Rochester Institute of TechnologyRIT Scholar Works

Theses Thesis/Dissertation Collections

3-1-1968

A Study of Hydrogen BondingMary Ann Conklin

Follow this and additional works at: http://scholarworks.rit.edu/theses

This Thesis is brought to you for free and open access by the Thesis/Dissertation Collections at RIT Scholar Works. It has been accepted for inclusionin Theses by an authorized administrator of RIT Scholar Works. For more information, please contact ritscholarworks@rit.edu.

Recommended CitationConklin, Mary Ann, "A Study of Hydrogen Bonding" (1968). Thesis. Rochester Institute of Technology. Accessed from

A STUDY OF HYDROGEN BONDI NG

MARY ANN CONKLIN

MARCH, 1968

THESIS

SU BM I TTED I N PART I AL FULF I LU~ENT OF THE

REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE

APPROVED:

Earl Krakower Project Advisor

Robert L. Craven Sta ff Cha j rman

T. E. Strader Li brary

Rochester Institute of Technology Rochester, New York

Acknowledgements

The author is grateful to the Institute

faculty for accepting her in their Master's program

and for giving her valuable training during her

residence at Rochester Institute of Technology.

The author wishes to express her esteem of

Dr. Earl Krakower and her appreciation of his efforts

in making bhis work come to its conclusion. Much of

the credit for this thesis must be given to his

dedication to productive and fine research, his thor

oughness and conscienciousness in scientific endeavor,

and his ability to communicate his purpose in the

research project to the author.

The author also wishes to thank Dr. Louis Daignault

for the mass spectrometric analyses. She is grateful for

all he taught her and values her association with him.

Thanks are extended to Frederick Delles who aided

the author particularly in the early stages of the work;

and to Mitchell Bogdanowicz for his help in the gas

chromotography experiments.

Dr. Harry Agahigian is also acknowledged for the

very fine NMR training he gave the author prior to the

commencement of this research.

This dedication, to Robert Frost, is .madesince the

poet well expresses the author's feelings in his''Reluctance"

and his "Road Not

ABSTRACT

The proton magnetic resonance technique lends itself to the

study of hydrogen bond equilibria. Through PMR spectra, hydrogen

bonding can be detected. The specific proton which participates

intimately in the hydrogen bond may be identified.

The following work has investigated hydrogen bonding systems

in which the proton donor is trifluoroacetic acid and the proton

acceptor is an N-heterocycl ic base, such as quinazoline, quinoline,*

and isoquinol ine.

From measurements of the chemical shift versus concentration, and

from measurements of the line width at half peak amplitude versus

concentration for the time average exchange signal, the extent of

hydrogen bonding may be determined and the participating equilibria

may be characterized. Two equilibrium constants have been calculated

using the PMR data obtained in this study.

Solvent effects upon the hydrogen bonding equilibria have been

noted in several cases. These effects were similar to those found in

otheV previous investigations and can be explained.

The effect of hydrogen bonding upon the N-heterocycl ic ring has

been observed. The ring protons are chemically shifted upon hydrogen

bond association of the N-heterocycle with the acid. This observation

reveals a charge redistribution in the heterocycle upon complex

formation. There is insufficient data to calculate the charge densities

of these systems, however.

TABLE OF CONTENTS

IV. APPENDIXES ..

V. BIBLIOGRAPHY

Page

I . I NTRODUCT I ON

A. The Hydrogen Bond I

B. Magnetic Properties of Nuclei 7

C. NMR Applications to Hydrogen Bonding 18

D. Charge Density Considerations 30

II. EXPERIMENTATION

A. Synthesis and Preparation Techniques 36

i. Quinazoline 36

ii. Quinoline 39

iii. Isoquinoline 39

iv. Trifluoroacetic Acid 40

v. Chloroform 40

vi. Acid-base mixtures 40

B. Instrumentation 41

I I. RESULTS AND DISCUSSION

A. Acid-base Systems. Two Component Systems 45

B. Solvent Effect

i. Three component system 56

i i . Two component system

a. Dimethy I su I foxide-tri f I uoroactic Acid 60

b. Chloroform- isoqu inol i ne 61

c. Chemical Shift of N-heterocycl ic Ring Protons 63

List of Illustrations

Figure Title to follow

page

I Preparation of Quinazoline 37

II Block diagram R-20 43

III Trifluoroacetic acid spectrum 45

IV Chemical Shift vs Concentration

Trifluoroacetic-Carbon Tetrachloride 45

V Quinazoline spectrum . 46

VI Quinoline spectrum 46

VII Isoquinoline spectrum 47

VIII Chemical Shift vs Concentration

Quinazoline-Trifluoroacetic acid 47

IX Chemical Shift vs Concentration

Quinoline-Trifluoroacetic acid 47

X Chemical Shift vs Concentration

Isoquinoline-Trifluoroacetic acid 47

XI Illustration of line width vs

concentration 53

XII Width vs Concentration

Quinoline-Trifluoroacetic acid 53

XIII Width vs Concentration

Isoquinoline-Trifluoroacetic acid 53

XIV Chemical Shift vs Concentration

three component system 59

XV Chemical Shift vs Concentration

comparison plot for the two and

three component systems 59

XVI Chemical Shift vs Concentration

Trifluoroacetic -Dimethyl sulfoxide 60

XVII Chemical Shift vs Concentration

Isoquinoline-Chloroform 62

Figure Title to follow

page

XVIII N-hetero ring protons

Chemical. Shift vs Concentration

Quinazoline-Trifluoroacetic acid 63

IXX N-hetero ring protons

Chemical Shift vs Concentration

Quinoline-Trifluoroacetic acid 65

XX N-hetero ring protons

Chemical Shift vs Concentration

Isoquinoline-Trifluoroacetic acid 65

XXI N-hetero ring protons

Chemical Shift vs Concentration

three component system 65

INTRODUCTION

The Hydrogen Bond

The formation of a hydrogen bond results from an interaction

between a donor molecule, A-H, and an acceptor group, B. This inter

action usually involves two functional groups: the acidic (e.g.,car-

boxyl or hydroxyl group) and the basic (e.g., nitrogen in N-heterocycl es)

The proton donor, A-H, and the proton acceptor, B, molecules have

strongly electronegative atoms between which the hydrogen atom lies.

A stable hydrogen bond will only be formed if the charge distribution

of the A-H bond orbital is such that the proton is sufficiently un

screened. The molecules associate rapidly and reversibly to form

molecular aggregates known as complexes. The hydrogen lies between

the two electronegative atoms and shares an electron pair with each (30).

This association is written as:

A-H + B * A-H...B (C)

The bond formed is relatively weak; usual bond energies are in the 2-15

Kcal/mole range (17,31). For example, the dimer of trifluoroacetic acid

gas has a bond energy 6.85 0.2 Kcal per hydrogen bond per mole;

aniline in solution has a bond energy of 1.93 Kcal/mole while methanol

in carbon tetrachloride has a bond energy 13 Kcal/mole (13).

It is difficult to suggest a universally acceptable definition of

a hydrogen bond. However, Pimental and McClellan offer their criteria

for hydrogen bond formation:

"A H-bond exists between a functional group, A-H, and an atom

or group of atoms, B, in the same or different molecule when

(a) there is evidence of bond formation (association or

chelation); and (b) there is evidence that this new bond

linking A-H and B specifically involves the hydrogen atom

already bonded toA."

(54)

2

There are three theories of hydrogen bonding (36): the electro

static approach (52); valence-bond approach (17); and the molecular-

orbital approach (45).

The Is configuration of hydrogen (with the 2s and 2p orbitals

having too high energies) allows hydrogen to form only one covalent

bond (17,31).

The electrostatic approach considers the case of two dipoles and

their attraction.

-A-i+H &S~3t.

The closer B approaches H the stronger the electrostatic link between

them is. Some pairs of electrons, such as those found in N-heterocycl ics,

often determine the strength and direction of hydrogen bonds. This

approach has been supported by studies about solvent effects of H

bonding (48). The electrostatic approach is supported theoretically by

quantum mechanical rules (52). Another supporting evidence is that

hydrogen bonds are formed by fluorine; e.g., HF and specifically a

symmetrical H-bond is formed in HF . There is a possibility that the

nature of the actual hydrogen bond lies somewhere between the electron-

pair structure and the ionic structure. The electrostatic approach

provides an explanation as to the fact that only electronegative atoms

form such bonds (31).

The valence bond approach presents possible structures contributing

to the hydrogen bond, such as

covalent A-H bond

ionic A-H bond

covalent H-B bond

ci7;

(1) A-H B

(2) A H . . . . B

(3) A"HB+

(4) AH . . . B

3

Proponents of this approach indicate that the covalent contribution

may be significant (17,18). This treatment has received experimental

support from studies on the proton accepting power of a series of

compounds (49). The order of proton accepting power or hydrogen bonding

strength is the reverse of the order of magnitude of electrostatic

attraction. It was suggested that, for atoms having non-bonding or

lone pair electrons, a possible contribution is made by the charge

transfer force arising from stabilization due to electron migration

from a base to an anti-bonding hydroxyl orbital f an acid (48). The

proton donor and proton acceptor act as electron acceptor and electron

donor. Further work in the vacuum ultraviolet region was suggested

for examination of a possible resonance between structures A-H B

and A H B which if found would support this approach.

The molecular orbital approach (53) considers the molecular orbital

on either side of the hydrogen atom. The molecular orbital contains

two electrons by which the hydrogen is weakly bonded to both the

electronegative atoms, A and B. The electron pair in the A-H bond

are tightly drawn to the electronegative atom. This bonding electron

pair is pulled from the first valence shell into the second valence

shel I of hydrogen. The hydrogen then may form a weak bond with some

electron-pair donor such as nitrogen (44).

Certain conditions are necessary in the potentially hydrogen

bonded system for the association. They include a geometrical config

uration which allows H bonding and an electron affinity of the atoms

sharing the hydrogen. How our choice of compounds would fit these

criteria will be discussed later.

4

The particular arrangement of the three atomic centers involved

in the hydrogen bond is not yet established. The linearity of A-H...B

is said to be energetically favored (54), but work revealing that cyclic

dimers are often formed by hydrogen bonding molecules supports a non

linear arrangement of the three atomic centers. One example is the

dimer of acetic acid which is known to be self-associating.

H .0...H-C1 H

I // \ I

H-C-C C-C-H

\ SH

0-H...0'H

There are two types of hydrogen bond. The first, and most common,

is the intermolecular hydrogen bond in which the groups, A-H and B,

are in different molecules. Intermolecular hydrogen bonds can produce

polymers; that is, the complexes are not restricted to dimer size.

An example of intermolecular hydrogen bonding is given by trifluoro

acetic acid and quinoline.

0. .0

Xc^

tF-C-F

or by trifluoroacetic acid itself.

F O...H-0. F

I sf \ \

F-C-C C-C-F

I \ #F 0-H...0 F

The other class of hydrogen bonding is intramolecular, where the

groups A-H and B are part of the same molecule. Intramolecular hydro

gen bonding is also called chelation and is commonly found in ortho-

5

compounds where five-, six-, or seven-membered rings are formed by this

chelation. For example, 5,8-d ihydroxy-a-naphthoquinone is known to

Form intramolecular hydrogen bonds:

In contrast to intermolecular hydrogen bonding, the species

resulting from intramolecular hydrogen bonding does not have an

increased molecular weight, increased melting or boiling point, nor

decreased vapor pressure.

It shall be shown in subsequent discussion that certain properties

of intermolecularly hydrogen bonded systems are remarkably concentra

tion dependent. Such concentration dependence either does not exist

or is very small for intramolecu larly bonded systems.

Association in hydrogen bonding is a rapid and reversible process.

For a two component system there should be several equilibria operating

at once. Possible processes include:

A-H + B - AH . . . B

A-H. . . B

->

A . . .H B

A . . . H B->

+ H +BH+

B + B B + H +B

A-H + f- + H-A 43

The formation of a hydrogen bond is sensitive to several variables;

e.g., temperature, concentration, and solvent (if the hydrogen bond

system is in solution). Often the presence of hydrogen bonding in a

6

system is detected by the sensitivity of the acid-base system to these

(variables.

The hydrogen bond can alter the mass, shape, atomic and electronic

arrangements of the molecule. Hence, hydrogen bond formation modifies

!both physical and chemical properties. It is this feature that makes

the following work possible.

Proton magnetic resonance spectroscopy. PMR, is a useful technique

which is applicable to the study of hydrogen bonding. PMR directly

indicates the role of the hydrogen atom in hydrogen bond formation

and supplies information about the electronic charge distribution of

the hydrogen bond system (55).

Magnetic Properties of Nuclei

The PMR technique involves radiof requency-induced transitions

between energy states of a nucleus polarized by a magnetic field. The

nucleus giving rise to the PMR spectrum possesses angular momentum

and spin.

The maximum measurable vector of the angular momentum of the

nucleus must be, according to quantum mechanical rules, an integral or

half-integral of h/2ir or tf; h being Planck's constant, tf is called the

modified Planck's constant. This maximum measurable vector of the

angular momentum, the spin quantum number, is designated by I. If

1=0, the nucleus does not give rise to a NMR signal. The nucleus may

have (2 1+ I) states; the permitted values of the moment along any

one direction may be described by the magnetic quantum number, m.

m = I, (I - I), (1 -

2), ... (-T + I), - I.

The magnetic moment of the nucleus is parallel to the spin

quantum number. When the spin quantum number is zero, the magnetic

number is zero. When I has a finite value, the magnetic vector, u, is

described by my/ 1.

8

The magnetic properties of the nucleus may be described by a

ratio since the vectors y and I are parallel.

h"

P= J-(I'K) = /(I %-) Equation I

}f is the magnetogyric ratio.

The spin of the nucleus is related to the electrical quadruple

moment, which describes the distribution of electric charge about the

nucleus. The electrical quadruple moment, 0, measures the non-sphericity

of the electric charge distribution.

Only nuclei with I > I possess electrical quadruple moments; that

is, hydrogen H having 1 = { does not possess an electrical quadruple

moment so that PMR investigations should not involve any direct

interaction of the spin with the electric charge distribution. However,

14nitrogen N possessing a spin 1=1, has an electric quadruple moment,

Q = 2 x I0"2.

The nucleus in a homogeneous external field H with the magnetic

moment vector u in the z direction possesses energy-

yH . y may

have (2 1+ I) distinct, equally-spaced states which are separated by

The nucleus which absorbs radiof requency quanta will undergo

transitions between energy levels expressed by different magnetic

quantum numbers; e.g., m = 0 and m =-{ if 1 = i for the nucleus. From

the Bohr relation E = hv where E is the energy of the transition, h is

the Planck constant, and v is the frequency of the radiation absorbed,

since the neighboring energy levels are separated by y H / 1.

hv = yH / 1 Equation 2

pHov

= tt~ = Ht*t Equation 3

Ih o JhM

This frequency may also be expressed in terms of the magnetogyric ratio.

^ =

IH=

24mT)Equation 4

/Ho2tt

Equation 5

v is the frequency of the radiation required for the nucleus in the

magnetic field H to undergo the transition between neighboring levels.

If the environment of the nucleus is ignored, the transition

occurs at a characteristic frequency which is pnoportional only to the

external applied magnetic field, H,and the magnetogyric ratio, o .

The magnetic environment of the nucleus, however, may be effected

by neighboring nuclei and electrons. In solids, where molecular

translation and rotation are inhibited, the magnetic moments of

neighboring nuclei alter the magnetic field felt by the nucleus in

question. The strength of the field due to a neighboring nucleus at

distance R is within + 2y/R and- 2y/R ; the spectral signal of the

nucleus in question would be broadened since the resonance condition

occurs over a range of frequencies. In liquids and gas samples, however,

where molecules translate and rotate freely, this magnetic dipole

broadening cancels out to zero. Due to rapid molecular motion, there

fore, this dipole-dipole interaction may be neglected.

Electrons about the nucleus in question also alter the magnetic

field felt by the nucleus. The magnetic field induces an orbital

motion thereby setting up currents within the molecule and a secondary

magnetic field, each of which is proportional to the external field.

The molecule acquires a diamagnetic moment because of this induced

10

motion of its electrons. The field felt by the nucleus is proportional

to and slightly smaller than the external field.

H. = H ( I -

cr-) Equation 6local o

The parameter, <^, is the screening constant and is dependent upon

the electronic environment of the nucleus. Non-equivalent nuclei

experience different diamagnetic electron screening and, as a result,

give rise to different resonance frequencies. This change in the

resonance frequency is called the chemical shift, ?. Each chemically

distinct nuclei exhibits a characteristic chemical shift. The chemical

shift is proportional to the external magnetic field, H,and may be

expressed in gauss, or hertz (where one hertz is equal to one cycle

per second. )

The resonance frequency of the nucleus is also effected by inter

actions known as electron-coupled spin interactions or spin-spin

coupling. This interaction between nuclear spins of non-equivalent

nuclei occurs through the bonding electrons; it is independent of the

external magnetic field. The energy of the interaction is proportional

to the product of nuclear spin-vectors. - The proportionality constant,

J, is known as the spin-spin coupling constant.

If the interaction of two non-equivalent nuclei is considered, one

nucleus via the bonding electrons feels the spin orientations of the

second nucleus corresponding to (2 1 + I). If for the second nucleus

I = i then the first nucleus sees 2{{) + I or 2 orientations and its

resonance frequency signal will be split into a doublet, or into two.

This multiplicity of the PMR signal of equivalent nuclei is thus

determined by the neighboring group of equivalent nuclei. The neighboring

1 1

group because of spin-spin coupling splits the peak of the nuclei in

question into (2n 1+1) multiplets, where I is the spin quantum number

of the neighboring nuclei and n is the number of nuclei on the neigh

boring group.

In summary of the foregoing discussion, the magnetic nucleus, i,

giving rise to the PMR absorption signal undergoes transitions between

energy levels. The Hamiltonian operator, ~)\ ,describes these levels

and is given by

X= ^f I Y. H (I -o-.) I(i) Equation 7

i

orJ^

=j-; I Y. H. T(i) Equation 8

i

The Hamiltonian operator, 2 , describing the spin-spin coupling between

nuclei, i and j, is given by

J( =.f. J.. I(i) I(j) Equation 9J J

So long as dipole-dipole interaction is neglected (which it may be if

molecular motion is high and this effect is reduced to zero), the

complete Hamiltonian to be used shall be

Equation 10

Now, it has been stated above that PMR spectroscopy involves the

transitions of a nucleus in a magnetic field between energy levels

designated by the magnetic quantum numbers; e.g., m = + , m = 0 when

I=i. These transitions require the absorption and release of energy.

The transfer of this energy is accomplished in part by nuclear

precession, a property of the magnetic nucleus.

12

When a nucleus, i, is in a magnetic field, H., it precesses about

an axis parallel to the direction of the field. This is illustrated

below:

The angular velocity of the precessing nucleus is id. . This u.,

the Larmor precession frequency, depends upon the magnitude of the

angular spin momentum, 1., and the magnetic moment, y., of the nucleus,

or the magnetogyric ratio of the nucleus V. . w. depends also on the

field seen by the nucleus, H..

u. = V.H.I ii

Equation 11

This nuclear precession is an integral part of one relaxation

mechanism. Relaxation of the nucleus begins as soon as it absorbs

energy. Atoms or molecules move rapidly as a result of thermal energy

Such thermal motions set up oscillating magnetic fields which may have

the same frequency vector as the precessing nuclei. The effect of such

thermal motion could be then to change the magnetic quantum number of

the precessing nuclei. That is, transfer of energy from the relaxing

nucleus may occur by its transition from one state to another because

of the magnetic field induced by thermal motion of neighboring magnetic

nuclei and molecules. Eventually thermal equilibrium with the other

degrees of freedom is established. This relaxation depends on

temperature, concentration, and viscosity. It is called longitudinal

13

relaxation, or the spin-lattice relaxation time, T., explained here,

in part, by one of many relaxation mechanisms.

The second form of relaxation is called transverse relaxation,

or the spin-spin relaxation time, T. If several nuclei precess about

axes parallel to the same magnetic field and these nuclei do so in phase,

there results a rotating magnetic vector perpendicular to the axis

of the magnetic field. When this precession of the nuclei fall out of

phase, the rotating magnetic vector would diminish to zero. The rate

of this type of relaxation is known as T. Any factor which tends to

hasten the loss of phase of the several precessing nuclei shortens the

spin-spin relaxation time, T?. Such factors include a non-homogeneous

applied field, and non-homogeneous internal field within the sample as

effected by sample viscosity. These relaxation mechanisms have an

effect on the character of the PMR signal. The line width of the PMR

signal is determined by, among other effects, the relaxation times. The

line width is of the order I /T and; especial ly noticeable in viscous

liquids or solids, when T is decreased, the line width becomes very

broad.

Because the PMR technique involves the placement of molecules in a

magnetic field, an essential part of the background is a discussion of

magnetochemical properties of molecules.

An applied magnetic field polarizes a molecule placed in it by

inducing electronic orbital currents and spin alignments within the

molecule. Aspects of the magnetic susceptibility of molecules are

relevant to the understanding of PMR spectroscopy. The magnetic field,

H, and the induced magnetic moment per unit volume, M, are related by

H = XyM Equation 12

14

where X is the volume magnetic susceptibility and depends only on the

substance placed in the magnetic field.

The induced magnetic moment is parallel to the field when X is

positive. In this case the substance is said to be paramagnetic; para

magnetism usually occurs in substances having electrons with unpaired

spins. Substances are diamagnetic when X is negative.

The molar magnetic susceptibility, X, may be defined by

X = v Equation 13m

where M is molecular weight and d is density.

In substances having little symmetry, the relationship H = X M

may not hold, and magnetic susceptibility may have to be expressed by

three vectors alonq three directions I, 2, and 3: X = l/3(X. + X + X,),a ' '

m I 2 3

The substance is said to be diamagnetical ly anisotropic if

X. =

Xy r X.,.

The Bloch equations apply to the macroscopic moment, M, when all

nuclei are acted on by the same field, H . M is the resultant magnetic

moment per unit volume, for several nuclei of magnetogyric ratio, J ,

in a magnetic field, H, acting in the z direction and i>5 is the

angular frequency of Larmor precession.

iM = YMH =a) M Equation 14

dt o o

Since M has components M,

M,

Mr x' y'

z

Equation I 5x

Equation I5y

-rr~~

0 Equation I5z

dMX

dt=

u Mo y

dM

to Mdt O X

dMz

0

These equations for the x, y, and z components of the magnetic moment

per unit volume must be modified by consideration of fluctuations and

relaxation effects so that since M approaches M,

zrK o'

dMz

M - Mz o

dt T

dMX

dt

M

u M - =*

o y T2

dM

Y -

dt

M

-uM- /

ox T

Equation I6z

and

Equation I 6x

Equation I 6y

where T. is the longitudinal relaxation time and T is the transverse

relaxation time.

In a case of rapid molecular motion the local magnetic field

changes rapidly and the above equations must be modified. The corre

lation time, / ,is introduced by Bloch as a measure of the fluctuation

rate of the local magnetic field. If the fluctuations are very rapid,

then

w"/ < < Io'c

the/decay of all components is equal; M,

M,

M becomes identical and' K ^ x' y'

z *

T.=

V

If H. is a field perpendicular to H and is rotating with angular

frequency, u), then it has component

(H.) = H. cos cj t Equation 17I x I

^

(H.) =-H. sin u t

I y I

16

The complete Bloch equations follow:

dM M

- = Y(M H + M H. sin wt)- -^ Equation I8x

v

y o z I TM

dt

dM M

-j^- = YW H. cos tot- M H ) - =^ Equation 1 8y

dt z I x o T_'

dM M_

M

-rr- = )f(-M H, sin tut- M H. cos u>t)

-

^dt x I x I T.

Equation I 8z

These Bloch equations refer to a fixed coordinate system described by

directions x, y, and z. The coordinate system may be allowed to rotate

about the z axis with angular velocity -to. The new vectors of the

macroscopic moment, the in-phase component of M, u, which is parallel

to the direction of the applied field, H,and the out-of-phase com

ponent, v, which is perpendicular to the direction of H., may be related

to the fixed macroscopic moment vectors by

M = u cos cut- v sin cot Equation 1 9x

xM

M =-u sin <jjt

- v cos tot Equation 1 9y

Substitution of these relationships into the Bloch equation yields

-rr + ^ + (to -

to)v = 0 Equation 20dt T o

N

dv v

-rr + zr- - (oi -

to)u + vH.M = 0 Equation 20vdt T o I z

M

dM M - M

-rr- +-^f

- Yh.V = 0 Equation 20zdt T. I

^

If the rf field H. is not large, M is nearly M the equilibrium value

of that polarization so that

dv v-rr + =r-

- (to -

to)u + vH.M = 0 Equation 20vdt T o I z

M

17

becomes

dv. + Y_ _ ( _ u)u + vH M = o Equation 2I\

dt T o I o

18

NMR Applications to Hydrogen Bonding

The formation of the hydrogen bond changes the physical and chemical

roperties of the molecule. Hydrogen bond formation thus affects a

change in the electron distribution of the molecule. As previously

stated, the electron environment of the proton influences the magnetic

field the proton experiences. That is, the hydrogen in the unassociated

molecule experiences a different magnetic shielding than the hydrogen

in the associated state.

There would be, upon hydrogen bond formation, two possible effects:

(ignoring the possibility of intermolecular, donor-acceptor, electron

currents) (55).

(I) The proton of the acid, A-H, experiences a field due directly

to currents induced in the B atom of the base; this may lead

to a contribution to the proton chemical shift.

(2) The base atom, B, disturbs the electronic structure of the

A-H bond and, thereby, alters the magnetic susceptibility

of the proton. A change in the shielding constant results;

this change would be reflected in the chemical shift of the

proton (55).

Upon hydrogen bond formation there is a down-field shift of the

A-H proton PMR signal unless the base B is an aromatic pi-electron

system (54,60). This implies that in association there is a decrease

in the diamagnetic shielding; i.e., the proton is more"bare"

because

the proton's electron environment has been repulsed by the base electrons

(66).

This downfield shift due to association has been observed frequently

(41,42,43). In a proton study of quinoxaline

in inert solvent, dichloromethane, CH_CI,

and in trifluoroacetic acid, CF-COOH, Blears and Danyluk (14) noted

19

downfield shifts for all protons when the protonating solvent was used.

They attributed the deshielding to the formation of a conjugate acid

with the heterocyclic nitrogen accepting the proton. Upon association,

the charge density about the ring is changed, and this alters the

shielding of the ring protons. This change in electron distribution is

not yet we I I understood.

Rapid equilibria between the species (conjugate acid and base, the

salt, the original base and acid) in acid solutions was indicated,

according to Blears and Danyluk, by the symmetry of their spectra, type

AA'BB'. A more complicated spectra, type ABCD, wou I d have been expected

if the protonated heterocycle lifetime were significantly greater than

the exchange rates between it and the proton donor, or it and possible

tautomers (14).

It may be added here that the ABCD pattern arises from four non-

equivalent nuclei of the same species and a AB arises from four nuclei,

two of which are equivalent to each other and non-equivalent to the

remaining two. The more complicated spectrum would appear when the life

times of the four-equivalent nuclei are long enough for a characteristic

signal to appear for each.

It is assumed that the proton exchanges between the heterocycle

and the acid, and that it spends about equal time with each. When the

mean lifetimes of the protonated heterocycle and the acid are long

compared with the exchange rate time, sharp resonance peaks are

observed for the species. But when the lifetimes are shorter and compare

with the time of exchange; a single resonance peak is observed at a

chemical shift intermediate between the chemical shifts of the two

species (62).

20

This is to be expected upon consideration of the Bloch equations.

A case of a proton exchanging between two sites is pictured in figures

(a) through (d) below. Figure (a) depicts the spectrum when the life

time of either site is long; figure (d) depicts the spectrum when

exchange is rapid and lifetimes are short.

T(.u -

to. ) = 10a b

(b)

"y(co -

to, ) = 2a b

(d)

The hydrogen bond system is in a state of dynamic equilibria where

the hydrogen atom of the acid may be in any one of several chemical

environments.

ft) CF-jCooU + C.FjtOOH

21

0---H-0

o -

u-o

& /w <LP3 COO VA

<X*l <TT3 f3

<^x Cv CN

c^ 6 ^ VA -ft

+ CF3CQOV-\

6 - U-&

o

W-o>^F3

QC')A- (Lv\COo\A

..H-On

o"

x<2>

+ CF3Coov\VA-O

o

Nc-cr,

(W .. ^]

5Jp-CF

\AO-

0."^)

. u-o\

C-C.F.

o

'\&H- Os.

O'/C-^3

22

tS) \+- O

o

N

C-CF,// 3

E)fe>Uv...

ft"

] ;=

r h'7^ H

B\AV

+

ft"

o>

.?

C.-CF7

Qa.E)<s

') VA\.7o-C - C F,

-K

uv-V- CF3COO

QC) .*

rvi U .... o

C-^F,//

r>l Hv C_ F3 COO

*-)VI'..

.-O.

C-CF

//

-V- C F3 C OO

0-BWV

+ "% B A- "Vv\

G) A- f^-W P\- V-\ + C\

*The quinazoline equilibria are represented in this manner because

it is not yet known whether nitrogen 1 or nitrogen J>, or both

participate in the hydrogen.

23

Should the exchange of the proton between these various species occur

rapidly, the resonance signal of the proton in any one of its environ

ments cannot be separated from the signals of the proton in the other

possible species. That is, the resonance signals of the proton involved

in rapid equilibria may coalesce. The rapid equilibria cause only a

time-average environment of the proton to be detected, and one resonance

signal often called the "exchange-averagepeak"

results.

Because the width of the proton resonance signal may be thus

effected, it is well to consider the Bloch equations which, modified,

describe the system undergoing such equilibria or proton exchange.

The Bloch equations for the in-phase, u, and out-of-phase, v,

components of the macroscopic moment are expressions of a complex

equation. A complex moment, G, is defined by

G = u + iv Equation 22

so the equation becomes

|j + [j- - i(u -

co)]g =

-iVH(MoEquation 23

This equation for G describes the macroscopic moment, M, where all

nuclei see the same field, H .'o

However, nuclei in different chemical environments experience

different fields due to the shielding effect,

H. = H (I -<P) Equation 6i o

^

where <r> is the screening or shielding constant.

The resonance frequencies of non-equivalent nuclei are corres

pondingly different; if the Larmor frequencies are designated to. and

uD for different environments A and B. Then the macroscopic momentsD

will be independent and described by equations.

24

I - i ( co-

to) G. = - i V M . H Equation 24

i\A

dG- i ( ok -

to)

B

\K M. H. Equation 24a

B

In order to modify these expressions to allow for exchange between

environments A and B, the populations of the nuclei in these positons

need be considered. If /. is the mean lifetime of the nuclei in AA

and Tais the mean lifetime of the nuclei in B, the population at

D

A, p., and at B, pRshould satisfy

pA+

pb= ' Equation 25

and may be related to the lifetimes of the nuclei in these two states

by

r, rB

The Bloch equations then become

dG

B^A

+

nEquation 26

d*- i (w. -

to)

"A

A

C C

irH.Mo + -I -A.

1A ^B rA

Equation 27

dG

T2L B

- i (to_ -

to)

GA GRGD

=-iJH.M + ^A

- j-

B 'B TA rB

The total complex moment, G, is given by the sum G. + G .

G =-i^H.M

"I o

r +7 rA B A B

f-- i(coA-to)

uA

ji (toR

-

to)

I + = i (to. -

co) i (toR-

to) rBi i

Equation 28

25

If the lifetimes of the nuclei are very short, then this expression for

"he total complex moment becomes

G =-iVh.M

I o

^A+ %

ri- i (to -

to) r*+ri

T2B

i(coB- . to)'B

Equation 29

The imaginary part of the complex moment represents a mean resonance

signal with frequency

mean PAUA+

VbEquation 30

and with width

'A

VEquation 31

The transverse relaxation time, T, may be quite smaller than the

longitudinal relaxation time, T. ; and the two signals may not collapse

completely, but will instead produce a wider exchange-average peak.

This occurs when the exchange of the proton is not rapid enough to give

complete collapse.

JB(co - to

)2

- 7l) Equation 322 2

'A PB A B 8

A B

The exchange between sites as described by the above expressions

is precisely the circumstances of the proton in hydrogen bonding.

The line broadening effect due to hydrogen bond formation may

thus be explained. The degree of downfield shift varies for the protons

of the N-heterocycles depending upon which ring they are on. The

degree of shift may be used as a qualitative indication of the relative

electron densities of the carbon atoms to which the ring protons are

bonded (46). Also, the degree of PMR signal shift upon association is

26

thought to correlate with the hydrogen bond strength. The shielding

of the protons in the N-containing ring is effected by several factors

in that ring, namely; (I) the electric field of the ring nitrogen,

(2) the magnetic anisotropy of the nitrogen, (3) the diamagnetic

anisotropy of the pi-electron cloud, and (4) the pi-electron density

of the ring and the charge.

Nitrogen possesses a quadruple moment, 0, which it is recalled from

above, a measure of the non-sphericity of charge. This environment

*

about nitrogen would effect especially those protons nearest nitrogen.

Line spectrum may be broadened due to quadrupolar relaxation.

The downfield shift of all the heterocycle's protons may be

explained by the delocal i zation of the charge deficiency, and the

coincidental deshielding of the pi-electron cloud.

The PMR signal of the carboxyl proton of the acid in the system

also experiences a downfield shift upon hydrogen bonding. This proton,

since it is directly involved in the hydrogen bond, is expected to

exhibitthe greatest chemical shift dependence on concentration (65).

Upop hydrogen bond formation the polarity of the A-H bond increases (3).

The greatest downfield shift occurs when the proton is deshielded to

the greatest extent; this occurs when the hydrogen bond association is

the strongest (22,27,41).

The PMR signal of the acid proton involved in the hydrogen bond

may shift upfield upon association with an aromatic base (60). This

is caused by induced diamagnetism by the pi -electrons. The applied

magnetic field causes the pi-electrons to circulate around the entire

ring skeleton. Because of the circulating pi-electrons a secondary

or local magnetic field is set up which augments the applied field

27

outside of the ring but in the same plane. Above and below the

romatic ring, the secondary field opposes the applied field (28).

hould there be an association between the proton donor and the pi-

electron cloud, as in (a) the acid proton would be in the area and

the acid proton signal would appear upfield (60).

A/-, N..H-A

H

(a) (b)

The hydrogen bond shift is temperature and concentration dependent;

a feature that has often been used to identify the intermolecular

interaction as association or hydrogen bond formation (45). Early

workers (8,50) using ethanol observed the temperature deoendence of

the hydroxyl proton chemical shift, and the temperature independence

of the CH, and CH proton chemical shifts. Increased temperature caused

the hydroxyl chemical shift to move upfield as did dilution by inert

solvent. This can now be explained as the disassociation of hydrogen

bonded species causing the upfield (toward chemical shift of unassociated

states) shift. Solvent dilution, since it produces the same effect

as temperature increase, may be used more conveniently.

Often to obtain the chemical shift for the non-associated state,

the compound is successively diluted with inert solvent and the results

are extrapolated to infinite dilution.

It has been noted that the temperature dependency of the chemical

shift of acid-base systems may not be entirely due to a shift in the

hydrogen bond equi I ibria. The slight te-'ioerature deoer.dency of the

28

chemical shift of the proton participating in the hydrogen bond may

arise from temperature dependent changes in the effective length of

the bond H...B (28,68).

Chemical shifts are also susceptible to changes in solvent (64).

The equilibrium constant of association can be estimated from the

chemical shifts measured as a function of concentration (35,54).

For this calculation the unassociated and the complexed species

are assumed: ( I ) to have characteristic precessional frequencies,

v and v ; (2) each species has a lifetime longer* than 3 x 10 sec;

and (3) the observed frequency, v', is a weighted arithmetic mean of

these.

C = moles of complex

A =moles of acid or proton donor

B =moles of proton acceptor

v' =

a VC+

VAEquation 33 (35)

v. and v are obtained by extrapolation to infinite dilution of a plot

of values from a PMR study of pure acid.

vc(O (A + B - C)

K "

(A - C)(B - C)Equation 34

For systems in which the proton donor is self-associating another

expression may be used (54).

A' = moles of proton donor not associated with base

Y = fraction ofA'

which is monomer

A = A' + C

M = moles of monomer,Ya'

v' is measured PMR frequency

v1 = ( jA v + (G-jv Equation 33

29

v is characteristic precessional frequency; a weighted average

frequency of monomer ncjquency of monomer not hydrogen bonded to B.

Qv' = v + (v. - v) Equation 35

I I I Ic

.. ,,

l=

TTi r ix + Equation 36v'

- v K(v. -

v) dC v - vM

A A

A plot of : versus-rr-

should be straight line, from whichv - v do

v. - v and K may be determined (68).

From temperature dependence studies the enthalpy of association

AH may be estimated (35) by use of the expression

AH/RT2

= OlnK/aT)P

30

Charge Density Considerations

Other protons in the system experience altered resonances upon

hydrogen bond formation. The ring protons of the aromatic base show

a change in chemical shift as association occurs. This change in the

chemical shift may be used as an indication of a redistribution of

the heterocyclic pi -electrons.

Recall that the chemical shift is a function of the applied field,

H,and a local factor, the screening constant, er-. (See equation 8)

The shielding constant at nucleus I is the sum several contributions.

An expression of these combined contributions has been proposed.

<r = <^''

+ (r'''

+ Il'2+<rl'rIn3

1 d P2t\ Equation 37 (63)

/T- I Iwhere d

'

is the diamagnetic contribution from the electrons on

atom I,

CT* 'is the paramagnetic contribution from these electrons.

I 20-

'is the contribution from the electrons on atom 2, that is, a

neighboring anisotropy effect. <EP'

is the contribution due to ring

or non-localized electronic currents. These contributions determine the

screening constant which, in turn, determines the chemical shift (II).

The importance of ring currents in determining aromatic proton

chemical shifts has long been recognized and estimates of the contribution

have been attempted.

The theoretical contribution of the pi-electron ring current of

benzene to proton shielding has been estimated as -2.24 to -2.76 ppm.;

estimates based on experiment evaluate the contribution to be between

-1.48 to -1.44 ppm. (II). Despite this discrepancy, the contribution

is seen to be large enough that one wou I d expect detectable changes

in the -chemical shift upon redistribution of pi-electron density.

31

Conversely, one would expect changes in the electron distribution

to be indicated by changes in the shielding parameters of ring

protons.

Quantitative effects of the charge density distribution on

chemical shifts of ring protons are difficult to determine, but it

is valuable to correlate the order of chemical shifts with the order

of charge densities. This correlation is based on the assumption

that the chemical shift of the proton is displaced by an amountpro-

portional to the pi -electron density on the ring carbon atom to which

it is attached.

Early workers recognized this interrelationship of ring electron

distribution and ring proton chemical shift. Shifts of aromatic

protons upon dilution with solvent were attributed to a "ringcurrent"

effect (65).

Much work has been done to correlate the chemical shift of

aromatic nuclei with the known inductive and resonance effects of ring

substituent groups. Taft, et.al. (67) investigated substituted

f I uorobenzenes in carbon tetrachloride. Electron withdrawing groups,

or deactivating meta directors, were found to cause the fluorine

resonance to shift upfield, while electron releasing, or activating

ortho-para directors, caused a downfield shift of para fluorine and

an upfield shift of meta fluorine. Comparative results were found by

Corio and Dai ley (19) who studied substituted benzenes. The ortho, meta,

and para protons exhibited a shift to high field of the chemical shift

when electron withdrawing groups were substituted, and a low field

change in chemical shift when electron releasing groups were placed

on the ring. The meta directing groups shifted the meta and para

32

protons equally, but shifted the ortho proton a great deal. The

Corio and Dai ley qualitative determination of electron density (19)

indicated q < q = q : that is, the charge density at the meta

x> ttiNp' '

and para positions are about equal and both are greater than the

electron density at the ortho position.

In the attempts to correlate electron density with chemical

shifts several important conditions must be met. The degree of

hybridization of the attached ring carbon atom must remain unchanged.

There must be no change in the contribution to cr of the sigma bond

between the ring carbon and proton. Buckingham points out that this

second condition might not be met in reality. He states that

although inductive effects of substituents do contribute to pi-electron

distribution, since the protons are removed from the ring by the sigma

bonds, it is "dangerous to try to correlate proton signals with

pi-electron densities as the associated carbonatoms."

(16) Buckingham

concludes that the chemical shifts can be better accounted for in

terms of the electrostatic field created in the molecule.

Proton-transfer studies involving aromatic heterocycles have detected

alterations of ring proton chemical shifts due to hydrogen bonding (43,65).

These changes have been attributed to redistribution of pi-electron

density upon formation and dissociation of the hydrogen bond.

The first problem to be considered in this study is whether or

not the acid-base interaction is indeed the expected hydrogen bond

association. The second problem is to characterize, or find out as

much as possible about this association and the corresponding equilibria.

33

Many studies have shown that hydrogen bonds form between weak

iroton acidsand electron donor molecules such as those containing

luorine, oxygen, or nitrogen; or even aromatic hydrocarbons (69).

The bases used in this work are all aromatic N-heterocycles; the

molecules are planar and have their electronegative atom, the nitrogen,

in easily accessible positions on the two fused six-membered rings.

The study of these compounds should prove interesting because of

the effects of the nitrogen lone pair electrons in the ring pidistri-

bution and because of theoretical aspects regarding N-hetero aromatic

systems. The nitrogen in the pyridine ring causes reduction of the

pi-electron density at the carbon atoms relative to the carbons in the

benzene ring (46).

The outstanding feature common to these chosen bases, or proton

2 2 3acceptors, is the nitrogen in the ring. Nitrogen (Is

,2s

, 2p ) has

two sigma bonds, one pi bond, and one lone pair of electrons (32).

Quinoline, benzo(b) pyridine, or l-azanaphthalene occurs in coal

tar and distillation residues of petroleum. It is a weak base, pKa4.9l

(37,69). Spectra of quinoline have been obtained and characterized

(12,28,33,43,55,65).

Isoquinoline or 2-azanaphthalene, Beilstein #3078, is found in

nature as a constituent of coal tars. Although it does not occur in

biological systems itself, many compounds of biochemical interest

contain the isoquinoline unit. It is a weaker base than quinoline,

pKa. = 5'. 36 (27,69). Spectra of isoquinoline have been obtained and

studied (13,55).

Quinazoline, I,3-diazanaphthalene,

Beilstein #3480, is the third

base studied in this work. It is the strongest base of this series,

pKa= 3.51.

34

Quinazoline provides a good starting point for a series of hydrogen

bond studies because of the relationship of its structure to many

biologically important molecules (51).

The dynamic equilibria of biological systems depends a great deal

upon hydrogen bonding. In fact Pauling has written "... the significance

of the hydrogen bond for physiology is greater than that of any other

structuralfeature."

(51) The more information concerning the hydrogen

bond aptitude of individual components of those systems, such as the

N-heterocycl ics, the more light will be shed on biological processes.

The hetero ring of quinazoline

of purines commonly found in DNA

>)

))

resembles one ring

It is this particu lar

ring

N

which, by hydrogen bonding, participates in holding the

two strands of the DNA helix together.

unit

There is also a structural relationship to the pteridines, base

Nwhich have been found to be effective against

Npernicious anemia and some types of leukemia (I).

Folinic acid, which has many functional groups to

which it may attribute its activity

0 H H

I IN- CH2M,h70>-C-N-CCHCHC0oH

I 2 2 2

CO H

is necessary in mammalian cell division; a part of its complex structure

resembles the quinazoline N-hetero ring. There is a whole class of

naturally occurring materials called quinazoline alkaloids (10).

35

The PMR spectrum of quinazoline in inert and polar solvent has

ieen obtained and studied (6,13,28). It is considered afirst-

order spectrum (28) and has been analyzed as such.

Trifluoroacetic acid, CF-COOH, is an organic acid, pKa= 0.588

(25 C. ) (40). It has been used in hydrogen bond studies before (43),

and is the proton donor used throughout this work.

Chloroform, one solvent considered in this work, is able to form

hydrogen bonds by donating its single hydrogen. It has been shown to

associate with acetone and other oxygen-containing acceptor molecules

(23,30). PMR hydrogen bond studies of chloroform systems have been

conducted (27).

Chloroform is weakly self-associating but correction can be made

by studying it in inert solvent (41). The infinite dilution shifts of

chloroform in inert and in proton accepting solvents are first

obtained; the difference between these values is the chemical shift

between unassociated chloroform and the acid-base complex.

This work begins with the synthesis (where necessary) and purifi

cation of the molecules to be studied. Once this task has been

accomplished, the acid-base systems shall be studied neat and in

solution by (i) varying the mole fractions and concentrations to determine

the sensitivity of the spectra to this variable, and (ii) measuring

and varying cell temperatures to detect and identify the hydrogen bonded

species.

36

EXPERIMENTAL

Quinazoline preparation (5).

Reaction diagram, figure I.

4-Ch I oroqu i nazo I i ne

4-Hydroxyquinazol ine, Aldrich m.p. 218-218.5,21.9 g. (0.15 mole)

was mixed with phosphorus pentachloride, Eastman Organic Chemical,

50.0 g. (0.24 mole) in phosphorus oxychloride, \80 ml. The mixture

was refluxed for three hours; then the phosphorus oxychloride was

distilled off. The residue was mixed with chloroform, 170 ml. and

poured onto ice, 262 g. Ammonium hydroxide, Baker Analyzed Reagent,

0.898 sp. gr., 48 ml. was added to adjust to the pH 8. The mixture was

then extracted four times with 50 ml. of chloroform. The yellow residue,

23 g., was chromatographed on an alumina column, Alcoa F-20, 146 g.,

56 x 2 cm. The benzene eluents yielded 12.3 g. of a creamy white solid,

98-9

C, 40 yield.

Hydrazine Derivative

4-Chloroquinazol ine, 10.0 g. (0.06 mole) in 15 ml. chloroform was

added slowly to p-tol uenesu I fony I hydrazide, m.p.|ll-2 C, I I.I g.

(0.06 mole) in 40 ml. chloroform. The mixture was kept at 26 C. for

17 hours. The creamy white precipitate was collected and washed with

chloroform; air dried, m.p. 157-160 C, yield 24 g. (0.071 mole).

Ouinazol ine

A hydrazine derivative, 17.0 g. (0.05 mole) was added slowly to

sodium hydroxide, 500 ml. (20.0 g. sodium hydroxide, 150 ml. water,

37

350 ml. ethylene glycol) (2) and stirred for three hours at 86 C.

The reaction was quenched by the addition of 200 g. of ice (9,57).

Quinazoline was extracted twelve times with 50 mj. chloroform.

The reddish brown oil 7.7 g. (0.05 mole) v/as chrornatographed on an

alumina column, Alcoa F20, 77 g., 29 x 2 cm. Benzene eluentswere

collected; the pale yellow white solid residue, 3.2508 g. (0.025 mole)

m.p. 43-4 C, was then vacuum distilled twice, m.p. 45.8 - 46.7 C.

The distillate gave a single peak on the Perkin-EJmer gas chromatograph

800 record using a Carbowax 20 M column at 200 C. The fragmentation

pattern from the CEC 104 mass spectrometer indicated no fragments

other than those accounted for by quinazoline, molecular weight 130.15.

Elemental analysis: Calculated: ft? = 73.I0;$H = 4.84;$N = 21.68

Found: fc = 73.50;$H = 4.61 ;%H= 21.45

Quinazol ine

Quinazoline, Aldrich Chemicals, was vacuum distilled twice at

1 .25 mm'.- 1 .30 mm. pressure. The center fraction, b.p. 70.0 C,

was reserved for investigations. This distillate, m.p. 46-.5-47.8 C.

gave a single peak on the Perkin Elmer gas chromatograph 800, Carbowax

20 M column, 200 C. The fragmentation pattern from the CEC 104 mass

spectrometer indicated no fragments other than those accounted for

by molecular weight 130.15.

Elemental anahysis: Calculated: %C = 77. I0;#H = 4.84; %H = 21.68

Found: %C = 73.78;$H = 4.64;$N = 21.15

38

PREPARATION OF QUINAZOLINE

+ PCI,

POCI, solvent

:MH2NH-S02{? "'V-CH3

V

NHTos

I

OH

NHNHS02-(/J-CH3

39

Qu i nol ine

Quinoline, Eastman Organic Chemicals, was vacuum distilled twice

at 5 mm.- 8 mm. pressure. The center cut, b.p. 99-99.5 C. was

25.4reserved for instrumental investigations. This sample

n^"

'= 1.6220

gave the expected NMR spectrum, a single peak upon gas chromatography

on the Perkin Elmer 800, Carbowax 20 M column,185

C. The fragment

ation pattern from the CEC 104 mass spectrometer indicated no fragments

other than those accounted for by quinoline, molecular weight 129.15.

Elemental Analysis: Calculated: C = 83.70?*; H = 5.60?; N = 10

Found: C = 81.01?; H = 5.43?

C = 83.16?; H =

N = 10.45?

N = 10.61?

Isoquinol i ne

Isoquinoline, Aldrich Chemicals, was vacuum distilled twice at

4 mm. -5 mm. pressure. The center fraction, b.p. 91-92 C. was reserved

254

for instrumental investigations. This sample, n_. = 1.6199, gave

the expected NMR spectrum. The fragmentation pattern from the CEC

104 'mass spectrometer indicated no fragments other than those accounted

for by isoquinoline, molecular weight 129.15.

Elemental Analysis: Calculated: C = 83.70?; H = 5.60; N = 10.80?

Found: C = 83.02?; H = 5.41?; N - 10.67?

Gas chromatography on the Perkin Elmer 800 Carbowax 20 M, 185 C.

revealed an impurity, possibly quinoline, of about 3?.

A sample of isoquinoline was purified by preparative gas chroma

tography on the Varian 1520, Carbowax 20 M column at 225 C. This

sample was used to make several acid-base mixtures and the instrumental

results were consistent with those results using isoquinoline, twice

distilled and not purified by g3S chromatography.

40

Trifluoroacetic Acid

Trifluoroacetic acid, Eastman Organic Chemicals, 110 g. was

distilled with 7 g. trifluoroacetic anhydride, Eastman Organic

Chemicals. The fraction b.p. 70.5-71.0 C. was reserved for

instrumental investigations.

Chloroform

Chloroform, Baker Analyzed Reagent, was purified by the techniques

ot Fieser (25), and as recommended in Weissburger's, Techniques in

Organic Chemistry, VII.

Preparation of Acid-base Mixtures

The two-component mixtures were prepared in small 3 or 10 ml.

vials with plastic stoppers. The base component was weighed into the

vial and then the more volatile trifluoroacetic acid was added. A

Mettler semi-automatic balance 170 g. limit, 0.1 mg. was used for

weighing. The sample mixtures were prepared in open air.

The three-component mixtures were made following the above

procedure except that the solvent, dimethy I sul foxide, was added last.

The total moles of isoquinoline and trifluoroacetic acid were calculated;

then the proper amount of dimethy I sul foxide to be added was calculated.

A pre-weighed amount of dimethy I sul foxide was added to the chilled

binary mixture; the weight of d imethy I su I foxide actually added was

determined by weight difference (weight of vial containing all three

components minus the weight of the vial containing two components).

The sample mixtures were pipetted into an NMR sample tube already

containing the reference tetramethy I si I ane. The tube was immediately

stoppered.

41

Instrumentation

The nuclear magnetic resonance experiments were performed on

an Hitachi Perkin-Elmer High Resolution Nuclear Magnetic Resonance

spectrometer, model R-20, frequency 60 MHz for hydrogen. The block

diagram of the instrument is illustrated in figure II. Two major

units comprise the R-20; the operating console and the magnet console.

The magnet console consists of a permanent magnet of 14,092

gauss and is thermostated to a constant 34 C. (so long as room

temperature is within the range 18 -28 C. ) to maintain field and

resolution stability. Normal thermal drift and external disturbances

are compensated for by a closed loop field locking system which keeps

an external control water sample in resonance by adjusting the

magnetic field. This control sample is positioned within I cm. from

the measured sample.

The magnetic sweep is accomplished by a linearly varying R.F.

field. The abscissa sweep spectrum may be observed in two ways on the

operating console: using an oscilloscope, or a recording chart which

lies on a flat bed x-y recorder. The recording charts are calibrated

within 0.2? or 0.2 Hz. For example, when a sweep width of 600 Hz

is used with the tetramethy I si I ane peak at the 0 Hz position on the

chart the chemical shift of a resonance signal at 600 Hz would be

within 1.2 Hz of its true position. This hydrogen bond study often

employed 1200 Hz sweep widths. Chemical shifts of the time average

exchange peak were observed in the range 700-1200 Hz. The maximum

instrument chart error would then be 2.4 Hz for the extreme low

field shift.

42

The recording chart speed may be set at 30, 60, 125, 250, 500,

1000, 2000, and 4000 seconds for the full chart abscissa; and the

bscissa may be set to represent 30, 60, 120, 600, and 1200 Hz sweep

w;idth. The chemical shift and line width parameters may be read with

accuracy from the calibrated chart (see above) or may be determined

by use of an electronic digital counter, with accuracy to 0. I Hz.

Line widths were determined by use of the digital counter, Model

TR 3824X.

The resolution attained is 0.3 Hz and, under normal operating

conditions where room temperature changed not more than 1 C, is

stable.

Also, the spectrum is reproducible. For five successive sweeps

at 250 second sweep time the average deviation is 0.4 cps., providing

room temperature does not change by more than I C.

The sensitivity specification for the R-20 is a signal to noiee

ratio of 12/1 on the largest peak of I? (volume) ethylbenzenedeutero-

chloroforrn quartet. These experiments were run when operation at a

20 to 25:1 signal to noise level was obtained.

The operating console also contains a series of dials which make

it possible to:

(I) select the input frequency level; the H| level used in these

experiments was 3.2 x10^

micro-volts.

(2) select the sensitivity; the sensitivities most often used

in these experiments were in the range 5 to 100. Sensitivitywas adjusted according to the concentration of the sample.

(3) select the mode of representation, either dispersion or

absorption peaks. The absorption signal was recorded in this

work.

(4) integrate the spectrum. Integration was not necessary for

these experiments and, therefore, was not carried out.

(5) select the band pass width. In recording the spectra the

. usual setting was 0.1 second on the time constant dial.

43

Sample tubes were 15 cm. long and 5 mm OD glass cylinders with

a rounded closed end. The usual sample size was 0.5 ml., and never

less than 0.3 ml .

4.99/67MC

5M1C

/^VCD

r

SV/EEP

CCT

PEC-Q

/-/

777'

D6CAP

6E7Z

CRO

B

EM

98KC

OSC

A DP.

---59.9/ic--

MULTI

x /2

033

MOp B.P.F

MEASURME,

n

b"tlKCy-\ <

/OOKCt/KC

L.

r"

ATT

J

BALANCED

MOO,

\

N

IOOKC

B.P.F.

ZKCKC

V-F

CONV.

f/U(E_PN_ REU/V/t]

E1UL Tl.

x 12

59.9MC S.S.B

MOP.

60MC

B.P.F,

CONTROL

/OOKC

OQC

STABIL IZEP

SUPPLIES.

H.T, & CT.

-*- TO UNITS

MAQENT

TEMP.

CONTROL

TO MAC?NET

FIG. // BLOCK DIA&RAM OF

tHANNEL

'RIDGE

60,005

rI

COMC

R.EAMP,

D\CSL

5KC9

MPI0\5KC\PHA&

ftriP\BM.\ PST.

^0

/

wf

\|

O II

^//^BALANCEWPICATOR INTEGRATOR

MACrNET

5KC FIELD

MOP

5RC PHASE

SHIFTER

DC. AMP

C R 0

U'R(?

^RfC.

D~

RECORQER

(CALIBRATEP

CHART)

ERROR OUTPUT

WNEL 0LOCK-

ON /ND/CATOR

HIGH FREQ. COMPENSATION

R^20H.R. NMR SPECTROMETER

45

RESULTS AND DISCUSSION

Acid-base; Two Component Systems

Qui nazo I ine-Tri f I uoroacetic Acid

Quinoline-Trifluoroacetic Acid

I soqu i no I i ne-Tr i f I uoroacet i c Ac i d

The PMR spectrum of trifluoroacetic acid consists of a single

resonance signal at 690 Hz from tetramethy I si lane. The full line width

at half the signal amplitude, Av,, is 0.6 Hz for liquid trifluoroacetic2

*

acid (Figure III). This sample is a mixture of trifluoroacetic acid

monomers, dimers, and higher order mers formed by hydrogen bonding.

The observed resonance signal is therefore a time average exchange

peak which results from the rapid equilibria A and B. See chart

following pg. 20. An infinite dilution study of trifluoroacetic acid

in carbon tetrachloride indicated the chemical shift of the monomer f0

be 606 Hz from tetramethy I si lane (Figure IV).

The PMR spectra of the bases, quinazoline, quinoline, and iso

quinoline consist of a series of resolved signals upfield from the

trifluoroacetic acid resonance. These spectra have been previously

characterized (12,13,55). The proton signal assignments are illustrated

in Figures V, VI, and VI I.

The PMR spectrum of quinazoline results from the two hetero-ring

proton spins and the four aromatic non-hetero-ring protons. The two

hetero-ring protons give rise to singlets at low field relative to

the other proton signals.

There has been some dispute concerning the relative chemical

shifts of. the H-2 and the H-4 of quinazoline. Black and Hefferman

Figure III tlo follow page 45

F-C-C

O

xo-H

Lj''

Figure IV to follow page 45

Chemical Shift vs Concentration Trifluoroacetic acid-Carbon

Tetrachloride reference tetramethy I si I ane at 60 MHz

720

"j". 7:i: -j7

7-;:-:;;77--:- ;. oo - 6

710

; v t:--:: 77 0 7.-

. .

-

.

;-

_ ..

N 7001

. ._ .

,

X;...-

: ,

'

.

'

'9o

"

j.::".'-

: .s --::.......... -';.!.'"

>. . '..

VD

690 XT -.-.- .. .

'

.. 7,. .._ ... '..... *. . -

' '

+-

..--.i.-.:_--.:r::L-.:-:.::-l::

-.'-

7j---.--

; -.. ..A.:. -\ -y yro

0)

c .y.... . \ . . .-'. . . . . \ y. . . . .

'

.. '. y. y >...''. y y y y ; ..:.(. .

ro 680yyyy.r.yy .

i '.:--!:; : :: -y -:j-.: ;. :;

. . -y_\yy y'-\. ::

w. .,.-_...-.

-----

"'- '-: !

'----!

:--'

- -P- -[

>~

sz 670 , _ . .... :. ..l_...; ._.:_!.. . .'.t-

...... ;...'.'.+ -

j- - -

,

- - -

f-

i- - ,_ . . ~ . -- ; . . . . .

j. . ... -

o . . [ t . ._ | . _ | . . 1.._ .

.\.... ^ ... . . . _ . . I

F ,

..-.(...-. .. ,._

--'_;" -

7 .

' !"

roi_ .

-

. . . i . ; _L. . i . _ . . ;+-

660 - - - -; - i ----.1 - ;-..:. -_ ,-.

-[ p

<d' *

T"

-

' - -- |-

p- -- - - - -

;- '- j. - - -

--

';.:__.::.._'_'..,

7 . . :;: L _.:.;; :'.":. _\ :.;/.:_ ". :.;". '..:_..

B:... :;::.!..:.::[ ::

-

. . .

; " '

. ;:-

. .

--.:-

_ , .

o1_s-

6507"

7777 - -

.

- '!'-'--'":-

N-- - - - ...

...;. . -. '--- . .

,.... - .... ......... ....

X 7 ; ;'

"

: -..-', .:-:-..-..'-.:..-:.:.

-.._.;..-

'. / 7 -.\ :;"

. :.": . --;. 'y . y y :

c: 640. . . ._ _.- _ ......... -~

y~ "

~.~ - -----. ... ~~

\ ., iry.'-.y

+-.._-..:.:-.-.._ *--.-- -- -

'- - . _

s-.... .:: .-.j;:.;..:.| :.:-. _--r.:::..;.i_. -..:.

'

". .... '-. .

sz

en630 ::..-

1:

ro

-;-

,.-:_-::

;z--.---\-:----yv-

.

.-

o

V 620,

. . . : L .

-

7- -

- .

-

j-

SZ

O

-

\

- -- . ...j

= -_.f_. __.

: ... . ... __...(

610.::i:...-:'-':--.-.:-;:-::-.-:-r-::- r:"

-r:-

y.'r:r::,:.- ----_- -.- ---------

-"PPry- :--

r

_.-"_--

>

y yj-"-":-

L;-^7-

t-

.9.

600- '''-' --- -

-,' - -- .'_'..__ -- r--~,~ - - _

-j --

. : !'

:'

.. i ::..::: yyyyyyy.y.: .-.:.::. :...: ..:

590

0.0 0.2 0.4 0.6 0.8

Mole fraction Carbon tetrachloride

1.0

46

report two resolved signals, 562 Hz for H-2 and 554 Hz for H-4 from

tetramethy I si lane at 100 M Hz; that is, they report that the H-4 absorbs

at higher field than the H-2. Two independent groups (6,38) report

strong evidence that the H-4 absorbs at lower field than the H-2.

This study found that both H-2 and H-4 give rise to a resonance

chemically shifted 545 Hz from tetramethy I si I ane. The high field

triplet is assigned H-6. The remaining protons, H-5, H-7, and H-8

give rise to resonances which lie between the H-4 and the H-6

resonances (13).

The PMR spectrum of quinoline consists of many resolved signals

resulting from the three protons on the hetero ring and the four

protons on the adjacent ring (Figure VI). The proton in the two

position gives rise to a quartet at low field; the proton on the

eight position gives rise to a multiplet immediately upfield from the

H-2 quartet. The highest field quartet is due to H-3. The resonance

of the H-4 is found between. the H-8 resonance and the nearly coin

cidental H-5, H-6, and H-7 signals. The H-4 doublet is difficult to

identify because it is not signficantly downfield from the other

aromatic non-hetero ring proton signals. Spin-spin coupling constants

have been suggested for this system (61):

|J34| = 8.3 Hz |J23| = 4.4 Hz |J24| = 1.7 Hz

There is also evidence of cross-coupling between H-4 and H-8

|J48| ='0.8 Hz

The PMR spectrum of isoquinoline is similar to the spectra of

quinoline and quinazoline in that the protons a to the nitrogen are

at low field relative to the other protons of the system (Figure VII).

IC j

Figure V ;to follow page 46

2. 1

.f\.

../W^v-A/v.-

R 1

ijI

JJiiillll

M HI I

77 . I

7; 'ill1

i ' :i

' v

(i v,

P^-pf-y

IOC

^"'Figure VI

0

:o to follow page 46 <1C0

j q?'+*/*'

"\<*-.>v-';/'I

S1.I1

'.iil

8

vw

if!ii i ;

! !

f I

i 1

\'

1

\\fA.n<v.'

yM

V K/'7/7iW r y

' >V^.-^v-l7.'

47

The low field singlet is due to the H-l. The doublet, high field of the

H-l singlet, is due to the H-3. The H-4 gives rise to the resonance to

i

the low field side of the remaining signals but is nearly coincidental

with them.

In the two component acid-base system, the singlet at low field

is identified as the time average exchange peak. The series of

resonances upfield is identified as being due to the base protons. The

entire spectrum is modified considerably when the concentration of

acid and base is changed.

The chemical shift and half-line width of the time average exchange

peak is concentration dependent as can be seen in figures VIII through

XI. In the acid-base system the proton may be in any one of several

environments as seen in the equilibria A through E (page 21-22).

Each existing species among which the acid proton is rapidly

exchanging contributes to the observed time average exchange signal.

As the relative concentrations of the acid and base are altered, the

concentrations of these different species also changes, e.g., the

complexes formed in equilibria QaC, QC, and IC, may be formed in

greater quantities as the base concentration increases from 0.0 to

0.5 fraction. This change in the relative proportions of the various

species is revealed in the time-average exchange signal. For example,

as the mole ratio quinol ine/tri f I uoroacetic acid is increased from

0.0 to 0.3 equi I ibri urn (QC) is shifted toward complex formation. In

the complex, the H-bonded proton is less shielded than in the unassoc-

iated trifluoroacetic acid. As more of the complex is formed, the

acid proton is further deshielded and the time average exchange peak

is shifted to low field. Conversely, as the complex dissociates, the

ico,---

r Figure VII to follow page 47

J\> J jK^/K,

H

!<!

p

t

i^\vv-^ '--V.^-^.sJ^^/.OJ.

-

Figure VIII to follow page 4-7

Quinazoline - Trifluoroacetic acid

Chemical Shift vs Concentration Time average exchange peak

1100

Is!

X

o

oc

ro

1050

- 1000V)

>-

JZ+-

0)

Ero

l_+-

CD

EO

N+-

l_

0)

x:

CO

ro

o

E0>

SZ

o

950

900

850

800

750

725

r -"-!

G

0,. \. :

lr._::--

[

K7:7v|t 7 -i

p 7777f! - -- Vi ' 1-

G

::7 77.

7-K:.-(.-:.:.-;:

.

t _

:t.

7 :::

...,.

.:::.:(.

-

r-L

0.0 0.2 0.4 0.6

Mole fraction Quinazoline

0.8 1.0

Figure IX to follow page 47

Qu i nol i ne -

Chemi

Trifluoroacetic acid

cal Shift vs Concentration Time average exchange peak

1000

8

N

X

o

ro

d)

c

ro

sz+-

<D

Ero

+-

EO

i_

<DSZ

sz

in

ro

o

E(!)

SZ

o

950

900

850

800

750

;;;;:!

"

'

. : . / . r

1.... yy.y j

^vt"- - \;y:

. "-..;

: 7

- ".-,"'

:_."-f

7-

;,-"..;. t-.P. - .--

: i-1-. \y

:...:. a -.. G ..--.;~.:-.-.;-----:.-

- - ii -:7-

'

,

777, 7:7;/: r,:"": 7 :-7;,.-.l-.

:"

:.,

7'

/ |- . . .

,. . .

J- .

-^- . .

;;.r.-. :-:-.. r: ...:.-. G . .

. 7' y .;._.._ : . .

-.;.L-.--.i:.. O .yy.Ayyy .

-r -.::-;.::.-. .-!-. :i:.: I :..:::.., --.;:.

-; j.:

G

O

r-; : 1 .

GO

..-.

77 "-

G

O

700

650

7 Q'

7 ji.

0.0

,...

1

0.2

. _l_;.:---.

0.4 0.6

Mole fraction Quinoline

0.8 1.0

Figure X to follow page 47

Isoquinoline - Trifluoroacetic acid

Chemical Shift vs Concentration Time average exchange peak

1100.---,..-.---:.-;'-: ~'::r::l-y '

-'. ......|.--... .

..... yr. .. .

-.<-. -::::.L

:' '

".'

''

\-:;1;;j _7;.4__

; j ; . ; 7 -..-..-'; 7'

.: .-... :77- ]p-.y: .7

;g

N

:<:'

' ! .

,--..--.

f..... . .

j. .

j |

..... _ .

r

>: 1050 '- ; -- :-r-- --

7--

...-.:-:_-; \y.yzy. [y: ._;_:._7;__ -.--;..;.-. ---..:.: J;i-.'7-..-7^_--::

o-

, - --,.... : - . S ...!-..1

. ( .

,-

VO :-:_::- 7: -..;... .: [:_. . . ..<yy_. 7 ytyil-.: . ;..-;_:.\~z.. yy.p.zryyr \..y.yy.zz.L..::._-.--7-.._

+-" " '

-7:

--' ' --

r_ :_ --;

-

' '"

r-

--i- ' '- ! yy '

ro-

1- -- :"-...- J - _

'

:-_;..._.-. _ ,_ip^.._:: [.__-.__. ._..._. . .-..,..

oc

-..-' -!.."

''

t_ -f

:-;-

-I .

'.!-

: Pro - - - : : . . . j. . _

,. 1 _ . . : . [ i . .._

< :...,...

!. 7 :

"

1 1.!"

f J"-" '

"%j"

i

' "

_

f"

in1000

>---'--

|- -

-[ -,-

-| 1 - - j-''

-~1-:pp.;--..:<:p

- -

7j\ -;;\pp~

y.^rrr. . ,.

:

.-.r

sz - - - i - - - - - - t - - . .- - j. - -.1 ,

. - J j. . ... . \ ... .S-l-% -- j .. - : - ."-.._._.

+-^

_ ... .

'["

i 1

~

-

' ' i'

tr ;

~ t~

CD" -

!" *

"1 ' i " " " ! - 1 ._-.-. l ... [. - - ,- - .. - . - . -

P- - -- - .'-.. -...'. | 1 -r . . . , . ._..._.1 . L i.. - ..,...-.. - --

ro . .... l , . .-..,..

-

._;

-

-_.r-

GL+-

i:

'

'

7.-.7I-

:

'

-

7"

'P'

"'>- '

~

... -. - -: -7fl)+-

950 -: -.

':

r- - I -

-! -;-;--

E ... ; ..::::.:.!

. :..[ -. : .: . yyy .'{.;. \:.:r..\ yrr.:'::.yy y _ .yy.y.y

r.'

. . \:... :_ ;-.--. r . .

'

;.-;' -

()

U -

r- - 1 -

;-

[- - -

; - - - - - - - - - - -

;- \ - - -

,; - ;

;- - -

s-1

- 1 |-

'--; '-

7 i -

'

1 -

.

N-.

"

....::."-;--. :.. r. s ::-. .::.--,--. r. .--; ;-.--; -.:.:: r .-yy.J;.-:-.-.-

,-...rr :.. ....

-...-

-.,

r:

H~-,_._i_.r. .l . .r._. .. r i J..j..: :r. i

.:; ..7":.[. :.

:-.:"'

\yy.y-. _ ..y.

I_-- - - "

; |'

;' " "

f' '

:' " " * ' "

r ;' " "

;" ......

a) -

-y- -r-.r-i-..,... .! ; . .

.., ;. i.

t -ji--- --.--.

sz - ; ! - i ... 1. .. I -

(... ......I ; . _. . . [

a900 -- '- -

yr\yz yy\ zyyy\rr zrzyi yrrP:r~y. yy.-rrr'rry 'yy-yyr:~" 7 '"

r

4-..... ;.-[_--.-- - - .-----.-:-.----.-:-.-. 4_-..l.....l !..._..__;.,.

H- '

..

;-"' / ,

: '(

- - 1 -

-J.-

r. -

sz: r j r ;

'" " -'

<

(t

'

['-

tn l ! '-- -r-- ' _..

- - f . 1r q ..|

. ; . .

_,. .

ra

o 850.

A' ' '''

- '

.-".-

1 r- -

,

- ;-_

--

,Y : i

-

. j '-!;..[ i

' ' "

1 .

-

b0)

A '..

-[-

'.

-|!-...!_

x-. . . 1 f. - -:- [ . .1 .,. 1. -

-7I-

-l -

.-'..,-,...

o :-.-. ,-. "-,- 7 ;-"-

7-: -:-...^-: P\ --:-!-

,T. '7

800

-_"--?-:-

-:^.:-..i r -r.:r. r7"

; ;_/-; .

-

.:.:.:..j.:..:: :-

.-/7 -

;."

. ;.:l: r .;-.[. -.:. :t

..

::.:;/:::- -

:..;.-.:-:--};:::..: . .--:::.-.:::; . :

,

-[- ----;----;-7---L--

r.....T.-..!.:..__. -7---;_--.---,--

--;.L.-.-7-_..--:.---!-r-..-_:.^:;....;.

77.77-f:777;]:. :0, ...777:77:[::::::,:7-::7:--:i:-:-.7:-:;;,;:.:,| 777:77^7777:7

750

::.r:::-': O, , -----,

- ...

, ,

. . . - : ['--

-

-: -T: - ';-

-;! f7 -1

--:--

;- :. "-

-t.'.: .-.

700

0.0 0.2 0.4 0.6

Mole fraction Isoquinoline

0.8 1.0

48

protons are shielded further by their electronic environment and the time

average exchange signal shifts to high field.

The formation of a salt from 0.3 to 0.9 mole fraction quinoline,

and 0.25 to 0.9 mole fraction isoquinoline prevented chemical shift

measurements throughout the entire concentration range. The complete

association of the acid-base ion pairs is thought to occur at a 1:1

mole ratio of base to acid (42,43).

From 0.0 to 0.5 mole fraction base, the predominant equilibria

(C) and (D)

B + H-A t B...H-A t [W"...A~~]

would be progressively shifted to the right, as evidenced by the down- '

field shift of the time average exchange signal. As seen in figures i

VIM, IX, and X only a small amount of base is needed for the complex

to be formed.

The complex dissociates into ions in excess base and the time

average exchange signal shifts to high field when equilibrium E

operates:

[BH+...A~] tBH+

+

The extrapolation of the chemical shift versus concentration graph to

infinite dilution of acid in base may be taken to be the chemical

shift of the base cation. In the quinol i ne - tri f I uoroacetic acid system,

this acid cation is called the quinol inium ion; in the

isoquinol ine-trif luoroacetic acid system the cation is the isoquinol inium

ion.

The time average exchange signal is a weighted sum of the signals

of the unassociated trifluoroacetic acid, v., and of the complex, v~.

That is, the chemical shift,v'

,of the observed peak may be given as (35)

v' = | vc + ^-=-^ VA Equation 33

49

pon rearrangement, this relationship becomes a means by which complex

toncentration can be estimated:

v1-

vAC = A Equation 37

VC~

VA

Since the chemical shift observed for the acid is a result of

monomer, dimers, and polymers, the chemical shift for the acid-base

hydrogen bonded species is considered essentially equal to that of the

acid (42). That is, the time average exchange peak is the weighted

sum(35) :

, _

C A - C - H HV

~

AVC+

A VA+

A VH

where H represents the hydrogen bonded species. Since v.,= v.

this beomes

iC A - C - H . H

V =

A VC+

A VA+

A VA

which rearranges tov'

- VAC = A Equation 37

VC"

VA

The resonance frequency of the complex, v, is found by extrapolating

the plot of the chemical shift vs. concentration to the expected minimum

at 0.5 mole fraction (43).

The calculation of theequi I ibrium constant for equilibrium QE of

the quinol ine-trif I uoroacetic acid system follows.

[qh+...a_] * [qh+] + [a-]

K_ CQH+]rA-][QH...AJ

v1 is the observed frequency of the time average exchange signal, v

is the frequency of the complex ion-pair, found by extrapolation to 0.5

mole fraction to be 1112 Hz from tetramethy I si lane. V(,H+ is the

50

frequency of the quinol inium ion, found by extrapolation to 1.0 mole

fraction quinoline to be 700 Hz from TMS. C is the concentration of

the complex ion-pair and QH is the concentration of the quinol inium

ion. The observed frequency is the weighted sum given as

Assuming a total of 100 moles initially,

CqI + CtfaI = 100

and

[TFAl = Cc3 + Tqh+J

At 0.90 mole fraction quinoline and 0.10 mole fraction trifluoro

acetic acid,v' = 992 Hz using 10.0 moles of acid initially.

tc - qh+3 . roH+D

%,

.

VQH+VC

1

C

C

V-

V

QH C

v'-

vc[QH +}

C 700 - 1112 412

rQH+l992-11.2 120

[tfa"] = c + Cqh+3

10 moles= C + TQH+3

c = io - Tqh+]

C = 3.43CQH+3

3.43[QH+

J = 10 -

QH+

rQH+3 =

j= 2.26 moles

= 3.43

51

7.74 moles

u. CQH+] Fa~1 C2.26ir2.26"] _ ,,

KE"

QH+...A-}"

CT7741

= '66

= 6.6 x10"'

A similar calculation for the equilibrium IE of the isoquinol ine-

trif I uoroacetic acid system follows:

ClQH+...A\] ^ Dqh+3 + Ta"!

K =ClQH+DfA"3

[|QH+...A"1

v' is the observed frequency of the time average exchange signal and

at 0.05 mole fraction trifluoroacetic acid v= 985 Hz. v is the

frequency of the complex ion-pair, found by extrapolation to 0.5 mole

fraction to be 1112 Hz. vjir\|-i+l's ~^e frequency f the isoquininol inium

ion, found by extrapolation to 1.0 mole fraction isoquinoline to be

750 Hz.

The observed frequency is the weighted sum

DQH+...A"3 - C|QH+J v C IOH+3 v

r.QH+...A-l(IQH...A)

C|QH+_A-;jU0H

Equation 33

Assuming a total 100 moles initially

Ciq] + Ttfa] = ioo

CtfaI = Dqh+...a""3 + r IQH+

1

vc ^iqh+...a"3

Cioh+...a-1 W^lC IQH+1 v

'-

vc

= 2.97

52

v -

Ciqh+D Ta~3KE

"

r + ^[IQH ...A J

DQH+...A~] + ClQH+] = 5 moles

ClQH+...A~] = 5 - ClQH+D

OQH+...A"] = 2.97 C IQH+1

2.97DQH+T) = 5 - ClQH+]

DQH+] =

~y= 1.26 ClQH+...A"3 = 5 - 1.26 = 3.74

Ke=

CL26^I.263= A25'__ 4>25x

|0-l

The value of the equilibrium constant for the qu i nol ine-tri f I uoroacetic

_2

acid system, found in a previous study (43)- K = 6.3 x 10,differs

from the value found here. The previous work (43) involved chemical

shift measurements at 40 M Hz, while this work was done on a 60 M Hz

instrument. Due to stability factors introduced in modern instrumen

tation the work is sufficiently more refined and therefore the chemical

shift measurements obtained would appear to be more reliable than

those obtained with the 40 M Hz instrument.

A comparison of the equilibrium constants for the quinol ine-

tri f I uoroacetic acid system, K = 6.6 x 10,and the isoqui nol ine-

tri f I uoroacetic acid system for which K = 4.3 x 10,corresponds to

their relative base strengths. Quinoline is the stronger of the two

bases and is therefore the better proton acceptor. The quinol inium

ion would be expected to be more stable than the i soquinol in ium ion.

This is revealed in the equilibrium constant calculation. For both

systems the complex ion-pair is favored, but it appears to be less

favored in the case of quinoline.

53

The equilibrium constants for equilibria C and D were not

ii

calculated using the above procedure. Due to the nature of the

i

Chemical system no reliable data could be obtained over this region.

j

A computerized calculation using iterative procedure is needed for

the refinement of the data regarding equilibria C and D and would be

useful for the calculations involving equilibrium E.

The half-line width of the time average exchange peak is con

centration dependent as seen in figures XI, XII, and XIII. Line widths

were measured for the qui nol i ne-acid and isoquinoline acid systems.

Line widths were not measured for the quinazol i ne-tri f I uoroacetic

acid series since the mixtures were very viscous and additional line

shape broadening resulted.

The line widths were obtained using the electronic digital

frequency counter and are accurate to 0. I Hz. The spectrum was

expanded to the minimum sweep width which fully presented the time-

average exchange signal; this sweep width used is 30 Hz full scale

(XI).

As stated in the introduction, there are several causes of line

broadening. Should the lifetimes of the acid proton in its different

magnetic environments be long compared to the exchange rate but not

long enough to produce resolved resonance peaks representing each

environment, (i.e., fXto -

w, ) = /2),a broad signal appears. That

a b

is, when the predominant equilibrium is slow, there will be exchange

broadening of the time average exchange peak. The types of equilibria

operating should be related to the concentration of the system. For

example, in high acid concentration equilibrium C:

B + A-H ^ B...H-A

Figure Xi

Time average exchange signal

Quinoline-Trifluoroacetic acid

Traces of actual resonance signals. Chemical

hertz, from tetramethylsilane. The resonance

are not relative from one trace to another.

shifts in

intensities

Trifluoroacetic acid

1.0000 mole fraction

0.8 Hz

690 Hz

Quinoline-Trifluoroacetic

0.19^-1 mole fraction

Quinoline

Quinoline-Trifluoroacetic

0.9170 mole fraction

Quinoline

968 Hz

Figure XII to follow page 53

Time average exchange peak Width at half amplitude vs

Concentration Quinoline-Trifluoroacetic acid

. L :

'"\

o

IT-

F,o

0-

o

24

22

20

18

-_:..-..

i -

l- ---

....

f_

y-

T-"

- -"f:

f- ...-.

16

14

12

.10

8

-fyy

o'

-

; 7 ,

-

^ ---|-,,_,- *E-~ ' - P^yFyyy P:P

:"l~'"c

P

.

>. ... . _. ..

i_-...[

..[._.._ ....

'

. i

6-- - - - i - -

...- - L I 1 -

i i'

-r.Ay. O^P^

0..0 0..2

7 \.'---^M:-B^-^-FFp

4

2

0

^pFFyPpFp^APlPzFzF^FpF.M:^^

Pv.

0..4 0..6 0.8 1.0

Mole fraction Quinoline

Figure XIII to follow page 53

Time average exchange peak Width at half amplitude vs

Concentration Isoquinoline-Trifluoroacetic acid

i .-71

: i .>

. . _ * :'. -

'

-V-

-;-.r--

T.;L-.r 7::.-7 : i ::

16

7 "77

.-. 7

:--(

-

-

.:.

! .. . . yyy y

|7 '7-1j

-

:- - i

- - - , - -- . . .

_r.

... _ - ._.

..^. _ . . :.. : --7 .

[---

14O [Ay~--

{y-prPr <

-

f- - '. --

j ...-;'....i .

;

^

- -:- 7--:1.-,. _ .._... ._

\zy:

y

.777 O-:.

t.--.! ".

-I"

i . . .-;_:. .

-

':

- - 7."

.77 ;

' ' '

A

- -

r 7: p| : . : . 77, ,. .

1'.-' 1 : -

,.: .

1 : ..".-.. | . .

-\

[ _r. . . .

O . . . . ^ . .. : - -

1 [ 1 _. 1 ! - .... . . ;

12

: .

:-: r.r

7- .6? ._

_

-

1 . : -, : .

1 . : .

. . (.-

rl

- ,._:_.-

-

1'

- '. :

'..A. .

10

:Pyy:-'-r

::y7o::":"-tTr77n77:

[zyj.rz.z.y.

r r'.... .rr:f_

:-._; -::. 7V :_-r:

7 -77--"!.:777 7 ;-7-

8f o 777- ; . r; ": :.-.

'

i : ; . J:7;77| : 7: : :77:|'777-7- :"77

777.7:-'

['-'-'"-

ayy '.y . y ... f _ _ 'yy. yy :..:-.:;:.--!.: : - 1- -_- t: h .:::..: i "-"-"-

6---;-~--

._ L-

--_-;-

l : .-. ... ....

"T-f

Tf-7 7.77! .r

"7~77-

. 7 . '.77"

7-

z

;;;- -~-

\:;!"

-

-.:-.:7:-

7.7-

A-'z\z-y'y f -

7-0-..7-7-7

l_ 'SEP'. TiZW

'.'.-'

\r:.:.-''

a - y -

'

1"_" ' ~

7. !;":.:

4

?

-_-_:.:.-.::.-.

:.:!-

l

"---7- --- :l _-_.-;-;_

r _. . .

:77T 77,

. 1

b.-. .-. -.--7.-

-

: :' -

zzz : i

--.-7:::-.-:...._

-&-

.:.-:

0-- - - - 7" "~~ . _-.- - . ~-.-~7--

.7.777--.-.-.j--.--~7.-j.--7-

. [Z Pi ~Z7 P.-

'"-77

0.0 0.2 0..4 0,6 0,8

Mole fraction Isoquinoline

1-.0

i 54

is more likely than equilibrium E:

BH+...A"3 -

BH+

+

i The line width dependency on concentration may then be a means

i

of characterizing the equilibria of the system.

Equilibrium processes resulting from the formation of an H bond

are usually rapid and would not, therefore, be considered to contribute

extensively to the exchange broadening. The dissociation of the

complex into the protonated base and trifluoroacetic acid conjugate

base is also considered a rapid process(42,43).*

The greatest line widths occur in the high acid concentration

range where equilibrium D between the hydrogen bonded acid-base

species and the ion-pair complex, operates.

[b...h-a~] ^ Cb-h+...a"]

This slower equilibrium is the reason for the exchange broadening.

An interesting note is that as the base concentration is increased

just beyond the region of maximum line width, a drastic decrease in

line width occurs. This would indicate that faster equilibria begin

to operate at this point. The equilibrium C

B...H-A ^ B-H+...A"}

would be shifted toward ion pair formation upon addition of base.

Once the complex is formed, the rapid dissociation process becomes

possible. This shift to greater concentration of the ion pair

does indeed occur, as indicated by the formation of the insoluble

salt beyond the 0.3 mole fraction base region.

Another cause of line broadening is the quadruple relaxation

mechanism. Nitrogen (1=1) possesses an electric quadruple moment,

which imparts an asymmetrical charge distribution about the nucleus.

55

Spin-spin interactions with a nitrogen nucleus influence the spin-spin

relaxation time, T and therefore alter the line width of the proton

signal. Notice that the maximum line width for both the quinol ine-acid

and isoquinol ine-acid systems occur in the 0.7 to 0.8 mole fraction

acid range. The quinol i ne-acid systems show the more dramatic line

broadening, the maximum line width being at least 22 Hz. Whereas, the

maximum line width for the i soqui nol ine-tri f I uoroacetic acid system

is about 14 Hz. Of the two bases, quinoline is the stronger. The

hydrogen bond association between the quinol ine ^nd acid is no doubt

stronger than that of the isoquinoline. The proton is probably held

closer to nitrogen in the quinoline molecule than is possible in the

isoquinol ine-acid association, and as such, is more subject to the

strong quadruple relaxation mechanism.

It is suggested here that if the line width vs. concentration

measurements could be made for the qui nazo I ine-acid system, an even

greater maximum width of the time average exchange peak would be

observed since quinazoline is the most basic of the three.

The fully protonated base ion would show a very broad signal

due to the strong quadruple effects of nitrogen. This work was conducted

at room temperature, and no resolved signal was observed for either

the quinol inium ion orthe isoquinol ium ion. Signals of the protonated

bases may be detected if the equilibria are sufficiently slowed, as

they should be at decreased temperatures. Low temperature experiments

are planned.

56

Acid-Base-Solvent: Three Component System

The high resolution NMR technique used in this study requires

samples in liquid state. Unfortunately, the acid-base two component systems

form salts in the mole fraction ranges 0.30 to 0.85 quinoline for

quinol ine-trif I uoroacetic acid; 0.25 to 0.92 isoquinoline for isoquinol ine-

tri f I uoroacetic acid; and 0.45 to 1.00 quinazoline for quinazoline-

tri f I uoroacetic acid. In order to study the acid-base systems over the

complete concentration range, 0.0 to 1.0 mole fraction of one component,

a search for a solvent was made. The salt dissolved most easily in

dimethylsul foxide but even in DMSO, a 70 mole percent solvent was

required to dissolve the salt formed throughout the 0.0 to 1.0 acid/

base range.

The isoquinol ine-tri f I uoroacetic acid system was selected for the

preliminary solvent study since isoquinoline is the weakest base in

this series. The isoquinol ine/tri f I uoroacetic acid mole ratio was

varied as in the two component system and the concentration of dimethy I -

sulfoxide was held at 70 mole percent for all solutions in the series.

The plot of chemical shift versus i soquinol ine/tri f I uoroacetic acid

concentration may be seen in Figure XIV. The expected low field shift

between 0.0 to 0.5 mole fraction isoquinoline did not occur (42).

When the chemical shift is plotted as a function of the mole

fraction trifluoroacetic acid in the three component system and this

plot is compared with the chemical shift versus concentration of the

two component acid-base systems, as in Figure XV, the effect of the

solvent can be clearly observed. There is still a low field shift;

however, the low field shift in the three component system is remarkably

less than that of the two component system. One reason for this

57

apparent decreased chemical shift dependency on concentration is the

dissociation of the isoquinol ine-acid complex due to the increased

dielectric constant of the mixture. Dimethy I sul foxide has a dielectric

constant 48.9(20

C.) and 45.5(40

C. ) (39). Another reason is

the dilution of the acid-base mixture in the large amount of dimethyl-

sul foxide.

The solvent, dimethy I su I foxide, is not an inert solvent. Rather

it is anticipated that DMSO participates in the hydrogen bonding.

Dimethy I sul foxide is known to be a strong electren donor, or base.

DMSO has been used as a basic solvent in hydrogen bond studies because

of the availability of the electrons in the p-orbital of its oxygen

and because it does not self-associate (4). A recent NMR study of

succinimide and dimethy I sul foxide found the N-H proton chemical shift

to be concentration and temperature dependent (56). These

observations are indicative of hydrogen bond associations. DMSO is

known to form strong hydrogen bonds with phenols and mineral acids

(39) as evidenced by the downfield shift of hydroxyl proton in NMR

studies. There has been no evidence, however, that proton transfer to

the sulfoxide oxygen occurs, only that hydrogen bond complexes form.

For example, the association of acetic acid and DMSO is thought to

form 1:1 and 2:1 complexes as shown below:

and

CH3-C

^0-H...CCHT1

3

)=S

Ch

CH,1 3

CH^1 3

0 0-H. ..0 0-H...0

CH,\ 3

=s

I

CH3

(72)

58

The isoquinol i ne-tri f I uoroacetic acid-dimethyl sul foxide system,

then, may be viewed as several competing equilibria. Not only

(A)

CF3COOH + CF COOH ^ CF.

/.0...H-0

C-CF.

0-H...0x

(B)

CF COOH

/

CF, CF, CF,J,

3 ' 3I

3

,c c. c

0 N0...H-0 0...H-0 N0

(C)+ CF,COOH

N 3 .H-0

CF,

Ao

(D)

and

(E)

but also

(H)

CF3C00H

?F3

N ...H-0^

0

CF.

i+..:</ \

CH-S-CH3

Ji ,0

CF,-(/ ^3N0-H...0=S

CH

++ CF COO

.59

The low field shift for the three component system would not be as

great as it is for a system in which there are only equilibria A, B,

C, and D since equilibria E operates in the three component system.

Recall that the resonance signal observed is a time average

exchange peak. The environment of the proton in the DMSO complex

contributes to this signal. The proton involved in a DMSO-acid

complex would not be as deshi elded as a proton involved in the

isoquinoline complex. Since proton transfer to the DMSO is unlikely

(4), the proton would remain closer to the shielding environment

of the acid then it does in the isoquinol ine-acid system where proton

transfer to the base is likely.

Line width measurements, AvL, were not possible for this2

three component system due to viscosity broadening of the mixtures

at34

C.

o

-P

r-l

o

H

BCD

o

Figure XIV to follow page 59

Isoquinoline-Trifluoroacetic acid -Dimethyl sul foxide

Time average exchange peak

Chemical Shift vs Concentration

800

....

750

700'

0)

c3

rH

co 0

650

x:-p

o>

ens

p

CD Q^ 600 ;

o

u

H

N

-P

I 550

cH

P

<H 7H \

x:

CO500

450

0\

400";"

"""

0.0 0.2 0.4 0.6 0.8 1.0

Moles Isoquinoline / Moles Trifluoroacetic acid

Figure XV to follow page 59

Time average exchange peak

Chemical Shift vs Concentration

Trifluoroacetic acid-Isoquinoline O.

Trifluoroacetic acid- Isoquinoline-DliSO <

Trifluoroacetic acld-Dimethylsulfoxide ?

OVO

-P

rj

CD

C5

HH

W

H

&-P

CD

Be>

-P

0)

-P

O

Jh

<H

N

P

CD

.3

H

P

H

CO

Hcj

o

H

B0)

1100

1000

900

800

700

600

500

1+00

77''tr-.!77-''

'I'

Fp:F F:\~F'

o :.

.77:^77-:C? 77.._

77777- 7:i O

"- -^ .-7.-77-.7-_. 77 ~

|,7.-r._.|. .::;:::.-.:: ^

PFF-rFFo7- -,7"-P-~

--.-:}--

: D O :.."-..

--7-7-^-D 777 r-7777777-777

-U-y , . . .,.777777777

'

7 ,7

7v_~."

"~.Trri.""t~ i _

l*t

77777777-;--'

<j<p7717:777777 -\ .

P' P\P--"PPzEzyrPy

7-07.:07.:.,: .-:--_::-.777: 0:.7: .. .. - ,::_ ..

,

:_.-^-^ ,

-7- ___..,_.-

7

:"-.::[77: 7:777'

-.

'

7 77 1_'

7 ._:_;:. . . _

7770.---7-71 ..-y-

-7*--:i -' -

: : ..

i- d<

777.1.7 ::.!.-;:;-..r.:7:!.;.-:.;7;

'^rl :-"T""-i"-rr-t-"."-f < \ -

1.0 0.8 0.6 0.*f 0.2

Mole fraction Trifluoroacetic

0.0

60

Dimethylsul foxide-Trif luoroacetic Acid System

The PMR spectra of each of these components indicates a

resonance signal at 690 Hz for trifluoroacetic acid and at 163 Hz

for dimethylsul foxide.

Trifluoroacetic acid is a known proton donor in hydrogen bond

systems and DMSO is a known electron donor as previously discussed.

This binary system gave the expected chemical shift vs. concen

tration plot (Figure XVI). Association does occur as evidenced by

the drastic low field shift of the acid signal a*s the 0.5 mole

fraction range is approached from either concentration extreme. The

methyl protons of DMSO did not show any modifications in their spectral

signal .

No line width measurements were taken because of the viscosity

of the system.

Figure XVI to follow page 60

Trifluoroacetic acid - Dimethylsulfoxid:

Chemical Shift vs Concentration

o

vD

4^>

crj

CD

%H

H

W

H

>>^-P

0)pir

cti

U-P

0

-p

Bo

u

N

-P

UCD

H

P

<H-

CO

H

o

H

Bo

&o

8^0

800

750

700

650

600

550

,i:: 7: 0 ::.;:: ay 7.77(77;:: -.!..::-7-7|.7 .. -. i -

-

.. :. .7 z\...y.

y '"P''[-'P{.E r'~PJPPPr O 7"y-_'."7".;

- -

.~-;-v-:l--: ...;..y3^i;:....:.o

- ; ,-'-

-r"

'. . - .

- -

7 / .

" - -

\- ~

. 7

'

'" ~~

- -

\ *'('

'

' * r

{'

p K ' i . -.'. P.L"'

7..['. \: '. '.

yyyF'F'.-T'yyWp

'

'/ P y"' "y\

"

T\.F^PPF^^r 4""ife777'

- ,- - ! / -

' '

- 7~77rn. ".7; .f

~

~""| 7 7" "

7 7 : ; 7 7" '

7 77"

V~

7"

",

PP-;

-..:-.-:

-7--.-7:7::-

--,-:..-.-.-

7777:"

O -EEyPy~--~P~--^

yEPrryyyyyyyFEy. P'yyy yyPzPP QO ..-

-

---

--

.: , . - !-

- -

\ :'>::V:-: -:.;1-._--. ; -. ^

7 .

'"'

"

"':"

r~-T~ I" ^

:

ro"-. "..... ....

'

: '':. ; .. PJ -

.. [:.,';\ [_ "-.F\

"

..

-

-

o

-"

- -

--

l- - \ -

PEPPPPaFeeFF.-'

:- T -r -

'

- 7-

. .

(i.

|__ , _ .. j. _ ._

... ! . 7.7-7 L._,7i. | - _. l 7- ... :. 7, .,. 7. .. 7 - . o-

7-. i 7777-j"7::-

-"; 7 7::p:-. : ! --. -.!

7-J."

.

"\-777 7;".--

77-;"7-77-;777^t77-77:7"7777

"

-77--" -

.

-

-;"

j~

O

777J. .7

"

j -7-7.-1. 7. "|-7:-

-f i ;h-

:.K-:

500

o.o 0.2 0.4 0.6 0.8

Mole fraction Dimethylsulfoxide

1.0

61

Isoquinol ine-Ch loroform Solvent Effectsi

I The study of the isoquinol ine-ch loroform system was first under-

I

taken in an attempt to find an inert solvent for the acid-base systems.

Since isoquinoline does not self-associate, no significant change in

the spectrum was expected upon dilution with an inert solvent.

However, changes did occur for this isoquinol ine-ch loroform

system throughout the concentration range. As shown in Figure XVII,

the chemical shifts of all the hetero-ring protons and the chloroform

proton are concentration dependent.

The N-hetero-ring protons appear to be identically influenced by

the chloroform. This may be due to a ring current effect often

observed for aromatic molecules (15,65) and is not necessarily

indicative of a hydrogen bonding association. When a molecule such

as isoquinoline participates in hydrogen bond formation the N-hetero

ring protons are dissimilarly affected; this is shown by this work

as well as others (42,60,64).

It is now known that chloroform is capable of forming hydrogen

bonds. Chloroform can self-associate and forms weak hydrogen bonds

(26). The question that arises fromfhis study is whether the association

is indeed hydrogen bonding.

The usual association shift is to low field as the complex is

formed. The chloroform proton is observed to shift to high field in

the intermediate concentration range. The chloroform proton apparently

is not participating in the usual hydrogen bonding.

One explanation of the data is that the chloroform proton associates

with, or at least is effected by the pi-electron cloud of the N-hetero

ring of isoquinoline. The chloroform proton could approach the iso

quinoline molecule from the top or bottom of the planar molecule as

62

in figure (a) on page 22 of introduction. The proton would be then

subject to diamagnetic shielding by the pi-electron current (15).

The isoqu inol ine-ch lorof rom association, although small in

comparison to the base - trifluoroacetic acid association, was

sufficient to indicate that chloroform would not be suitable for

further work since it is not an inert solvent.

Figure XVII to follow page 62

Isoquinoline - Chloroform

Chemical Shift vs Concentration Ring and chloroform protons

--.. --,--

j-",

r~--

560 -_ j _ -_7!

7-:_I;i7_-_- '

'r_'_-_-: -

.7.

-

.--".7ovO

-p

ctf

o h-7%/-,:-^0:,-,,_.r .:.-::..;.

:->;.-

,^77 7.

H

7~"~

"

"O~ "" """ " ~

"""''

f """ ^ """

'"- "" -- r-

54o _.;_.._ __

. -.Ss!/ |"--

..x,.7| "'7 : -.77-:""-

:"

-j7.37-7 -..''

H

w --:

Hyy-'

_-- 7- -.\".

-,>

-r'

._____ - ..

^. _ ...

{>> - .- _: 1-

^^- - - ...!..j. . . .

^ -7 77 77 7.:.7I7-

-. - - /\ - 1.. ..... ..;....-.... . ; -

,...

-PL ... - 7 . 7. :. -7

-

O ^\- -

|...... - . . -

(D 777 777 7 -.7-7 h -\7-- r-

". -[-- - --(---r-

,

- ;- -

,

E - -

-.7.-

7-J-.7-77-yy-TSt-:;!::::-;-;.:-

-777-77:777777: .ztryy.

U 520 -'[-: - | -

!"

"07.' " '"" '"

"' 7

P i_-..77 :; 7 yy 0777' ;:.;

i:7^^7 .7 .7.

7:7777-

._ _:. ::. ., 7

<D

-P fc-^FF ----t -.-: ---_ ..

- "--

-

^>^73t: " ^-

:o"

0ri3 Q -7 -::--...

-ry:.~.:.-

.A..z-.. -.--. 77...-

..:t. .:_7:7.7 ..: -zzyy 7.-77

<H

-

tjr,E- 7.7::-77i::::._ ... ; 77777::. ... yyy

-

.... 7,"- mP.. 7.7 7 7! O 7 ".

7-

77: 7 77-7- :r-;-

7..:-:. 7 -77

N 500-U1

-r.- "

_ _-'7 7' '" '

+5 - - ^. '. ;.-.-..,..; t . . -

f-J ,- ... ; .---< . .- .... - ^\ . .. ._. -..;...

"

.. . . ; ;

<D r - - - -- -- - !-^. 7 -'- - ;. ;..._..._ 7 .. . . .

,d[yrr- -.-

:. 7 : . 7 . \X -. .

i -

. .._.::

-

;. 7777.7.;-:-

77777O77-

Ay : .

d Az.y.-yyyyry;---

-77-7-7-

0\. 7777:7:7.7777:7. 7. 0 7:7. 7.777 7O::..

H__: :;.7.::7.-.-.:...7.7: . -_,..-tX 7 7._..__.7:37777.7 :_:._..:3:7 .: : .7:7.

p

480r F- -

:

-

!^\Q-

l

-

- !-.

H- -

'

1 r- .

(. P- - .

;...

...

Z^^-.._.... : 1 . ,

CO

77777^7^

H

o L_". :..: i :[ .:::::.:. 1 p~

:::pppp :-"::.--..-..! ::*.:.-. r [:i\ _ 77777 :; . pplpep 7. r::.:\ ;. :i-; ; 77::-.. ;_-_..:

B<D 460 7 77___ 7. 7 7_71 i_ 777.7 ~__L ..77 _ 1 : 7 77 1.7 .

"7" " '

I"

'."lII".*! *,*.L"*.Tr

_7____

" "

"!

,3'- - - - ..-.---

r-- . - .

j.- - . .

,- - -..

j- -i - ....

r- -- - -

U

%F^F&^^^440

:HCI3S^A"

; 7-^A -'.;;...; .|.7:777 :[777777i.7 77:!77:77!:7 7707:7

420 - ' -"- '--;--}- -- \ - - -

yy-y 7 ,-- -

0.0 0.2 0.4 0.6 0.8

Mole fraction Isoquinoline

1.0

63

Chemical Shift of N-heterocycl ic Ring Protons

I The base spectrum also is concentration dependent. The chemical

shifts for the hetero-ring protons change upon addition to acid to

the N-heterocycl ics.

The chemical shifts were very difficult to measure because the

base spectra are complicated by overlapping signals and because the

spectra were drastically modified as hydrogen bonding and protonation

occurred. The results (which should be further investigated) perhaps

at higher H. frequencies are shown in figures XVIII, IXX, XX, XXI,

and XXI I.

However, qualitative conclusions may be drawn from these data.

There is insufficient information for any quantitative results. Also,

the magnitude of ring proton shifts is thought to be medium dependent

and consequently charge density arguments may not be supported by the

data presented here (42,60,64).

The two N-hetero protons of quinazoline seem to be similarly

influenced by the formation of the hydrogen bond. Each exhibits a very

slight high field shift toward higher quinazoline concentrations.

The limited range of concentration over which liquid samples were

possible prohibited measurements beyond 0.45 mole fraction quinazoline.

However, the chemical shifts observed were all downfield of the chemical

shift of 1.0 mole fraction quinazoline. The quinazoline N-hetero

ring protons are more deshielded as acid concentration is increased.

(Figure IXX)

'

This would indicate that the electron density is reduced somewhat

at the 2 and 4 ring carbons in the associated quinazoline. The

decrease in the charge density seems to be about equal for both sites.

Quinazoline ~ Trifluoroacetic acid

N-hetero ring protons

Chemical Slri ft vs Concentration

it::

OvO

fi-P

o <a

H CD

FSN Cj

-P H

fl) W

"ft

CD+5 B<H cd

H *H^"PW CD

-P

H

o

H

BCD

.d

o

610

600

590

580

570

560

550

'54o

? E3

o

D

cSP-

E3

O

-Qr

_^o

7 77-7T7 7.

777

CD.

D?

0

O O

"O

7-7

77:

37X

:2f747

.:.-.t:

-

.-.:..t.....

0.0 0.2 0.4 0.6 0.8

Mole fraction Quinazoline

a

1.0

64

An important implication may be drawn from this observation.

The effects of ring substituents on ring proton chemical shifts seem

to be about equal for meta and para positions, whi le the ortho

position is effected differently (19,67).

The H-2 and H-4 are ortho and para to nitrogen I: in quinazoline

they are both ortho to nitrogen 2. If the hydrogen bond association

hence the effect upon pi-electron redistribution, was through nitrogen

I, the H-2 and the H-4 chemical shifts should be changed dissimilarly.

That is, the A6 would be greater or less than A6.. However, if the

hydrogen bond association involves the nitrogen, to which H-2 and H-4

are both ortho, the effect on the chemical shifts would be about the

same for each proton. More experimentation is necessary, but at this

point, the data leads to suspicions that the nitrogen in position 3

is the one which participates in hydrogen bonding.

The N-hetero ring protons of quinoline also exhibit concentration

dependent chemical shifts. (Figure XX)

Because of the formation of insoluble salt in the intermediate

range the plot is not continuous; nevertheless, between high quinoline

concentrations to high acid concentrations there is a downfield shift

for all N-hetero ring protons. The shift of the H-2 is approximately

10-15 Hz; for the H-3 the change is about 50 Hz; and the chemical

shift of the H-4 is about 80 Hz. At high acid concentrations, the H-4

signal becomes nearly coincidental with that of the H-2.

These observations compare with previous studies that have found

that positions meta and para to the position of disturbance are

subject to greater modification in their chemical shifts (19). It

65

certainly seems from this that the charge density of the 4-carbon is

decreased to a greater extent than that of the 2-carbon. However,

since the magnitude of the change is medium dependent, no definite

conclusions may be drawn.

The N-hetero ring protons of isoquinoline exhibit similar changes.

(Figure XXI )

At low acid concentrations, the chemical shift of the H-l and

H-3 fluctuate about the values obtained for neat isoquinoline. At

high acid concentrations, however, both are shifted downfield. The

H-l is effected less, the change being~

20 Hz. The H-3 is shifted a

great deal, 50-60 Hz; the chemical shift of the H-3 becomes coin

cidental with that of the H-l. This indicates that for some reason the

H-3 has been deshielded.

The N-hetero ring protons of isoquinoline exhibit an interesting

solvent effect. (Figure XXII) At high acid concentrations, the H-3

is shifted 60 to 70 Hz upfield by the addition of dimethy I sul foxide

to the system. This observation would parallel the findings of

Krakower and those of Schaefer who noted the medium dependency of

N-hetero ring proton chemical shifts (42,64). The change in chemical

shift due to solvent for the H-l is not proportional to the change

for the H-3. This leads to doubts about whether the chemical shifts

of ring prorons can justifiably be used to determine even qualitative

charge density distributions.

wchesibimsimmoFiBB,,

Figure IXX to follow page 65

N

X

O

VO

CDc

10

in

>

sz+-

CD

Eroi_+-

CD

EO

1_

CDx:

to

550

540

530

520

510

500

490

480

470

460

450

~

440

<dsz

O

430

420

410

400

Quinoline - Trifluoroacetic acid

Ring protons Chemical Shift vs Concentration

H-2 OH--3a

....7.3

Q -

l . -. ; -

_

1. . I

7-

"

.-. ;

"

'

;-

j

H-4 ? H--8 0

.7, .. t- _')

'_ \ _ I'

0 j '. i'."

1 yy 1: ".. 7. .

"

.!

"

7: ;.'. .7 . i

Q O S. Oi-7-7

I-- -

!i::.- " ( 77

D

-7 :: (.:

...77.L

^30

7;:~

[-. 1

7 '-- ::7-

--.--[--.- - O

1\ \:z: . i O"-

.- : | : :. . .

'

:.-.:!

' -u-

~_,

. 7. . . L.zz : 7. y :'.:..: 7. ~T *

-7""

. . .

y .".

'

i:A-EfFa-

\'

,\ . 1

\ 7-

7:7'

O77 :.-. 77-j.:

Fee ..... 15... -

,

77f

- -

-t- -

1- -

r

h"

'

.

i :.7-i *, 1

r- - -

1-7-7"-F-PEP

1

. .7...(_..-

'

... . j. .

:.-

. , 1 ;'

. ... 0 .

'

- -

;- - - - -

r- 1 - - - -

;- - *

1- ...

j-- - -

!

'L-'

i-

-1- '-

--I

i 777 :[7 -77- 0- ;/

'i7i7.:~~" ' "

t11"

~:~_,,

..- j-..

-77 "t "^ .

-'

7 77 r~i"

7 1 7 ::::."

~~:.7

~:~T~~:

:. 73

v

7 v y

77-;

"-'

-~-.y

v v ,

-

;

[:'-.-

: j -

- V- -

:i '.. .

' 7i~

7*

D

-T-H--0

---Oh

. . j .

7"'

.77

- -

j- -

777?

'

y -7

""7:7 i

D

'_"_'.

77.7:77"

.7

::::-::!'

~~'.'\ ~~E'"\

" '

'.'

. 7"

.7 j

7 7"

\

v. .

"" "

i

v - !

: A

777 -7-

. 37 ...

1~"

I

7 :

^ _t. ...!_. _

-

i- -

1--

- i

t'

\ ....

. . . . .

:FF.77:1 7 77"7|

. ._

"

7t- -

J

7~

'777.71

" - - -

1 03-.

777... .. -i .

..y ": 1 77^

...

j - > -

~~i -r.-..:-.--

1

- - (-:-

'--- -'--

0.0 0.2 0.4 0.6

Mole fraction Ouinoline

0.8 1.0

Figure XX to follow page 65

1 1 soqui no I ine - Trifluoroacetic acid

N-hetero ring protons Chemical Shift vs Concentration

590-

580

N

X21 570

O

(0 560

CO

y-

EO1_H-

550

N-f-

l

540

CDSZ

c

5304-

H-

sz

CO 520

(0

V

E(11

510sz

O

500

0.0

H-1 a

H-3

0.4 0.6 0.8

Mole fraction Isoquinoline

Figure XXI to follow page 65

Isoquinoline-Trifluoroacetic acid-Dimethyl sulfoxide

N-hetero ring protons

Chemical Shift vs Concentration

580 ... ,

.7 .. GG 7. . ...77:7:77:7 77777 7I7777777 :7:-7'

. .7

7777[777-77:o::"-7.

N

:--:!:-'F::^

@

^7^^"^

-

?""

-

' F-F^FP P- .:.,

570-

7

-:77:-o7.-7.77. eGS-4T4-

-

.:-...tr-..:,|r.,-r-:-: -.: 7

<l,. ... .._ . .

. _ __ . . !__._.._|^

... _, f .

^

. _ ...

-7.-77 "..;..: 00"

:PP 7 7777;77-:7:|777 7 ;7 7-7

O

p

E cd 70.-

7.-77 7 -.-;--. ',7 .7

7-.J--7.-

yy.77..-

7 .7 .-[.77-7 .O.-.7.7:.77.77 ;.",.:

O

U CD550

"-

,

"

--7 7-

777 7 77377-7.-7. -..7777.0777.. Ij" "

: .

- 7" ~-.7 7

7-

cdtsl H

N"

.

. ;.. . .:.: . : 7[-.7 : \~y. yy. _7- .37 j z: 37.77. .7077

! - i i - ; - -

[- i ;

' - - - -

P -H

Jh K> J^v.;:.::;77:7.3

0.'

-.7 -.'77 . ;::::::.:

'

r zyzz:\.:: .z:z. .].yy .7 --7-7.7

CD r.7-

. 77 ...:.:7 077- -

17:7-7-

77---

[-

-_ \---_ [ .-y.zy'y.";~

..-.'>;..-.-. . -7-7-777.

-P

. . 1 . . .7 '. _ . .

,_. _

T... _ :.._.:.. ', . . . . .

.--. r 7..7;

.. -_.> 7 .-,. ..:,..

'

1 . L ..7 . ,. ; :...::._. ..r .....

rrr.

H CD

P 05

J^O_^:

-

_

.

7-7-J7'

---.j:

- 77-

'-"-- " " ;-- ; --7-

Or 7

- 7.7- 77 ...... ....-r..|3 . -

"

_.. .|.7 . .. _.. _. 7__ .

_

. _..

H P520

----

Baa+^|h- --:-'-.--'-.-

-:- -

w +i

h . :__ L :..._.....;

3.. ..-_....- J. _ . . _ _. 7 1 : .... 7 .... . .

03

O 51Q7

.07_77 7___.0- 7_-.-i__73.-___-...-._37 .-7.77777 77 _ _ .

H;-- . ...;.(T>

'-

..00 Q , ...... i ,- . . . . ,

6 1 G_l_---t '" -.-

1 , ;

CD77 7 7.77 777-70:7377:3.-77737:777:.7-|.7.7...7-7-7777777-3777 77 7

"

:

O 500(__;:;.

7--777... 7:7777-7-77777:177:77.7 77:7; 777.77: 77. :!:.. 7. .yy 77

777 7

0 --r 1 777 t. { ..__: i_3 ::.._!.. j . ::::. 7::33i.:._: :.:_b zz.aaaa aa ya yy .z.y

0.0 0.2 0.4 0.6 0.8

Mole fraction Isoquinoline

1.0

DMSO

H-l

H-3

70fo 60% 50%

O

D

e

Appendixes

Tabulated data in support of figures

IV, VIII- X, XII - XXI.

Trifluoroacetic acid - Carbon Tetrachloride

Chemical Shift vs Concentration

Chemical shift in hertz from tetramethylsilane at 60

Mole fractions Chemical Shift

Trifluoroacetic Carbon Tetrachloride Hz

1.0000 0.0000 690

0.6.75 0.3825 704

0.3143 0.6857 714

0.2695 0.7305 -71 4

0.14S0 0.8520 71 4

O.0839 0.9161 714

0.0231 0.9769 697

0.0031 0.9969 606

Quinazoline - Trifluoroacetic acid^-^

11

Time average exchangepeal"

(tae) and ring proton peahs

Chemical Shift vs Concention

chemical shift in hertz from tetramethysilane at 60 MHz

Mole fraction

Quinazoline Trifluoroacetic

Chemical shift, Hz

tae H-2 H-4

0.0107 0.9893 748 572 598

0.0870 0.9130 742 573 598

0.1484 0.8516 792 570 594

0.2170 0.7830 864 572 599

0.2299 0.7701 876 572 600

0.2933 0.7077 954 576 606

0.3585 0. 641 5 1010 568 594

0.3957 0e6043 1044 568 592

0.4223 0. 5777 1060 564 588

1.0000 0.0000 - 545 545

Quinoline - Trifluoroacetic acid

Time average exchangepeak-

Chemical Shift vs Concentration

chemical shift in hertz from tetramethylsilane at 60 MHz

Line width at half peak amplitude vs Concentration

Mole fraction Chemical shift width

Quinoline Trifluoroacetic Hz Hz

0.0000 1.0000

0.0954 0.9046

0.1332 0.8668

0.1712 0.8.288

0.1871 0.8129

0.1888 0.8112

0. 1 941 0.8059

0. 2037 0.7963

0. 2084 0.7916

0.2305 0.7695

0.2675 0.7325

0.3024 0.6976

0.8918 0.1082

0.8925 0.1075

0.9170 O.0830

0.9330 0.0670

0.9655 0.0345

654 0.6

726 1.4

79 3.1

782 9.7

806 13.5

806 15.0

812 16. 5

830 18.3

836 21.9

858 19.5

904 8.0

952 3.6

1014 1.4

1010 1.8

968 3.1

922 2.9

754 1.3

Isoquinoline - Trifluoroacetic acid

Time average exchange peak

Chemical Shift vs Concentration

chemical shift in hertz from tetramethylsilane at 60 MHz

Line width at half peak amplitude vs Concentration

Mole fraction

I soquinoline Trifluoroacetic

0.0960 0.9040

0.1070 0.8930

0.1093 0.8907

0.1350 0.8650

0.1713 0.8287

0.1901 0.8099

0.1912 0.8084

0.1918 0.8082

0.1991 0.8009

0.1993 0.8007

0.2049 0.7951

0.2053 0.7947

0.2120 0.7880

0.2206 0.7794

0.2319 0.7681

O.9326 0.0674

0.9417 0.0583

0.9460 0.0540

0.9532 0.0468

Chemical shift

Hz

width

Hz

718 .0.6

737 1.2

738 1.3

748 2.1

780 4.2

814 7--5

81 4 8.4

812 8.5

828 8.8

829 9.0

830 11.3

830 12.2

838 12.7

850 13.2

864 11.6

1100 0.9

964 1.4

1080 1.0

969 1.0

Isoquinoline - Trifluoroacetic acid - Dime thylsulfoxide

Time average exchange peal-

Chemical Shift vs Concentration

Chemical shift in hertz from tetramethlsllane at 60 kHz

Moles IQ/Moles TFA Mole fraction Chemical

Isoquinoline Trifluoroacetic DMSO Shift

Hz

0.0000

0.0960

0.2066

0.3921

0.5275

0.5973

0.6888

O.7636

0.7964

0.8916

o.ooco 0.3333 0.7012 764

0.0289 0.3013 o. 7007 758

0.0339 0.2645 0.7016 750

0.0973 0. 2026 0.7001 727

0.1425 0.1575 0.7000 686

0.1651 0.1342 0.7007 656

0.1980 0.1037 0.6983 604

0.2187 0.0788 0.7025 544!

0.2322 0.0679 0.6999 518

0.2627 0.0361 0.7012 418

0.3125 0.0848 0.6027 506

0.3615 0.0405 0.5980 398

0.3742 0.0279 0.5979 348

0.3924 0.1038 0.5038 656

0.4345 0.0646 0. 5009 504

0.4132 0.0600 0.5268 496

7 j. fluoroac e ti c acid

Time average exchange psak

Dime thy1sulfoxide

600

550

N

: vO ^OO

P

a

EH

o 450

fH

pen

<H 0)

H ^

CO

Hcd

o

H

B<D

O

DMSO 0.6983DMSO 0.5979D77S5 0. 5009

0.7012 mole fraction o

0.6027 mole fraction a

0.5268 mole fraction <

4oo

35o;

0.08 0-.06. o.o4 0.02 0,00

Trifluoroacetic acid - Dimethylsulfoxide

Chemical Shift vs Concentration

Chemical shift in hertz from tetfamethylsllane at 60 MHz

Mole fraction Chemical Shift

Dimethylsulfoxide Trifluoroacetic Hz

0.1128 0.8872 704

0.3000 0.7000 756

0.4935 0.5065 829

0.6950 0.3050 776

O.7012 0.2988 764

O.8970 0.1030 540

0.0000 1.0000 690

;l soqui no I ine - Chloroformi

i

Ring proton and Chloroform proton peaks

Chemical Shift and Line Width vs. Concentration

Chemical shift in hertz from tetramethy I si I ane at 60 MHz

Mole fraction Chernica 1 Shift

Isoquinol ine H-1 H-3 CHC

Hz width Hz width Hz

0.000 - - - - 434

0.008 551 2 509 7'

433

0.016 550 2 507 6 432

0.024 549 2 506 7 432

0.032 549 2 506 7 432

0.040 549 2 505 6 432

0.048 547 2 503 7 430

0.055 546 2 503 7 430

0.063 547 2 503 7 430

0.070 547 2

'

504 7 431

0.103 545 2 502 7 432

0.292 529 3 488 7 432

0.496 521 2 480 7 440

0.697 516 2 476 7 450

Quinoline ~ Trifluoroacetic acid

'Ring protons

Chemical Shift vs Concentration

chemical shift in "hertz from tetramethylsilane at 60 MHz

Mole

quinoline

fraction

trifluoroacetic

Chemical

H-2 H-3

shift :

H-4

7n Hz

H-8

0.0954 0. 9046 548 476 540 -

0.1332 0.8668 544 471 544 -

0.1712 O.8288 538 468 538 -

0.1871 0.8129 546 469 542 -

0.1888 0.8112 546 468 542 -

0.1941 0.8059 546 471 541 -

O.2037 0.7963 550 472 546 -

0.2084 0.7916 546 469 541 ~

0.2305 0.7695 5br4 463 537 -

0.2675 0.7325 54o 462 536 -

0.3024 0.6976 544 467 540 -

O.8918 0.1082 526 419 462 486

0.8925 0.1075 525 41 8 460 483

0.9170 O.O83O 527 419 451 488

0.9330 0.0670 525 415 459 486

0.9655 0.0345 522 412 456 486

1.0000 0.0000 540 416 460 494

Isoquinoline - Trifluoroacetic acid

N-hetero ring protons

Chemical Shift vs Concentration

Chemical shift in 'hertz from tetramethylsilane at 60 MHz

Mole fraction

Isoquinoline Trifluoroacetic

0.0960 0.9040

0.1350 0.8650

0.1713 O.8287

0.1901 0.8099

0.1912 0.8084

0.1918 0.8082

0. 2049 0.7951

0. 2053 0.7947

0.2120 0.7880

0.2206 0. 7794

0.2319 0.7681

0.9326 0.0674

0.9360 0.0640

0..941 7 0.0583

0.9460 0.0540

0.9935 0.0065

0.9999 0.0001

1.0000 0.0000

Chemical

H-1

Shift

H-3

570 562

570 564

570 570

574 574

576 576

576 576

576 576

575 575

576

'

576

576 576

575 575

552 509

552 509

548 505

552 508

556 514

554 512

551 510

Isoquinoline - Trifluoroacetic acid --Dimethy I sul foxide

N-hetero ring protons

Chemical Shift vs Concentration

Chemical shift in hertz from tetramethy I si lane at 60 MHz

Mole fraction

soqui no I ine Trifluoroacetic DMSO

0.0973 0.2026 0.7001

0.0339 0.2645 0.7016

0.1425 0.1575 0.7000

0.1651 0.1342 0.7007

0.1980 0.1037 0.6983

0.2187 0.0788 0.7025

0.2322 0.0679 0.6999

0.2627 0.0361 0.7012

0.3742 0.0279 0.5979

0.3615 0.0405 0.5980

0.3441 0.0569 0.5990

0.3125 0.0848 0.6027

0.4575 0.0435 0.4990

0.4134 0.0599 0.5267

0.4353 0.0646 0.5001

Chemical

H-1

Shift

H-3

578 510

578 507

572 508

568 507

562 505

556 505

558 505

558 507

570 520

570 524

570 519

574 520

569 520

570 520

572 520

Bibliography

i

i

1 Acheson, An Introduction to the Chemistry of

j .Hetero.cxci.ic_Comppunds , Interscience, (London, i960)

2 Albert and Royer, _JCS_,

1949, 1148.

3 Aliksandrovand Sokolov, Doklady Akad. Nauk SSSR, 124,

115- Chemical Abstracts, 1161

4 Allerhand and Schleyer, JACS, _8_5.1715. (1963)

5 Armarego, J. Appl. Chem., 11, 70, (196l)

6 Armarego and Batterham, _JCS,.1966,.

750.

7 Armarego and Smith, JCS, 1965, 5360.

8 Arnold and Packard, JCP,1_9. 1608, (1957).

9 Bamford and Stevens, JCS, 1952, 4735-

10 Bentley, Chemistry of Natural Products, VII, Interscience,

(London, 1965")".

11 Black, Brown, and Heffernan, Aust. J. Chem., 20, 1305, (1967)

12 Black and Heffernan, Aust. J. Chem., 17, 558, (1964).

13 Black and Heffernan, Aust. J. Chem., 18, 707, (1965).

14 Blears and Danyluk, Tetrahedron, 23, 2927, (1967).

.15 Bothner-By and Click, JCP, 26, 1651, (1957).

16 Buckingham, CJC, 38, 300, (i960).

17 Cartme11 and Fowles, Valency and Molecular Structure,

Butterworths, (London, 1961 ) .

18 Coulson and Danielson, Arkiv Fysik,8,239- C.A.,49, 21340.

19 Corio and Dailey, JACS, 78, 3043, (1956)

20 Culbertson, Decius, and Christensen, JACS, 74, 4834, (1952).

21 Curd, Landquist, and Rose, JCS, 1947, 1759

22 Davie s, U.S. At. Energy Coram. UCRL-8909, (1959).

23 Ebsworth and Sheldrick Trans. Faraday Soc, 63, 1071,(1967).

24 Endicott, Wick, Mercury, and Sherrill, JACS, 68, 1299

25 Fieser, Experiments in Organic Chemistry, Heath and Company

(Boston, 1964).

2o Findley, Keniery, Kidman, and Pickles, Trans. FaradaySoc, 63, 846, (1967).

27 Freyman, Freyman, Koechlirn, Martin, and Marvel, Arch.Sci.

(Geneva) 12, Fascspec. 207, (1959), C.A., 54, lTToT

28 Gawer and Dailey, J.C.Phys., 42, 2658, (1965).

29 Gilette and Sherman, JACS, 58, 1135

30 Gordy, JACS, 60, 605, (1938)

31 Gorman, J. Chem.Ed., 33, 468, (1956)

32 Green and Linnett, JCS, i960, 4959

33 Hearn, Morton, and Simpson, JCS, 1951. 3329

34 Hilbert, Wulf, Hendricks, and Liddel, JACS, 58, 548, (1936)

35 Huggins, Pimentel, and Shoolery, J.C.Phys., 23, 1244, (1955)

36 Jones, J. Chem.Ed., 22, 74, (1945)

37 Karer and Schmid, Helv. Chem.Acta, 29, 1853, (1946)

38 Katritsky, Reavil, Swinbourne, JCS, 1966, 351

39 Kharasch and Thyagarajan, Quarterly Reports Sulfur C

Chemistry, 1, 1-86, (1966)

40 Kirk and Othmer (eds.). Enclyclopedia of Chemical

/Technology, 771, Interscience, (London, 1965~T

41/ Korinek and Schneider, CJC, 1157, (1957).

42 Krakower, thesis, (University of British Columbia, 1963)

43 Krakower and Reeves, Spectrochimica Acta, 20, 71, (1964)

44 Lewis, Valence the Structure Atoms and Molecules,

Chemical Catalogue Company, (New York, 192FF

45 Liddel and Ramsey, J.C.Phys., 19, 1608, (1951)

46 Lyle, Chemical and Engineering "News, Jan. 10, 1966, 72.

47 Meyer, Saika, Gutowsky, JACS, 775, 4567, (1953)

48 Nagakura, JACS, J6, 3070, (1954)

49 Nagakura and Gouterman, J.C.Phys., 26, 88l, (1957)

50 Packard and Arnold, Phys . Re_v., 83, 210, (1951)-

!

51 Pauling, The Nature of the Chemical Bond, Cornell

! Univ. Press, (Ithaca, 1948).

52 Pauling, JACS, 33, 1367

53 Pimentel, J .ChemL.JPhy

s . , 446, (1951).

54 Pimentel and McClellan, The Hydrogen Bond, W.H. Freeman

and Company, (San Francisco, i960).

55 Pople, Schneider, and Bernstein, High Resolution Magnetic

Resonance, McGraw-Hill, (New York, 1959)-

56 Porter and Brey, J.Phys.Chem., 71,. 3779, (1967).

57 Powelland Whiting, Tetrahedron, 305

58 Ramsey, Phys. Rev., 78, 699, (1951).

59 Reeves, CJC, 39, 1711, (1961).

60 Reeves and Schneider, CJC, 251, (1957).

61 Reeves and Stromme, CJC, 39, 2318, (1961).

62 Roberts, Nuclear Magnetic Resonance, McGraw-Hill Book

Company Inc^ (New York, 1959)-

63 Saika and Shichter, J. Chem. Phys. , 22, 26, (1954).

64 Schaeffer, CJC, 39, 1864, (1961).

65 Schaefer and Schneider, JCP, 32, 1218, (i960).

66 Shooley, J. Chem. Phys. , 23, 805, (1955).

67 Taft, Ehrenson, Lewis, and Glick, JACS, 81, 5352, (1959).

68 Takahashi and Li, J.Phys.Chem., 68, 2136, (1964).

69 Tamres, JACS, 74, 3375, (1952).

70 Tamres, Searles, Leighly, and Mohrman, JACS, 76, 3983, (1954).

71 Weissberger, (ed. ) Organic Solvents, Techniques of Organic

Chemistry,VII, Interscience (New York, 1955).

72 Vogel, JCS, 1948, 1833

73 Lindberg and Majanj^ suomen.Kem., 38B,21, (1965) C.A. 62,13023a.

74 Muller and Reiter, J. Chem. Phys. , 42, 3265, (1965)-