shifts were referred
to TMS (tetramethylsilane) which was used as an internal
standard. Carbon-l3 NMR spectra were obtained on a
Varian CTF-20 Spectrometer with a magnetic field of
18.7 kiloguass and resonance frequency of 20 MHz. Samples
in 8 mm NMR tubes were placed inside 10 mm NMR tubes
containing a 50% deuterium oxide-acetone mixture. Deu-
terium oxide served as an external lock with acetone serv-

ing as the secondary external standard. Complete proton

61

decoupling was employed. Chemical shifts were corrected
for bulk diamagnetic susceptibility according to the rela-
tionship of Live and Chan (100) for a nonsuperconducting
spectrometer. Downfield chemical shifts from the free
ligand and TMS are paramagnetic (deshielding) shifts.
Upfield chemical shifts from the free ligand are diamag-

netic (shielding) shifts.

Infrared

Infrared spectra in the range of A000 - 600 cm.1 were

obtained on a Perkin-Elmer A57 Grating Infrared Spectro-
photometer. A standard of polystyrene was used for wave-
length calibration. Nujol mulls were made of the solid

samples and run on sodium chloride mull plates.

2. Other Analyses

 

Solvent Purity

 

The purity of the various solvents was checked by the
use of a Varian Aerograph Model 920 Gas Chromatograph with
a porapak QS 80/100 mesh column and helium carrier gas.
An Omniscribe recorder (Houston Instruments) was used to
record the spectra and a Hamilton 702 NWG microliter syringe

was used for sample injections.

62

Water Analysis

 

Water analyses were performed on a Karl Fisher Photo-

volt Aquatest II Titrimeter.

Melting Point Analysis

Melting point analyses were performed on a Fisher-

Johns melting point apparatus.

3. NMR Data Handling

 

Complex formation constants of the 1806 complexes were
obtained by computer fitting the chemical shift-mole ratio
data using a CD0 6500 computer program system and a non-

linear least squares curve fitting program, KINFIT A (101,
9,50).

A. Solubility Measurements

Experimental Technique and Instrumentation

 

All reagents were purified and dried for each experi-
ment as described above. The ligands were crudely weighed
to ensure that enough ligand was added to each solvent to
produce a saturated solution. The minimum amount to be
added was previously determined by adding ligand from a

weighed amount to a known volume of solvent until saturation

63

was reached and then reweighing the remaining ligand. The
solvents were added to 50 ml Erlenmeyer flasks containing
the ligand in an inert atmosphere (N2) dry box except in
the case of water. The solutions were sealed first with
teflon tape and then with parafilm to prevent moisture
contamination from the atmosphere and the shaker bath
liquid.

The solutions were placed in a Wilkins-Anderson
shaker-bath equipped with a Precision Scientific Micro
Set Thermoregulator which controlled the temperature of
the shaker-bath to within i0.5°0. The temperature was
initially set for 12-2A hours approximately 10°C above the
final equilibrium temperature and then reset to the equilib—
rium temperature in an attempt to reduce the time required
to reach equilibrium. Equilibrium was considered to be
reached after two consecutive concentration measurements
yielded the same results within experimental error. The
shaking of the solutions was stopped at least one hour
before samples were taken for concentration determination.

Samples of the solutions were removed by pipeting
through a medium size pore filter stick attached to the
pipet by a small piece of tygon tubing. A vacuum pump
was employed to provide suction. This procedure was per-
formed as quickly as possible to prevent moisture contamina-
tion. However, as was found later, the solubility of the

ligands studied in water was small enough to be considered

6A

negligible as compared to their solubility in the other
solvents used. The solutions were resealed after removal
of each sample and shaking was resumed. The filtered
samples were then diluted to suitable concentrations for
measurement.

Concentration measurements were made by UV Spectroscopy.
Molar absorptivities of the ligands in the various solvents

were determined by use of Beer's Law (92).

A = Ebc (28)

A = measured absorbance

e = molar absorptivity

b = cell path length (cm)

c = concentration (moles liter-l)
Known concentrations of the ligands were prepared and

their absorbances measured. The molar absorptivity was

determined at each concentration and averaged. This pro-

cedure was repeated until a molar absorptivity was ob-

tained that was reproducible within experimental error.

5. Validity of Method

The use of uv spectroscopy in the determination of the
solubility of different substances is well documented in

the literature (78-81). The system chosen to test the

65

accuracy of the experimental technique used in this study
was that of benzoic acid in water. This system was chosen
because it lends itself to study by other techniques as
well. The solubility of benzoic acid was determined by
the gravimetric and titrometry methods as well as by uv
spectroscopy for comparison.

Saturated solutions of benzoic acid in water at
25 i 0.5°C were obtained by the method previously described.
Three one milliliter samples were taken from each of the
saturated solutions. One sample being diluted and measured
by uv spectroscopy, another being diluted and measured
potentiometrically, and the third sample being placed in
weighed vials which were dried for several days at ap-
proximately 35°C. In the gravimetric method the vials,
after removal from the oven,were dried further by placing
in a vacuum oven overnight and then into a dissicator for
at least a half hour before weighing the resulting residues.
A carbonate free solution of sodium hydroxide,standardized
with a primary standard potassium acid phthalate solution,
was used as the titrant in the potentiometric titration (102).
The molar absorptivity of benzoic acid in water was de-

1

termined to be 876 cm'1 liter' moles at "max of 273 nm and

a pH value of A.0 < pH > 3.0. Literature values of 933
cm‘1 liter-1 moles at pH 1—3 and 79A cm"l liter-1 moles
at pH A.0 were found also at "max of 273 nm (103).

Table 8 shows the results of the determinations at

66

Table 8. Solubility of Benzoic Acid as Determined by UV
Spectroscopy, Titration, and Gravimetric Methods.

 

Solubility of Benzoic Acid (M)

 

 

Sample UV Spectroscopy Titration Gravimetric
0.023 0.026 0.005
2 0.02A 0.02A _____
3 0.023 ----- 0.001
A 0.023 ----- 0.002
5 0.023 0.026 0.003
Blank 0.0 0 0

Average 0.023:0.001 0.025:0.001 0.003:0.002

 

67

25 : 0.5°C. The average values of the determinations
0.023 1 0.001 M and 0.025 t 0.001 M for uv and titration
methods respectively are identical within experimental
error. However, the average value found for the gravi-
metric method of 0.003 t 0.002 is smaller than the other
values obtained by ten fold. The results found in Table 8
and the average value of the solubility for benzoic acid
by the gravimetric method shows both the inaccuracy and
unreproducibility of the method. Literature value for the
solubility of benzoic acid in water is 0.0278 M (10A)
at 25°C.

It can be concluded that the solubility data obtained
by the solubility method utilizing uv spectroscopy should

result in reliable values.

CHAPTER III

RESULTS AND DISCUSSION

68

A. MOLAR FREE ENERGIES OF TRANSFER FROM WATER TO SEVERAL

NONAQUEOUS SOLVENTS FOR DB1806 AND B1505

 

The ultraviolet characteristics and solubilities for
DB1806 and B1505 in various solvents are listed in Table
9. Transfer coefficients and molar free energies of trans-
fer from water, the reference solvent, to several non-
aqueous solvents are given in Table 10. The errors in the
molar absorptivity values were determined by calculation
of the variances of at least three separate determinations.
The large error in the molar absorptivity value of DB1806
in water is due to its low solubility which resulted in
low absorbance values. (m0.1 absorbance units at satura-
tion). This error is carried into the solubility calculated
for DB1806 in water and therefore, into values of the
transfer activity coefficients and molar free energies of
transfer. The error in the solubility values for DB1806
and B1505 in the various solvents was calculated from the
relative standard deviations in the molar absorptivities
and absorbance measurements. The maximum error in the molar
free energies of transfer for DB1806 is estimated to be
:0.5 Kcal mol—l.
The equations used in the calculation of the transfer

activity coefficients and the molar free energies of

69

70

.seq m.smam 56 emeaesmpao A00

.onme mm; acoEogzmon coflpmgpcoocoo pmma opp oEHp ecu Hangs
sump mochm ecu OQCH woomaa who: mngEmm map mEHp 030 spam mam czocm moEHu Esfippfiaflzdm ADV

.moSHm> mocmnpomnm oumfigaogdom o>fiw ow mums mCOszHHQ Amv

 

 

 

mo.osaa.o ages: 02 commas me.oa:e.o seammem ANN mom:
He.oaeo.o ages: 02 0mm a mfi.eaae.o omammmm msm 0mm momfim

moo.oammo.o exams :2 emmfi H moo.osome.o maamemm 05m mooum.a

Hooo.eaoaoo.o axame_mme mmua eoe.oaeom.o Hmfismeam esm mom:

aeo.oamqo.o exams 02 com H seo.oaom:.o emza5eem mum z<
m-eage.aamloaxo.m mxmms :2 aeoz mo.oamo.o meagm.aamoaxm mam 0mm momfimn
oaav amass atonemm haematoaee geesaaatOmee e: as sem>aom enemas

awaoz xme
gpaaahsaom assessaaaem soapsfiae K
.OOm.O H mm pm

mesa>fiom msoase>.ea momfim use somama no maHpAHaesHom ens maaamasapemhaeo pmfioasmepa: .m edema

.pgo>aom monopogom Amv

 

71

 

Hlease seem m.oam.au m.oao.mu mom:

0 H gopMS momam
HIoHoE Hwom m.0nm.:| 5.0AH.0I MOQIN.H
HIoaoe awox m.0hm.:| 0.0wm.5| z<
HImage Heeg m.oaH.mn m.osm.mu mom:

Loam
o H Amv 3 momeQ
26> om go> om

Amzwosvmgos+0mmv W00 Amwooswmcocibmmv a» pco>aom pcmwfiq

 

.msem>aom assasm> oe tape: Ease momam see momfimm
no Loamcmpe mo mofiwpocm comm 0:0 mucoHOHmmmoo 59H>Hpo< hommcmpe .0H ofipme

72

transfer for DB1806 and B1505 are given below.

G°(i) = RT in HE; (1A)
t H O+solvents ‘ S
2 s i
and
wSi
in y = in ———- (15)
t sSi

Table 11 shows a comparison of solvent properties and
free energies of transfer for DB1806 and B1505. The ratio
of the dielectric constant to the dipole moment of the
solvents, e/u, as mentioned earlier, gives a qualitative
measurement of solvent structure (90).

The dielectric constant measures qualitatively the
dipolar orientations of a solvent. The greater the dipolar
orientations of the solvent the more "structure" the sol-
vent is expected to exhibit and consequently, the more dif—
ficult it should be to make a "hole" in the solvent for a
symmetrical nonelectrolyte solute such as DB1806. There-
fore, the solubility of DB1806 would be expected to increase
as the dielectric constant of the solvent decreases. How-
ever, the results show that this is not the case. The
values for the e/p ratio follow the order water > methanol
> acetonitrile > 1,2-Dichloroethane (See Table 12). The

solubilities of DB1806 increases in the reverse order and

73

 

 

 

 

30.03 0.030.31 05.3 5.00 mom:

00.03 0 00.3 3.05 000 00030

35.0 0.030.31 03.3 00.03 00010.3

00.0 0.030.31 00.0 0.00 z<

30.03 0.033.01 05.3 0.00 008:

00.03 0 00.3 3.03 000 000300
:\0 a 9:080: mHOQHQ mao%\30ox : .pcmEQz o .pc0pmcoo ucm>aom 020003

peapaeoo 6300683030 000 m3oa30 6300683030
.mpgm>aom wSOHL0> map on 30303 Eogm W04

.Lm0mc0he mo mmHMLwcm mosh L030: 0g0 mmfippmaopm pcm>aom mo comHL0QEoo .33 03909

Table 12.

7A

Comparison of Solubilities and AG° (H20 +

 

 

t
Solvents) of DB1806 with the e/u Ratio.
. A03 (H2O+Solvents)

Solvent Solubility M Kcal mol-l e/u
H20 3.0x10'5i1.6x10’5 0 A2.38
MeOH 0.0010:0.0001 -2.1:0.5 19.2u
AN 0.0A2:0.00A —u.310.5 9.90
1,2-DCE 0.098:0.003 -u.8:0.5 8.71

 

75

indicates that the less structured the solvent the more
soluble the ligand as expected, but the e/u ratio gives
a better indication of solvent structure than does the
dielectric constant alone.

All values of the molar free energies of transfer
were negative indicating that the ligands are more solvated
by the nonaqueous solvents than by water. The order of the
values for the free energies of transfer for DB1806 parallel
that of the E/U ratio. The lower the e/u ratio the more
negative values of AG: for DB1806 becomes, again indicating
that the less structure there is in the solvent the greater
is its ability to solvate the DB1806 ligand. The magni-
tudes of AG°

t
and -l.2 i 0.2 Kcal mol-

(H2O + MeOH) for DB1806 and B1505 of -2.1 i 0.5
1 respectively, are as expected
because of the smaller size of B1505 which would not be
expected to cause as great of a disturbance to the struc—
ture of the solvent in cavity formation as DB1806 and should
result in a smaller free energy of solvation. Therefore,
the difference in the free energies of solvation for B1505
in water and methanol should be smaller than for DB1806
as the results indicate.

Attainment of equilibrium conditions for the systems
investigated was extremely difficult. The equilibrium
times required varied from approximately 2-1/2 weeks to

6 weeks as measured from the time the samples were prepared

to the time the last concentration determination was made

76

(See Table 1).

When two consecutive concentration determinations
resulted in the same value, equilibrium was considered to
be reached. The length of time required to reach equilibrium
greatly limited the quantity of data acquired. In addition
to the problem of equilibrium times, uv scattering due to
solvent contamination or solute particles and sample con-
tamination during concentration determinations, especially
in the cases of low solubilities, often caused the need
for repeated duplication of measurements. Extreme care had
to be exercised in the cleaning of glassware to prevent
any contamination of the samples and especially contamina-
tion by substances that might complex with the ligands.
Blanks, which were treated in the same manner as the samples,
were checked periodically to aid in the detection of the
possible uv scattering.

In summary, the values of the molar free energies of
transfer from water to the nonaqueous solvents investi-
gated for DB1806 and B1505 indicate that the ligands are
more strongly solvated by the nonaqueous solvents than by
water. Also the parallel order of the values of the
8/u ratio and the molar free energies of transfer suggest
that the ratio, used as a solvent structure gauge, may
be a good qualitative indicator of the order of the free

energies of transfer for the crown ethers.

77

B. PROTON AND CARBON-13 NMR of 1806 COMPLEXES OF SOME

ALKALI METAL SALTS

1. Proton NMR

 

The chemical shifts obtained at 60 MHz from proton
nmr mole ratio studies of 1806 complexed with NaClOu in
deuterated acetone (AC-d6), and with KPF6 in deuterated
solvents of acetone (AC-d6), acetonitrile (AN-d6), and di-
methylsulfoxide (DMSO-d6) are shown in Table 13. A
constant concentration of 0.05 M of 1806 was used and the
concentrations of the salts was varied to yield the dif—
ferent mole ratios shown. Chemical shifts were good to ap-
proximately 12 Hz. The results show that there is virtually
no change in the chemical shift of 1806 on complexation with
either salt in any solvents used. Also samples of un-
complexed 1806 in methanol and in dimethylsulfoxide and
of 1806 in dimethylsulfoxide with a 25 fold excess of
KPF6 were obtained at 180 MHz. The results are shown
in Table 1A. The higher field did produce larger changes
in the chemical shifts. The 29.5A Hz difference in the
chemical shifts for 1806 in methanol and dimethylsulfoxide
reflect the difference in solvation of the ligand in the
different solvents. However, there is only a 2.05 Hz
difference in the chemical shifts between the free and

complexed forms of 1806.

78

Table 13. Proton NMR Chemical Shift-Mole Ratio Data for
1806 Complexation Studies (01806 = 0.05 M;
60 Hz = 1 ppm).

 

 

 

 

 

 

 

 

 

 

Salt: NaClOu Salt:. KPF6

Solvent: AC-d6 Solvent: AC-d6
CNa+ CK+

C1806 5 (:2 Hz) 6 (ppm) C18C6 6 (:2 Hz) 6 (ppm)

0 212 3.53 0 212 3.53

0.31 212 3.53 0.76 212 3.53
0.A7 212 3.53 1.01 21A 3.57
0.63 21A 3.57 1.26 21A 3.57
0.78 212 3.53 1.52 21A 3.57
0.9A 215 3.58 1.77 21A 3.57
1.10 215 3.58 2.53 21A 3.57
1.25 213 3.55

Salt: KPF6 Salt: KPF6

Solvent: AN-d3 Solvent: DMSO-d6

CK+ A CK+

C1806 a (:2 Hz) 6 (ppm) C1806 6 (i2 Hz) 6 (ppm)

0 209 3.A8 0 210 3.50

0.25 209 3.A8 0.25 208 3.A7
0.50 209 3.A8 0.50 207 3.A5
0.75 211 3.52 0.76 208 3.A7
1.00 212 3.53 1.01 210 3.50
1.25 212 3.53 1.52 212 3.53
1.50 211 3.52 2.02 211 3.52
2.00 210 3.50 2.53 210 3.50

 

79

Table 1A. Proton NMR Chemical Shift Data Obtained at 180 MHz
for 1806 and 1806-K+ in MeOH and DMSO.

 

 

Solvent Salt 6 (Hz)
MeOH None 733.15
DMSO None 762.69
DMSO KPF6a 76A.6u

 

a25 Fold excess salt.

Live and Chan (38) observed changes in the chemical
shifts for benzo-substituted crown ethers as big as 89.A Hz
on the complexation of DB30010 with Ba(ClOu)2 in acetone
and 36.2 Hz for the complexation of DB1806 with NaI in
deuterated chloroform. The spectra were obtained at 220 MHz
The higher field used and the presence of the benzo groups
on the crown ethers are the factors responsible for these
large chemical shifts.

Calculation of formation constants for the complexa-
tion of crown ethers with metal cations from nmr-mole ratio
data requires changes in the chemical shifts of the free
and complexed ligand that are large enough to be measured
accurately and have a small standard deviation. Formation
constants for DB1806, B1806, and DB30010 complexed with
various alkali and alkaline earth metal cations were
reported in the Live and Chan study. Formation constants

could not be obtained from the results presented here.

80

It is concluded that proton nmr operated at 60 and 180
MHz does not provide accurate enough data for the quantita—
tive investigation of the complexation of nonbenzo-substi-

tuted crown ethers such as 1806 with alkali metal cations.

2. Carbon—l3 NMR

 

Results of the carbon-l3 nmr chemical shift-mole ratio
studies for the complexation of 1806 and Na+, K+, Rb+,
and EtAN+ cations are shown in Table 15. Chemical shift
data of an 1806 concentration study in DMSO given in Table
16 shows that the chemical shift of the free 1806 ligand
is concentration independent.

In the following discussion all carbon-l3 nmr chemical
shifts are referred to TMS which is upfield (0 ppm). The
more deshielded the nuclei becomes the more downfield the
chemical shifts are and the more shielded the nuclei becomes
the more upfield the chemical shifts. Comparisons are made
to results obtained by Shih (A8) from potassium—39 nmr for
some 1806-K+ complexes in various solvents. Explanation
of the carbon—l3 and potassium-39 nmr conventions for

chemical shift data is shown below.

81

Table 15. Carbon-l3 NMR Chemical Shift-Mole Ratio Data
for 1806 Complexation Studies.

 

 

 

 

 

 

 

 

 

Salt: KPF6 01806=0'°5 M Salt: NaBOu 01806=0'O° M
Solvent: DMSO Solvent: DMSO
CK+ 5 ($0.05 ppm) CNa+ 5 ($0.05 ppm)
91806 C1806
0 71.19 0 71.21
0.23 71.08 0.23 71.11
0.u6 70.99 0.55 70.9A
0.70 70.88 0.69 70.86
0.93 - 70.82 0.83 70.86
1.16 70.8A 0.92 70.82
1.39 70.82 1.15 70.75
1.86 70.83 1.38 70.72
1.61 70.65
Salt: EtuNCLOu Salt: RbI
Solvent: DMSO Cl8o6=0°O6 M Solvent: DMSO 01806=°°05 A
CEtuN+ CRb+
C 6 (10.05 ppm) 0 6 (i0.05 ppm)
1806 1806
0 71.27 0 71.27
0.23 71.27 0.26 71.08
0.A5 71.26 0.A8 71.01
0.68 71.27 0.72 70.88
1.13 71.26 0.96 70.78
1.58 71.21 l.A5 70.73
2.6A 71.19 1.93 70.78
3.52 71.20

 

82

Table 15. Continued.

 

 

 

 

 

 

 

 

 

 

Salt: KPF6 Salt: KI _
_ C -0.05M
Solvent: AC ClBC5-O°O6 M. Solvent: MeOH 1806 —
CK+ CK+
6 (10.05 ppm) 6 (10.05 ppm)
C1806 C1806
0 70.50 0 70.59
0.25 70.38 0.25 70.37
0.50 70.28 0.A9 70.31
0.73 70.13 0.7A 70.2A
1.00 70.05 0.99 70.17
1.22 69.99 1.23 70.16
1.50 70.01 1.A8 70.17
2.0 70.00 1.72 70.17
2.5 69.97 1.97 70.17
Salt: KPF Salt: KI
6 0 =0.05 M 0 =0.05M
Solvent: DMF 1806 - Solvent: H20 1806 _
CK+ < ) CK+ . < >
6 i0.05 m 6 :0.05 ppm
01806 pp C1806
O 70.98 0 70.92
0.23 70.88 0.25 70.96
0-“5 70.75 0.50 70.99
0.69 70.66 0.75 71.03
0.92 70.56 1.00 71.07
1.39 70.52 1.50 71.08
1.85 70.51 2.01 71.13
2.77 70.52 3.01 71.15
3.70 70.50 3.51 71.17
A.01 71.17

 

83

Table 15. Continued.

 

Salt: RbI
Solvent : AN

Salt: RBI

C =0.05 M
Solvent : MeOH 18C6 —

018C6=OOO25M

 

 

 

 

 

C + c +
01806 6 (10.05 ppm) 01806 6 (10.05 ppm)
0 70.59 0 70.73
0.2 70.U6 0.2“ 70.59
0.“ 70.U0 0.00 70.56
0.6 70.29 0.96 70.51
1.0 70.12 0.55 7o.uu
1.“ 70.12 0.80 70.38
2.0 70.13 0.96 70.30
1.2 70.22
Salt: KI

0 =0.05 M
Solvent: DMSO 18C6 —

 

 

 

CK+ 6 (+0.05 ppm)
C18C6 _

0 71.27
0.22 71.13
0.43 71.07
0.65 70.97
0.86 70.89
1.08 70.89
1.51 70.82
1.9“ 70.82
3.02 70.83
3.88 70.80

 

8“

Table 16. 18C6 Concentration Study Data in DMSO Obtained by
Carbon-13 NMR.

 

 

01806 (E) 5 ($0.05 ppm)
0.011 71.26
0.021 71.23
0.0u3 71.23
0.086 71.2u
0.129 71.23

0.21“ 71.23

 

85

7MWS
Carbon—l3 nmr Spectra:

OEwW£ELI €-———-

1 J1
e——. 0er
uwvaaséyv/yum

 

 

Upfield shift results from an increase in electron density
of the carbon nuclei causing a diamagnetic (shielded)
chemical shift.

Downfield shift results from a decrease in electron density

 

of the carbon nuclei causing a paramagnetic (de-

shielded) chemical shift.

Potassium-39 nmr Spectra:

Upfield shift results from a decrease in electron density

 

of the potassium ion nuclei causing a diamagnetic
(shielded) chemical shift.

Downfield shift results from an increase in electron density

 

of the potassium ion nuclei causing a paramagnetic

(deshielded) chemical shift.

The limiting chemical shifts, 611m, which are the

chemical shifts of the complexes, the chemical shifts of

86

the free ligand, 6 (M.R. = 0), and log of the formation
constants, log Kf, calculated by the KINFIT computer pro-
gram, are shown in Table 17. The range of chemical shifts
from the free ligand to the complexed ligand is less than
0.55 ppm in all cases except for 18C6°Rb+ complex in AN

for which the results may be suspect because the mole ratio
values investigated were small due to the low solubility

of the salt. Potassium iodide was used to complex 18C6

in water and methanol because the complex formed with hexa-
fluorophosphate is insoluble and the solubility of the
potassium hexafluorophosphate salt is low in methanol.
Pyridine was a solvent of interest because of its low di-
electric constant and high dipole moment, but salt solu-
bility was too low in this solvent. According to potassium-
39 nmr results of Shih (”8), increasing the concentration of
potassium halide salts produces paramagnetic (deshielded)
chemical shifts in various solvents while an increase in the
concentration of potassium salts with an anion such as PFg
causes diamagnetic (shielded) chemical shifts. DeWitte
(105) found no concentration dependence of the 19F nmr
chemical shift on potassium hexafluorophospate solutions
indicating the absence of ionic association in these solu-
tions. The chemical shift changes indicate ionic associa-
tion of the salt due to the donicity and dielectric constant
of the solvent. As shown in Figures 2 and 3 both anions

I- and PFE gave the same paramagnetic chemical shifts in

87

Table 17. Carbon-l3 NMR Limiting Chemical Shifts and Log

Kf of 18C6-K+ and 18C6-Rb+ Complexes in Various

 

 

 

Solvents.
Salt Solvent aligand (ppm) 611m (ppm) log Kfa
KI H20 70.92 71.23:0.02 1.u10.1
MeOH 70.59 70.15 -------
DMSO 71.27 70.83 _______
KPF6 DMSO 71.19 70.8110.02 .......
DMF 70.98 70.5110.01 ———————
AC 70.50 69.98i0.0U -------
AN 70.73 70.39 -------
RbI MeOH 70.59 70.15 _______
AN 70.73 68.9Ui0.85 l.210.7
DMSO 71.27 70.7u1o,02 .......
H20 70.90 70.88 . m0
(a) Computer generated formation constants resulted in

standard deviations greater than or near the value
calculated for the formation constant in cases of
complexes considered to be strong due to the fact that
their chemical shift-mole ratio plots can be repre-
sented as two intersecting straight lines at M.R. of
one.

88

 

 

 

 

71.30
.1 , 1,.
7:10“ (1')
1
1‘; 1 DMSO
70.90 1" I (Pf-'6') DMSO
~K-
“8”") u u (1')
10.05
70.70
s I DMF
7050‘ 1 (”6.)
' ‘ AN
70.30 —\ (”'6’
L 11.91:
7010— . (I)
AC
‘ (PF-’6’)
6519!) 1 J 1 I 1 l 1 I
0 IO 2.0 30 4.0 5.0
MR: C” (c -005M)
' ° CIBCS lacs" ° -

Figure 2. Carbon-13 chemical shifts 15. K+/18C6 mole ratio
in various solvents.

89

 

 

 

 

71.40
7:00 i MeOH
8(pprn) j __ 0
£0.05 H2
70.60 '— AN Cues-0.025 M
70.20 -—
69.60
)-
69.40 -
P I 101450
69.00 J I 4 I 1 I J I 1 I
0 0.4 0.8 l.2 LG 20
(:FNJI'
MR!= —— (C =0.05M)
CIBCS IBCS -

Figure 3. Carbon-13 chemical shifts 13. Rb+/18C6 mole
ratio in various solvents.

90

in DMSO which suggest that the chemical shift changes
observed are not due to ionic association of the salts.
Shih also found that the chemical shift of the 1806~K+
complex is concentration independent which indicates the
absence of ion pairing between 18C6-K+ and the anion.

Figures 2 and 3 show chemical shift-mole ratio plots
for the complexation of 18C6 with potassium and rubidium
salts in various solvents. Mole ratio plots give an indica-
tion of the relative strength of the complex by the amount
of curvature in the plot. The greater is the curvature the
stronger is the complex. Plots that result in or that can be
represented as the intersection of two straight lines at
mole ratio of one within the standard deviations, are in—
dicative of the formation of a very strong complex. The
KINFIT computer program used in the calculation of the
limiting chemical shifts and formation constants, cannot
be used to calculate formation constants of these very
strong complexes. All plots, except in the cases of l8C6-K+
and 1806-Rb+ in water and 18C6-Rb+ in AN, can be repre-
sented as two intersecting straight lines within the error
bars shown. Consequently, formation constants for these
"strong" complexes could not be determined.

As mentioned before, the interaction of the alkali metal
cation and crown ether is electrostatic in nature. There-
fore, it could be deduced that the free 1806 ligand contains

a partial negative charge on the oxygen atoms and a partial

91

positive charge on the carbon atoms due to the greater
electronegativity of oxygen. On complexation, the electron
density of the carbon should decrease further due to the
interaction of the oxygen atoms with the metal cation caus-
ing a greater deshielding and therefore, a paramagnetic
chemical shift of the carbons in the complex than the car-
bons of the free ligand. However, diamagnetic chemical
shifts were observed in all solvents except 18C6°K+ and
18C6°Rb+ complexes in water. This can be interpreted as

the carbon nuclei of the solvated complexed 18C6 receiving
more electron density than the nuclei of the free solvated
ligand. Truter (58) suggests that when the K+ ion is inside
the cavity there is a repulsive interaction between the K+
ion and the oxygen atoms on the crown ether causing a de-
shielding of the K+ ion and subsequently causing an increase
in electron density of the carbon atoms. In the case of
18C6°K+ complex in water, hydrogen bonding between the oxygen
atoms of the ligand and water decreases the electron density
of the carbon nuclei and cause the paramagnetic shift
observed. As for the horizontal straight line observed

for the 1806°Rb+ complex, which indicates no complex forma-
tion, hydrogen bonding and the repulsive forces mentioned
above could just cancel each other. Another explanation

may be that the hydration of Rb+ ion may be strong enough

to prevent the formation of the complex. However, K+

ion, whose charge density is greater than that of Rb+

92

ion, is more hydrated and in this case, prevention of inter-
action between the ligand and K+ ion should be even greater,
but the chemical shift data shows the opposite to be true.
Izatt (106) found log Kf for 18C6°Rb+.in H2O of 1.56 by
calorimetry. Lichter and Roberts (107) suggests that re-
pulsion of the C-H bonding electrons towards the carbon
nucleus results in a shortening of the average radius of
that orbital, thus increasing the average of the inverse
cube of the effective orbital radius term of the Ramsey
equation (9H), <1/r3>, and inducing a paramagnetic shift in
the carbon-l3 nmr spectra.

Shih (“8) found potassium—39 paramagnetic shifts for
the 18C6°K+ complex in water and DMSO and diamagnetic shifts
in AC and DMF. These results indicate that DMSO and water
are better electron donors than the ligand and that the
ligand is a better donor than AC and DMF. Carbon—l3 nmr
data show that the complexation of the 1806 ligand with K+
and Rb+ ions all results in an increase in electron density
of the carbon atoms, except perhaps the 18C6'Rb+ complex in
H2O.

Table 18 shows the results for different 18C6-cation
complexes in DMSO. The ionic diameter of the cations (ex-
cept for EtuN+ which is assumed to be the largest) is
listed to provide a comparison with the limiting chemical
shifts for the complexes. The limiting chemical shift in

ppm follows the order of the ionic diameters, the larger

93

Table 18. Comparison of Carbon—13 NMR Limiting Chemical
Shifts of Various Salt Complexes of 18C6 in DMSO
with Ionic Diameter of the Cations.

 

 

Salt Diiggigra '6 (ppm)
lim
NaBOu 2.31 69.07
KPF6 2.98 70.8110.02
KI 2.98 70.83
RbI 3.28 70.7110.02
EtuNClOu -—--b 71.20C

 

(a) Reference (62).
(b) Diameter assumed to be greater than other cations.

(c) No complexation takes place.

9”

the diameter the smaller the diamagnetic shift from the free
ligand. The horizontal straight line resulting from the
interaction of EtuN+ and 18C6 (See Figure 1) indicates

that no complex is formed because the EtuN+ ion is too

large for the cavity of 1806. These results indicate the
importance of the size and therefore the charge density of
an ion in the determination of the cation—ligand inter-
action.

Tables 19 and 20 compare the solvent properties with
the limiting chemical shifts of 1806-K+ and 1806-Rb+ com-
plexes respectively. No linear correlations could be
determined from the results.

Formation constants for some of the 1806-K+ and 18C6-Rb+
complexes in various solvents are shown in Table 21. Table
22 shows a comparison of formation constants for the 1806°K+
complex obtained by carbon-l3 nmr and potassium—39 nmr (U8).
A strong cation-ligand interaction is expected between K+
cation and 18C6 because the diameter of K+ is Just right to
fit inside the 1806 cavity; however, solvents of high
donicity will highly solvate the K+ ion and thus prevent
a strong interaction. AC with the lowest donicity would
be expected to support the strongest interaction between
18C6 and the K+ ion. The order of formation constants

would be expected to be AC > DMF > DMSO > H20. Formation

constants determined from potassium-39 nmr data follows

this order but, those obtained by carbon-l3 do not reflect

95

 

 

7:.40
6341:0104
7:.204 i T
slim 7|.OO _' ' :\
$0.05 \.
‘\
I .
70.80 - 3;“- 23:6
:L b
N08
70.60 — 4’4
J J J J J

 

 

O 0.5 LO l.5 2.0 2.5

_ Ccoflon _

Figure H. Carbon—13 chemical shifts gs. cation/18C6 mole
ratio in DMSO.

96

Table 19. Comparison of the Carbon-13 NMR Limiting Chemical
Shifts of 1806-K+ Complexes and Solvent Properties.

 

 

Dielectric Donor

Salt Solvent Constant No. 611m (ppm)
KPF6 AC 20.7 17.0 69.9810.0U1

AN 38.8 IN.1 70.39

DMF 36.7 26.6 70.51i0.01

DMSO 46.7 28.9 70.8110.02
KI DMSO 96.7 28.9 70.83

MeOH 32.7 25.7 71.15

H20 78.5 33 71.23:0.02

 

97

Table 20. Comparison of the Car on-l3 NMR Limiting Chemical
Shifts of the 18C6-Rb Complexa and Solvent
Properties.

 

 

Dielectric Donor
Solvent Constant No. 6lim (ppm)
MeOH 32.7 25.7 70.12
AN 38.8 11.1 68.9410.85
DMSO “6.7 28.9 70.7110.02
H20 78.5 33 70.88

 

(a) RbI used.

Table 21. Formation Constants of 18C6-K+ and 18C6-Rb

98

+

Complexes in Various Solvents.

 

 

Salt Solvent Log Kf
KI H2O l.UO10.10
MeOH _________
DMSO _________
KPF6 DMSO _________
DMF m3
AC m3
AN ---------
Rbl MeOH _________
AN 1 2H 0.65
H2O mo

DMSO

 

99

Table 22. Comparison of Formation Constants Determined
From 130 and 39K NMR for 1806-K+ Complex8 in

Various Solvents.

 

l3C

 

Solvent Log Kf— NMR Log Kf- NMR
Ac m3 >0
DMF m3 2.70i0.01
DMSO --- 2.1910.23
H2O 1.1010.10 2.1710.13C

 

(a) KPF6 used except KI used in H2O.
(b) Reference (U8).

(c) Log Kf of 2.03 was found by Izatt (Reference 106).

100

this order as well. Alkali metal nmr chemical shift
changes reflect the competition between the crown ether
and the solvent for the metal cation while those of carbon-
13 nmr reflects the competition between the solvent and
cation for the ligand and is therefore at best, only an
indirect or qualitative indication of the former competi—
tion. An example of a formation constant determined by
the computer KINFIT program for a complex whose chemical
shift-mole ratio plot can be represented as two intersect-
ing straight lines is “123 1 3420 or log Kf = 3.61 1 3.53
for the 18C6°K+ complex with DMF. Lin (51) also reports
that log Kf for this complex in DMF could be greater than
three from carbon-l3 NMR data. Log Kf for 18C6'K+ in DMF
is reported as 2.70 1 0.0” from potassium-39 nmr data (See
Table 22). However, close agreement between the formation
constants obtained from carbon-l3 and potassium-39 nmr
data for the 12CL|°K+ complex was found (“8). Cation-
ligand interaction is expected to be weak in this complex
because the diameter of the K+ ion is too large for the
cavity of the 12C“ ligand. Values of log Kf for the
12C14°K+ complex obtained by potassium-39 nmr were 1.79

1 0.18, 218 1 0.16, and 0.31 1 0.0M in AC, AN, and DMSO
respectively. Values of log Kf obtained by carbon—l3 nmr
were 1.87 1 0.07, 2.26 1 0.07, and 0.67 1 0.1“ in AC, AN,
and DMSO respectively.

The results indicate that the carbon-13 nmr method is

101

unable to differentiate the relative strengths of complexes
that are expected to display strong cation-ligand inter-
actions. However, formation constants obtained for complexes
expected to have weak cation-ligand interactions agree well
with those obtained by alkali metal nmr as shown above.

The minimum value of the formation constant of a complex
for which carbon—l3 nmr data can be expected to provide an
accurate value as compared to values obtained by alkali
metal nmr has not been determined. It is concluded that
carbon-l3 nmr is not a sensitive enough probe for quanti-
tative determinations of complexation strength. An
increase in field strength to provide larger chemical shift
with less error may increase the sensitivity and useful-

ness of this method.

Future Work

 

This study only touches the surface in the investiga-
tion of transfer activity coefficients for the crown ethers.
Much more data are needed on the solubilities of the crown
ethers in order to continue. Transfer activity coefficients
for the crown ether complexes in various solvents will also
give valuable information concerning their solvation in
the solvents. However, a suitable method for the determina—
tion of the solubilities of the complexes must be found.

Proton nmr does not seem promising as a method of

102

investigating the complexation reactions of the crown
ethers except at high field strength. ~Such studies should

continue to be investigated by use of metal nuclei nmr.

LIST OF REFERENCES

10.

11.

12.

13.

l“.

15.

\lmmzw

REFERENCES

C. J. Pedersen, J. Am. Chem. Soc., 89, 7017, (1967).

I. Kolthof, Anal. Chem., 9;, 1R, (1979).

 

C. J. Pedersen, J. Am. Chem. Soc., 99, 2195, (1967).

 

Pedersen, J. Am. Chem. Soc., 99, 386, (1970).

 

Pedersen, J. Am. Chem. Soc., 99, 391, (1970).

 

 

J.
J.

C. J. Pedersen, J. Am. Chem. Soc., 99, 251 (1971).
J.

C. Pedersen, H. K. Frensdorff, Angew. Chem. Int.
Ed. English, 11’ 16 (1972).

 

 

C. J. Pedersen, Ch. 1 in Synthetic Multidentate
Macrocyclic Compounds, R. M. Izatt and J. J. Christen-

 

 

sen Eds., Academic Press, New York, NY (1978).

Fred Smetana, Ph.D. Thesis, Michigan State University,
East Lansing, MI, 1979.

J. S. Bradshaw, Ch. 2 in Synthetic Multidentate
Macrocyclic Compounds, R.7M.JIZatt and J. J. Christen-
sen Eds., Academic Press, New York, NY (1978).

 

 

D. N. Reinhoudt and F. DeJong, Ch. N, in Progress in
Macrocyclic Chemistry, Vol. 1, Ed. R. M. IEatt and
J. J. Christensen, John Wiley and Sons, New York,

NY (1979).

R. M. Izatt, J. D. Lamb, G. E. Maas, R. E. Asay,
J. S. Bradshaw, J. J. Christensen, J. Am. Chem. Soc.,

99, 2365 (1977).

L. Y. Martin, L. J. DeHayes, L. J. Zompa, D. H. Busch,
J. Am. Chem. Soc., 99, ”096 (197“).

 

 

 

R. M. Izatt, D. P. Nelson, J. H. Rytting, B. L.
Haymore, and J. J. Christensen, J. Am. Chem. Soc.,

23, 1619 (1971).

R. M. Izatt, N. E. Izatt, B. E. Rossiter, J. J.
Christensen, Science, 199, 99“ (1978).

 

103

16.

17.
18.

19.

20.

21.

22.

23.

2M.

25.

26.

27.

28.

29.

30.

31.

105

R. S. Cockrell, E. J. Harris, and B. C. Pressman,
Biochemistry, 9, 2326 (1966).

 

D. E. Fenton, Chem. Soc. Rev., 9, 325 (1977).

 

W. E. Morf, D. Ammann, R. Bissig, E. Pretsch, and

W. Simon, Ch. 1 in Progress in Macrocyclic Chemistry,
Vol. 1, Ed. R. M. Izatt and J. J. Christensen,

John Wiley and Sons, New York, NY (1979).

 

J. J. Grimaldi and J. M. Lehn, J. Am. Chem. Soc.,
101, 1333 (1979).

B. Dietrich, J.—M. Lehn, and J.—P. Sauvage, Tetra-
hedron Lett., 1969, 2885 and 2889.

 

 

E. Blasius, K. P. Janzen, H. Luxenburger, V. B.
Nguyen, H. Klotz, and J. Stockemer, J. of Chromatog-
raphx. 167. 307 (1978).

 

K. Kimura, T. Maeda, and T. Shono, Anal. Letters.,
All, 821 (1978).

 

Y. Takeda and H. Kato, Bull. Chem. Soc. Japan, 99,
1027 (1979).

K. J. Odell, E. M. Hyde, B. L. Shaw, and I. Shepherd,
J. Organometallic Chem., 168, 103 (1979).

 

 

K. Kimura, T. Maeda, H. Tamura and T. Shono, J.
Electroanal. Chem. Interfacial Electrdchem., 99,

91711979).

H. Sumujoshi, T. Tachibana and N. Keiji, Jpn. Kokai
Tokkyo Koho, Dojindo Laboratories (Chem. Abst. 99:
148113t).

 

R. M. Izatt and J. J. Christensen, Ed., Synthetic
Multidentate Macrocyclic Compounds, Academic Press,
NewFYork, NY (19787.

 

 

M. Kiritani, Jpn. Kokai Tokkyo Koho 79.31,305 (Chem.
Abst. 91:66350). “"

J. L. Dye, Angew. Chem. Int. Ed. Engl., 12’ 587
(1979).

J. L. Dye, J. M. Ceraso, M. T. Lok, B. L. Barnett
and F. J. Tehan, J. Am. Chem. Soc., 99, 608 (197U).

 

 

A. I. Popov, Pure and Appl. Chem., 91, 101 (1979).

 

32.

33.

3H.

35.

36.

37.

38.

39-

NO.

91.

H2.

#3.
NH.

“5.

H6.

A7.

106

E. Shchori, J. Jagur-Grodzinski, Z. Luz, and M.
Shporer, J. Am. Chem. Soc., 99, 7133 (1971).

E. Shchori, J. Jagur-Grodzinski, and M. Shporer,
J. Am. Chem. Soc., 99, 3392, (1973).

 

E. Mei, A. I. Popov, and J. L. Dye, J. Phys. Chem.,
g1, 1677 (1977).

E. Mei, L. Liu, J. L. Dye, and A. I. Popov, 1.
Solution Chem., 9, 771 (1977).

 

 

M. Shamsipur and A. I. Popov, J. Am. Chem. Soc.,
101, 1051 (1979).

 

J. L. Dye, Ch. 2, in Progress in Macrocyclic Chem-
istry Vol. 1, R. M. Izatt and J. J. Christensen, Eds.,
John Wiley and Sons, New York, NY (1979).

 

D. Live and S. Chan, J. Am. Chem. Soc., 9g, 3769,
(1976).

J. C. Lockhart, A. C. Robson, M. E. Thompson, P. D.
Tyson, and I. H. M. Wallace, J. Chem. Soc. Dalton
Trans. 1978, 611.

 

 

 

J. J. Christensen, D. J. Eatough, and R. M. Izatt,
Chem. Rev., 19, 351 (197“).

 

J. D. Lamb, R. M. Izatt, J. J. Christensen, and D. J.
Eatough,"Thermodynamics and Kinetics of Cation-
Macrocycle Interaction", in Chemistry of Macrocyclic
Compounds, G. A. Melson, Eds, Plenum, NewFYork, NY

T1979T_——

R. M. Izatt and J. J. Christensen, Eds., Progress in
Macrocyclic Chemistry, Wiley Interscience, New York,

NY (1979).
H. K. Frensdorff, J. Am. Chem. Soc., 99, 600 (1971).

 

 

 

 

E. Mei, Ph. D. Dissertation, Michigan State University,
E. Lansing, MI (1977).

A. Hourdakis, Ph.D. Dissertation, Michigan State Univer-
sity, E. Lansing, MI (1978).

J. G. Hoogerheide, Ph.D. Dissertation, Michigan State
University, E. Lansing, MI (1978).

P.-H. Heubel, Ph.D. Dissertation, Michigan State
University, E. Lansing, MI (1978).

MIIHIIIIIIHIIIIIIIIIIUIIIIIEIISIIIIIIIIIIIIIR'ES
31193 03143 2309