Citation
Base strength-reactivity effects in polyethylenimine esterolysis reaction

Material Information

Title:
Base strength-reactivity effects in polyethylenimine esterolysis reaction
Creator:
Lege, Curtis Stanley, 1946-
Copyright Date:
1979
Language:
English
Physical Description:
xi, 94 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Amines ( jstor )
Esters ( jstor )
Ethanol ( jstor )
Flasks ( jstor )
Imidazoles ( jstor )
Ions ( jstor )
Nucleophiles ( jstor )
pH ( jstor )
Polymers ( jstor )
Pyridines ( jstor )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Organic compounds -- Synthesis ( lcsh )
Polyethylenimine ( lcsh )
Polymers and polymerization ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 90-93.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Curtis Stanley Lege.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
023414625 ( AlephBibNum )
06997957 ( OCLC )
AAL4927 ( NOTIS )

Downloads

This item has the following downloads:


Full Text











BASE STRENGTH-REACTIVITY EFFECTS
IN POLYETHYLENIMINE ESTEROLYSIS REACTIONS










BY


CURTIS S. LEGE


A DISSERTATION PRESENTED TO THE GRADUATE
COUNCIL OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1979





































TO MY WIFE

AND

DAUGHTER

JOE ANN AND SPRING














ACKNOWLEDGMENTS


The author wishes to express his appreciation to his

research director, James A. Deyrup, for his thoughtful

guidance and criticism as well as his encouragement of

individual creativity in problem selection and pursuit. The

author is further indebted to Professor John A. Zoltewicz

for the many helpful discussions and use of equipment.

The author wishes to thank the faculty and staff of

the University of Florida and his fellow graduate students

for an enjoyable and fruitful experience. Special thanks

are extended to Tom Baugh and Eric Langenmeyer for their

friendship and advice.

The author is especially indebted to his wife who has

provided not only encouragement throughout but also has

made considerable contribution to the preparation of this

dissertation.















TABLE OF CONTENTS




Page


ACKNOWLEDGMENTS..... .................................... iii

LIST OF TABLES............................... ........... vii

LIST OF FIGURES.............................................. viii

ABSTRACT................................................ x


POLYMER ESTEROLYSIS REACTIONS:
DISCUSSION AND PROPOSAL..................

Introduction..............................

Discussion................................
Mechanism of Polymer Esterolysis
Reactions..........................
Polymer Catalyzed Hydrolysis........
Summary .............................
Proposal..................................
Introduction.........................
Nucleophiles.........................
Polymer..............................
Ester ...............................
Results..............................


AMINOLYSIS OF P-NITROPHENYL ESTERS BY
DODECYL POLYETHYLENIMINE...................

Results and Discussion...................
Characterization.....................
Basicity of PEI-D-NH2-HC1...........
Kinetic Analysis ...................
Bronsted Relationship....................
Summary...................................


CHAPTER I




















CHAPTER II








CHAPTER III HYDROLYSIS OF P-NITROPHENYL ESTERS
CATALYZED BY POLYMER BOUND HETEROCYCLES.. 44

Introduction............................. 44
Polyethylenimine Systems................. 44
Preparation.......................... 44
Primary Amine Analysis .............. 45
Further Kinetic Analysis ............ 49
Summary .............................. 51
Non Polyethylenimine Homopolymer Systems. 52
Conclusion............................... 58


CHAPTER IV EXPERIMENTAL............................. 60

Introduction ............................ 60
Syntheses................................. 61
Dodecyl Polyethylenimine PEI-D....... 61
Dodecyl Polyethylenimine Hydro-
chloride PEI-D-NH2-HC1........... 61
Dodecyl-4-methylenepyridine-poly-
ethylenimine Hydrochloride
PEI-D-Pyr-HC ............ ......... 62
Dodecyl-N-(2-pyridyl)-3-propylamine-
polyethylenimine Hydrochloride
PEI-D-APyr-HC .................... 62
Dodecyl-4(5)-methylenimidazole-poly-
ethylenimine Hydrochloride
PEI-D-Im-HC ..................... 63
Dodecyl-4(5)-methylenimidazole-
isopropyl-polyethylenimine
Hydrochloride PEI-D-Im-Ip-HC1.... 64
Dodecyl-4(5)-methylenimidazole-iso-
propyl-isopropyl-polyethylinimine
Hydrochloride PEI-D-Im-Ip2-HC1... 65
4(5)-Chloromethylimidazole Hydro-
chloride. ........................ 65
4(5)-Hydroxymethylimidazole
Hydrochloride...... ............... 65
3-Bromopropanal Dimethyl Acetal..... 66
N-(2-pyridyl)-3-aminopropanal
Dimethyl Acetal................... 66
N-(2-pyridyl)-3-aminopropanal
Hydrochloride..... ................. 66
p-Nitrophenyl Acetate PNPA.......... 67
p-Nitrophenol Caproate PNPC......... 67
Composition of PEI Derivatives by
Elemental Analysis............... 67
Composition of PEI Derivatives by
NMR Spectral Analysis............ 73
Determination of Primary Content of
PEI Derivatives Using Trinitro-
benzene Sulfonate (TNBS) ......... 73








Primary Amine Detection by Ninhydrin 74
Relative Concentrations of Imidazole
Containing Polymer Solutions..... 75
Potentiometric Titration of PEI-D-
NH2-HC1.......................... 76
Preparation and Standardization of
CO2 Free 0.1 N KOH Titrant....... 78
Preparation of Ester Solutions Used
for Kinetic Studies.............. 78
Preparation of Polymer Solutions for
Kinetic Studies .................. 78
Buffer Solutions.................... 79
Kinetic Method....................... 79
Background Rate ..................... 80
Pyridine Ionization in PEI-D-Pyr-HCl 81


APPENDIX................................................ 88

REFERENCES.............................................. 90

BIOGRAPHICAL SKETCH ..................................... 94















LIST OF TABLES


Table Page


II-1 PRIMARY AMINE AND ISOPROPYL CONTENT OF
PEI-D-NH2-HCl AND PEI-D-IpHC1 ................ 23

III-1 % PEI UNITS BOUND TO HETEROCYCLES............ 45

IV-1 COMPOSITION OF PEI DERIVATIVES FROM ELEMENTAL
ANALYSIS DATA................................ 70

IV-2 FRACTION OF PEI UNITS ALKYLATED FROM NMR
SPECTRAL DATA................................ 72

IV-3 PRIMARY AMINE CONTENT OF PEI DERIVATIVES
USING TNBS.................................... 74

IV-4 POTENTIOMETRIC TITRATION DATA FOR PEI-D-NH2-
HC1 ................................... ... 77

IV-5 ESTEROLYSIS RATES FOR PEI-D-NH2-HC1.......... 83

IV-6 ESTEROLYSIS RATES FOR PEI-D-Ip-HC1........... 84

IV-7 ESTEROLYSIS RATES FOR PEI-D-Im-HC1........... 84

IV-8 ESTEROLYSIS RATES FOR PEI-D-Im-Ip-HC......... 85

IV-9 ESTEROLYSIS RATES FOR PEI-D-Im-Ip2-HC1....... 85

IV-10 ESTEROLYSIS RATES FOR PEI-D-APyr-HCl......... 86

IV-11 ESTEROLYSIS RATES FOR PEI-D-Pyr-HC1.......... 87

IV-12 ESTEROLYSIS RATES FOR ETHYLENEDIAMINE HC1.. 87














LIST OF FIGURES


Figure Page


I-la Solvolysis of PNPA catalyzed by Poly-
4(5)-vinylimadazole and imidazole........ 10

I-lb Bifunctional mechanisms proposed for
poly-4(5)-vinylimidazole catalyzed
hydrolysis of PNPA....................... 10

1-2 Esterolysis of PNPA by PEI-D-Pyr-HCl..... 18

I-3 Esterolysis of PNPA by PEI-D-APyr-HC1.... 19

I-4 Esterolysis of PNPA by PEI-D-Im-HC1....... 20

I-5 Esterolysis of PNPA by PEI-D-NH2-HC1
compared to esterolysis of PNPA by PEI
bound heterocyclic systems............... 21

II-1 Comparison of pKa ap dependence on a for
PEI-D-NH -HC1 with low molecular weight
polyamines................................ 27

II-2 Comparison of esterolysis rates for
PEI-D-NH2-HC1 and PEI-D-Ip-HC1........... 29

II-3 Inhibition of ethylenediamine, PEI-D-NH2-
HC1 (pH=7.00) and PEI-D-NH?-HCl (pH=8.82)
esterolysis of PNPA by p-nitrophenol
anion (PNP)............................... 31

II-4 Plot of log(k /a) against pH for PEI-D-
NH2-HC1 esterolysis of PNPC and PNPA...... 33

II-5 Bronsted plot for the esterolysis of PNPA
by simple primary and cyclic secondary
amines .................................... 36

11-6 Bronsted plot for the esterolysis of PNPA
by imidazoles and anilines............... 37


viii








II-7 Plot of pKa,ap vs pH for PEI-D-NH -HC1....... 38
a,app 2
II-8 Bronsted plot for the esterolysis of PNPA and
PNPC by PEI-D-NH2-HC1......................... 40

II-9 plot of e vs pK or [a] for the esterolysis of
PNPC and PNPA by PEI-D-NH2-HC1................ 41

III-1 Effects of repetitive isopropylation of PEI-D-
Im-HCI on PNPA esterolysis reactivity......... 48

III-2 Comparison of PNPA esterolysis rates for
PEI-D-HC1 and PEI-D-Ip-HCl.................... 50

III-3 Plot of kobs against a for the esterolysis of
2,4-dinitrophenyl acetate by poly-4-vinylpyri-
dine and pyridine.............................. 54

III-4 Plot of k2 against a for the esterolysis of
PNPA by PEI-D-NH2-HC ............................ 55

III-5 Bronsted plot for the esterolysis of PNPA by
poly-4(5)-vinylimidazole and imidazoles....... 56

III-6 Bronsted plot for the esterolysis of 2,4-dini-
trophenylacetate by poly-(4)-vinylpyridine.... 57

IV-1 Typical spectrum of a PEI-D derivative (PEI-D-
APyr-HC ) ..................................... 71














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

BASE STRENGTH-REACTIVITY EFFECTS
IN POLYETHYLENIMINE ESTEROLYSIS REACTIONS

BY

CURTIS S. LEGE

December 1979

Chairman: James A. Deyrup
Major Department: Chemistry

The esterolysis of p-nitrophenyl acetate (PNP) and

p-nitrophenyl caproate (PNPC) by polyethylenimine (PEI)

derivatives was studied as a function of pH. The reactivity

of partly dodecylated PEI containing primary amines

(PEI-D-NH2-HC1), imidazole (PEI-D-Im-HC1), pyridine

(PEI-D-Pyr-HCl), and 2-aminopyridine (PEI-D-APyr-HC1) was

found to be pH dependent without exception. The

PEI-D-NH2-HC1 system was found to react with PNPA predomi-

nantly by nucleophilic attack of polymeric primary amine.

Potentiometric titration showed the DK or PEI-D-NH -HC1
a,app 2
also to be dependent. A Bronsted-like plot of log k for

polymeric free primary amine against pKapp for the ester-

olysis of PNPA with PEI-D-NH2-HCI had a slope (U value) of

0.81. A Bronsted plot of a series of low molecular weight

amines with PNPA taken from the literature had a slope of

0.83. The Bronsted-like plot of log k with pKapp, an
a, apt








approach unknown previously in polymer systems, quantita-

tively accounted for the pH dependence of the reaction of

PEI-D-NH2-HC1 with PNPA. The reactivity of PEI-D-NH2-HC1

toward PNPA is dependent on the pK of the nucleophile as in

the case of low molecular weight amine nucleophiles. The

slope of the plot of log k versus pKa,app for the esterolysis

of PNPC by PEI-D-NH2-HCI had a slope of 1.06. This slope was

dissected into a pH dependent polymeric enhancement component

(E) and a pH dependent nucleophilic component. The pKa,a
a,app
of the heterocycles in the heterocycle containing polymers

was unobtainable in the pH region of interest. However, the

pHdependencies were assigned to nucleophilic effects by

analogy. This approach proved useful in reviewing literature

reports of other polymeric esterolytic systems. Data for

PNPA esterolysis by poly-4(5)vinylimidazole and for

2,4-dinitrophenyl acetate esterolysis were plotted as

described above with slopes of 0.8 and 1.3, respectively.























xi














CHAPTER I


POLYMER ESTEROLYSIS REACTIONS: DISCUSSION AND PROPOSAL


Introduction


Esterolysis by polymer systems has received considerable

attention since the early days of polymer research. This

area of research was spurred as the mechanism of enzyme

reactions became better understood. It was hoped that

synthetic polymer systems might offer a method of modeling

or duplicating the action of the naturally occurring polymers

(enzymes). The aspects of enzyme action of greatest

interest in this field are high selectivity and high

efficiency. To this end synthetic polymer systems with

particular binding properties and high reactivity have been

sought.

One of the aspects of polymer esterolysis reactions

which has received a great deal of attention is coopera-

tivity, i.e., two or more functional groups in the same

polymer molecule participating in ester cleavage. The

original goal of the research to be discussed herein was to

find and study possible cooperative effects in nitrogenous

polymer systems. Chapter I will first discuss the mechanism

of polymer esterolysis and aspects of possible bifunctional

or cooperative mechanisms. Secondly, the design of a system






2

in which cooperative effects may be found will be discussed.

The remainder of this dissertation is a discussion and

interpretation of the results of the study and the impact of

this interpretation on the field of polymer esterolysis.


Discussion


Mechanism of Polymer Esterolysis Reactions

One of the features present in polymer reactions that is

generally absent in monomer reactions is the possibility of

binding preassociation. A simplified scheme for such a

reaction is shown in Equation I-1.


Equation I-1

P + S -a PS --- Reaction --- Products
I b


The substrate may (path a) or may not (path b) bind, depen-

ding on the structure of both the polymer, P, and the

substrate, S. Binding can greatly increase reaction rates

by increasing the concentration of substrate in the presence

of the polymer. There are at least two modes of binding,

hydrophobic and electrostatic, for synthetic polymer systems.

Hydrophobic binding is the mole of interest here; it requires

polymers and substrates of hydrophobic nature. Such

substrates frequently contain long aliphatic chains which

are usually located on the carboxyl end of the ester. The

electrostatic effects of concern are those produced by a

partly ionized polymer and their consequence on hydrophobic

binding and nucleophilicity. These effects are quite






3

important, but until now they have been impossible to

quantitate.

In esterolytic reactions of polymers, the binding can

certainly increase the reaction rate, but the cleavage por-

tion of the esterolysis mechanism is considered to be

independent of binding.f That is to say that the ester is

cleaved by the polymer in the same way, whether it is bound

to the polymer in a previous step or not (neglecting buffer

reactions, etc.). This is a reasonable supposition in that

nonspecific hydrophobic binding should not result in any

significant change in the details of the reaction mechanism.

This assumption, which is usually not stated, is very impor-

tant in understanding polymer esterolysis reactions as it

allows comparison to nonpolymeric reactions.

The specific types of esterolysis reactions of interest

to this study are those of nucleophilic amines with p-nitro-

phenyl esters in aqueous media. Satterthwait and Jencks2

have provided the mechanism shown in Equation I-2 for

esterolysis reactions which are first order both in p-nitro-

phenyl ester and nonpolymeric amine nucleophile. The

tetrahedral intermediate T is formed in a rapid and

reversible step. The rate determining step is the breakdown


Equation I-2

O H O
I ki+ I k
R-NH +Me--C--OR' X" R-N -C--OR' --- Products
k-1 I I
H Me
+





4

of T. That is, T is partitioned between starting mate-

rials (k-1) and products. The partitioning is governed by

the relative strengths of the C-N and C-OR bonds. The bond

strengths can be related to the pKa's of the two possible
3
leaving groups. The greater the pKa of the nucleophile,

the greater the percentage of tetrahedral intermediates

proceeding on to products. It is clear then that the pKa of

the nucleophilic amine is a very important measure of its

reactivity in aminolysis reactions. The correlation of

reaction rate with pKa, known as the Bronsted relation, is

very important to the arguments put forth later and will be

discussed in more detail in Chapter II.

The breakdown of the tetrahedral intermediate may be

catalyzed by general acid or general base mechanisms. The

mechanism of general base catalyzed breakdown of the tetrahe-

dral intermediate T is shown in Equation 1-3. This pathway

transforms the protonated amine group into a nonprotonated

amine. The amine pKa of importance for the anionic


Equation I-3

H 0 0
I+ 1 kb[B]\
R-N --C--OR' R--N--C--OR --- Products
S i k-b[BH ]
H Me H Me
T T


intermediate, T is the conversion of the amine to the

amide anion. This pKa is very much greater than that of the

alcohol. The intermediate T is thus prevented from breaking

down to starting materials. This general base catalyzed





5

pathway promotes favorable partitioning of T toward

products. The third order pathway contributes more in cases

where the amine pKa is significantly lower than that of the

alcohol moiety. The general acid pathway generates the T

intermediate via a more intricate set of proton transfers.

However this pathway includes proton transfers, which for

reactions of moderately basic amines, are thermodynamically

unfavorable forp-nitrophenyl ester reactions (proton donation

from a weak acid to a weaker base). The general acid

pathway is of higher energy than the general base and the

uncatalyzed pathways.4 Therefore, general acid catalyzed

reactions will not be considered further.

The above discussion was largely directed at primary

and secondary amines; however, it also applies to nucleo-

philic tertiary amines.3 The heterocyclic tertiary amines,

pyridine and imidazole, are of particular interest. Most

pyridines are incapable of losing a proton in the T interme-

diate ,and the T intermediate is therefore not accessible

(Equation 1-3). Nucleophilic reactions of pyridines with

p-nitrophenyl esters must pass through the second order

H




Pyridine Imidazole


uncatalyzed pathway. Imidazole on the other hand is not

restrained from access to the T intermediate. The imidazole

may lose a proton from the T- intermediate to form T-. This

pathway is analogous to the pathway proposed by Satterthwait








and Jencks for the general base catalyzed aminolysis of

esters by primary and secondary amines. Satterthwait and

Equation I-4

0-
0-
-I Kb[B],
H -N-N-C- OR --- N N C -OR -- Products
I-/ | K-b[BHt] \3
CH3 CH3

T T-


Jencks did not discuss the scheme in Equation 1-4. However,

this mechanism has been proposed for nonaqueous systems.5

This discussion is not meant to imply that there is a general

base catalyzed imidazole acylation reaction with p-nitro-

phenyl esters in aqueous systems, but for the purposes of

this discussion any general base catalysis will be assumed

to follow this mechanism.

The general base catalyzed reaction discussed above

favors partitioning of the intermediate T- to products by

formation of T (Equation I-3 or 1-4). If the pK of the

amine nucleophile is not greatly larger than the alcohol

leaving group then, the second order uncatalyzed pathway

(Equation I-2) would not partition as efficiently as the

third order general base catalyzed route. In terms of

activation parameters, the third order pathway for phenyl

esters in general is lower in activation enthalpy than the

second order pathway. However, third order reactions are of

lower activation entropy than second order reactions. The

decrease in activation entropy tends to overcome the decrease





7

in activation enthalpy. As a result the second order path-

way is frequently of lower energy than the third order

pathway.

It is clear from the above discussion that if the

problem of activation entropy reduction could be resolved,

esterolysis reaction rates could be enhanced via the general

base catalyzed route (Equation I-3 or I-4) in some cases.

If the general base and nucleophile were part of the same

molecule, then the general base pathway would no longer be

third order. The entropy of activation reduction would be

resolved largely. Addition of more general base species to

the nucleophile containing molecule should reduce the free

energy of activation for the general base reaction further.

A polymer containing nucleophiles and species capable of

serving the general base function should lower the free

energy of activation for the general base catalyzed reaction

below that of the uncatalyzed reaction. Such a bifunctional

or cooperative interaction should enhance the reaction rate.

The extent of the enhancement would depend on the nucleo-

phile-ester pair studied. As long as the formation of the

intermediate T_ (Equation I-2) is rapid and reversible, the

greater the pKa of the alcohol leaving group relative to the

pKa of the nucleophile, the more effective is the enhance-

ment. Conversely, if the pKa of the alcohol leaving group

is low relative to the pKa of the nucleophile, then the

partitioning of the intermediate T would be very efficient.

In the latter case the general base catalyzed reaction would





8

not be significantly lower, and no enhancement should be

observed.


Polymer Catalyzed Hydrolysis

The initially formed products of the ester aminolysis

reactions are amides. However, in some cases these reactions

can be viewed as the first step of ester hydrolysis reac-

tions. An amine may serve as a nucleophilic catalyst if the

initially formed amide is hydrolyzed at a reasonable rate

(Equation 1-5). In actual practice nucleophilic tertiary

amines, pyridines or imidazoles are usually used for these

reactions. These tertiary amines have a reasonable acylation

rate as well as deacylation rate.1 However, the first step

of the reaction, the aminolysis step, is the point of concern

here. The actual fate of the ester after release of alcohol

moiety is of little consequence. Thus the structure of the

amine nucleophile relative to its deacylation is also of

little consequence. Therefore, the aminolysis reaction by

primary amines (Chapter II) will be used as a model for


Equation I-5




O 2
0 -- ,C + 0
S/ H2

CH3-C CH3 CH -C
OR OR O
OR


aminolysis hydrolysiss) reactions of tertiary amines

(Chapter III)








Overberger et al.6 have published data that they

interpret as being consistent with a bifunctional mechanism.

They studied the hydrolysis of p-nitrophenyl acetate

catalyzed by poly-4(5)-vinylimidazole and imidazole as a

function of the fraction of unprotonated imidazole units6

(Figure I-la). They found that although the imidazole

esterolysis rate increases linearly with a, the polymer rate

increases nonlinearly. They concluded that neighboring

nonprotonated imidazoles interact cooperatively, enhancing

the rate of reaction. At low values of a few nonprotonated

imidazoles are neighbors, and the reaction presumably

proceeds through a noncooperative pathway. As the value of

a increases the cooperative interaction becomes more and

more the predominant pathway, and the reaction rate increases

accordingly. The mechanism of Jencks and Satterthwait

(Equation 1-2, I-4) although different from those proposed

by Overberger et al. (Figure I-lb) is also consistent with

their interpretation. In the terms discussed previously,

the reaction would proceed via the uncatalyzed (Equation I-2)

path as a increases.

It should be pointed out that the bifunctional inter-

pretation of Overberger et al. has not been without criticism.

Kunitake and Okahatab notes that later work on poly-4(5)-

vinylimidazole shows considerable conformation change, with

changes in solvent polarity.7 When the polymer is in its

more compact conformation, it provides a more efficient

catalysis of long chain phenyl esters. He suggests that (p 178)


























11 I I 1
0.60 0.80 1.00


Figure I-la.


Figure I-lb.


Solvolysis of PNPA catalyzed by poly-4(5)-
vinylimidazole (A) and imidazole (0).6
a 28.5 v/v% EtOH-H20, u=0.02, 300C.


Bifunctional mechanisms proposed for poly-4(5)-
vinylimidazole catalyzed hydrolysis of PNPA.6


a
kcat 30




10






11

"the increased catalytic efficiency observed at higher pH's

may similarly be explained on a microenvironmental effect."lb


Summary

In summary, there are two p-nitrophenyl ester aminolysis

reactions of concern. These are the uncatalyzed reaction

(Equation I-2) and the general base catalyzed reaction

(Equation 1-3, 1-4). The general base catalyzed reaction

would be faster than the uncatalyzed reaction (for nucleo-

phile pKa's not much greater than alcohol pKa's) if it were

not third order. The general base reaction can be made

second order by including the general base in the same

molecule as the nucleophile. This effect should be increased

by surrounding the nucleophile with general base functions

as might be found in a polymer system. The reaction of the

appropriate nucleophile ester pair should then be enhanced

in the appropriate polymer due to contribution of the

general base catalyzed reaction. Therefore it is reasonable

to look for the existence of bifunctional or cooperative

effects in polymeric esterolysis reactions based on the

above discussion, if suitable reactants are studied. Failure

to find a bifunctional effect could be due to choice of the

wrong reactants. It is only possible then to attempt to

choose a system in which the polymeric bifunctional reaction

may be unambiguosly observed.








Proposal


Introduction

There are several variables involved in the selection of

the proper system in which to look for bifunctional cataly-

sis. First one must choose the nucleophiles. Nucleophiles

both capable and incapable of bifunctional catalysis are

needed in order to distinguish the two pathways. Second,

the polymer must contain general bases capable of removing

a proton from the T- intermediate. Third, the appropriate


Equation I-6
H
HlH N


N -C OR k[B+H] C N-C --OR
-bI
CH3 CH
T T

ester must be chosen.


Nucleophiles

A goal of the work originally was to design a very

efficient ester hydrolysis catalyst. To this end heterocy-

clic nucleophiles were chosen. Imidazole, due to its

frequent occurrence in esterolytic enzymes, is often studied

O-

SN-C OR

CH3


in synthetic esterolytic polymer systems. Furthermore,

Overberger and Morumoto/ claimed a bifunctional mechanism








for this nucleophile in poly-4(5)-vinylimidazole. It was

decided, then to compare imidazole to other similar hetero-

cycles to determine if it had any special properties in

esterolysis reactions. The similar structure 2-aminopyridine

was also chosen. The T- intermediate formed from 2-amino-

pyridine is structurally capable of losing a proton

(Equation I-6) just as the T intermediate formed from imida-

zole (Equation 1-4). The pyridine nucleophile was chosen as

a nonbifunctional comparison. Unlike 2-aminopyridine and

imidazole, pyridine is structurally incapable of losing a

proton from T to form T Pyridine then cannot react via

the general base pathway but only through the uncatalyzed

pathway.


Polymer

Instead of studying a homopolymer as did Overberger et

al., a polymer system was chosen in which the basic sites

remained constant but the nucleophiles could be varied.

What was needed then was a polymer containing basic units

to which the nucleophilic sites could be attached. The

different nucleophiles then could be studied in the presence

of a general base of constant strength.

Polyethylenimine (PEI as furnished by Dow) seemed to be

the ideal system. Dow PEI 600 (Mw=40,000-60,000) is derived

from cationic ring opening polymerization of aziridine

(ethylenimine) as shown in Equation I-7.8 The structure of

the polymer can best be understood via a discussion of its

formation. If the reaction proceeded as in Equation I-7






14

the polymer would consist of only secondary amines (ignoring

end groups). However the polymer has been shown to contain

primary, secondary and tertiary amines. This observation is

readily incorporated into the mechanism by allowing the

aziridinium ion to react with both primary and secondary

amines. This mechanism would provide a branched polymer.

Every branch produces a tertiary and a primary amino group.

Thus the ratio of primary to tertiary amines is 1:1. The

end groups comprising less than 1% of the total amine

functions are ignored in this calculation.89 The tertiary

amine content has been determined by titration after

exhaustive benzoylation to be 25%. Simple arithmetic

reveals the primary amine content to be 25% and the secondary

amine content to be 50%. Thus the ratio of primary:secondary:

tertiary amines is 1:2:1.

Klotz10 has published extensively on the properties of

PEI and its derivatives. It has been shown that the primary


Equation I-7

+ N + H 2+ H
NH NHN |
NH + N H -- > N NH N- N 2
H H I N
H


amines of PEI are very reactive toward p-nitrophenyl esters.

This is especially true for partially dodecylated and

dodecanoylated derivatives reacting with apolar substrates.1

A great deal of this enhanced reactivity is apparently due to

binding. Of particular interest is a PEI 600 derivative of






15

which 10% of its nitrogens have been dodecylated, and 15% of

its nitrogens have been alkylated with 4(5)-chloromethylimi-

dazole.12 This "synsyme" is reported to approach a chymo-

trypsim in catalytic activity. The acylation rate (excess

nucleo-phile) is 2700 M-1 min-1 for the synzyme with

p-nitrophenyl caproate compared to 10,000 M-1 min-i for a

chymotrypsin with p-nitrophenyl acetate.

The synzyme system of Klotz was chosen as the polymer

to study. Replacement of the imidazole group with pyridine

or 2-aminopyridine is readily carried out owing to the

synthetic versatility of PEI. Klotz et al. have shown that

the polymer may be derivatized by alkylation,12 acylation,12

and reductive amination. 12,13


Ester

P-nitrophenyl acetate and caproate were chosen as

esters to study.11 The p-nitrophenyl esters are convenient

for kinetic analysis. The reactions are readily followed

spectrophotcmetrically.14 Furthermore the great majority

of polymer esterolysis studies have been carried out on

p-nitrophenyl esters.l The caproate ester binds very

strongly to the dodecylated PEI systems and therefore

proceeds through path a of Equation I-1. On the other hand,

the acetate does not bind nearly so well11 and must proceed

largely through path b. The use of these two esters provides

an insight into the binding effects of the polymer as well as

being comparable to the other studies in the area.








Results


It was reasoned that possible bifunctional effects

would become apparent in pH rate profiles of the polymer

esterclysis reactions. With increasing pH a larger and

larger fraction of the backbone nitrogens should be able to

act as general bases to convert the intermediate T- into T

enhancing the reaction rate. If such were the case, the

reaction rate should be pH dependent. If such were not the

case there should be no pH dependence unless polymer charge

density has an effect on rate. Overberger and Salamone,e

though noting some controversy, states ". . it appears that

the varying charge density does not significantly alter the

reactivity of a catalytically active polyion toward a neutral

substrate" (p 218).

The pH rate profiles for the aminoiysis reactions by

the various polymer reactions with p-nitrophenyl acetate

are shown in Figures 1-2, 1-3, I-4 and 1-5. It is clear

that the PEI-D-Pyr-HCl pyridinee system) is pH dependent.

As will be pointed out in Chapter III, the pyridine in

PEI-D-Pyr-HCl remains in the nonprotonated nucleophilic

form throughout the pH range. Furthermore the T interme-

diate formed from pyridine does not have access to T.

Therefore the pH dependence for the PEI-D-Pyr-HCI system

cannot be due either to ionization of the pyridine or a

bifunctional mechanism. In view of Overberger's statement

in this regard and its possible influence on polymer

reactions, the nature of this pH dependency is very






17

important. The remainder of this dissertation is an effort

to understand and quantitate this pH dependency.




















3.0-












2.0-


pH







Figure 1-2. Esterolysis of PNPA by PEI-D-Pyr-HC1 (slope =
0.74). See Chapter IV for experimental
details.



















































I I I

7 8


Figure 1-3.


Esterolysis of PNPA by PEI-D-APyr-HC1 (slope =
0.69). See Chapter IV for experimental details.


3.0-


.r 2.0.













I. 0-


I





















3.0













2.0 -



N









1.0
1.0 -


Figure 1-4.


Esterolysis of PNPA by PEI-D-Im-HC1 (slope =
0.56). See Chapter IV for experimental details.





















A5


3.0











2.0



0






1.0


7


b/
b c


7 8 9


Figure 1-5.


Esterolysis of PNPA by PEI-D-NH -HC1 compared
to esterolysis of PEI bound heterocyclic
systems. See Chapter IV for experimental
details.


PEI-D-Im-HC1
PEI-D-Pyr-HC1
PEI-D-APyr-HC1


-V














CHAPTER II


AMINOLYSIS OF P-NITROPHENYL ESTERS
BY DODECYL POLYETHYLENIMINE



Results and Discussion



Characterization

A large quantity of PEI-600 was alkylated with dodecyl

iodide.0 The PEI-D from this preparation was used to

prepare all the systems studied. This precaution served to

maintain a constant level of dodecylation from system to

system.

The simplest system studied, PEI-D-NH2-HC1, was the

HC1 salt of PEI-D. The high precision of the duplicate

elemental analysis shows that PEI-D-NH2-HC1 is homogenous

(this was true for all systems Table IV-1). The C/N ratic

from elemental analysis can be used to calculate the extent

of dodecylation as described in detail for Table IV-1. The

fraction of backbone nitrogens dodecylated in PEI-D-NH2-HC1

as determined from the C/N ratio, 11%, is in good agreement

with the value from NMR spectral analysis, 10% (Table IV-2)

The other polymer system to be discussed in this

chapter is PEI-D-Ip-HC1. This system was prepared by

exhaustive isopropylation of PEI-D with acetone and NaBH4





23

(Equation II-1). NMR spectral analysis of the extent of


Equation II-1


0
P -NH2 +
CH3 CH3


-H 20

+H2 0
2


N

CH3
3


NaBH
LA~


P H

N

CH3 H CH3


isopropylation was not possible due to interference of the

dodecyl group. However the extent of isopropylation can be

determined from elemental analysis. The C/N ratio for

PEI-D-Ip-HC1 gives a value of 0.32 for the fraction of

nitrogens isopropylated.

Primary amine determinations were made using trinitro-

benzene sulfonate in a modification of the procedure

developed by Satake et al.15 Reportedly this method is

specific for primary amines in the presence of secondary

amines. Primary amine analysis of PEI-D-NH2-HC1 provides

0.20 as the fraction of primary amines, whereas analysis

of PEI-D-Ip-HC1 gives a value of 0.01 (Table II-1).


TABLE II-1


PRIMARY AMINE AND ISOPROPYL CONTENT OF
PEI-D-NH2-HC1 AND PEI-D-Ip-HCl


Polymer % Primary Aminesa % Isopropylationa

PEI-D-NH2-HC1 20 ....

PEI-D-NH2-Ip-HCl 1 32

a Values rounded to the nearest whole %; see Chapter IV for
experimental details.





24.

The above result may be surprising as it requires

isopropylation of secondary amines with NaBH4. Only 20% of

the backbone nitrogens of PEI-D, the starting material, were

primary amines. However after isopropylation 32% of the

total backbone amines were isopropylated. Therefore a

significant number of secondary amines must have been isopro-

pylated.

It has been well established that unhindered ketones

can be aminated with secondary amines using the less reactive

NaCNBH3.6 However examples of this reaction with NaBH4 are

few. The reason for this selectivity seems to involve the

competition for hydride between the ketone(I) and the much

more reactive but less abundant iminium ion (II)


Equation II-2


CH R CH
C3 K \+/ 3
R2NH + 0 C N + H20

I 3 CH3

NaCNBH, or
I 4
NaBH4 NaBH4



CH3 R CH
HO H N-H
CH3 R CH3


(Equation 11-2). The NaCNBH3 reagent does not reduce the

ketone to a significant extent, but the more reactive NaBH4

does.16 Iminium ion reduction must occur to a greater





25

extent under the conditions used for preparation of PEI-D-

Ip-HC1 than usual. This can be explained in part by the use

of a huge excess of both NaBH4 and acetone. In addition the

polymer may very well promote formation of the iminium ion

as well. This possibility is suggested by the study of

acetone H-D exchange, catalyzed by PEI, carried out by
18
Hine. They have shown that PEI (as well as certain

diamines) catalyze H-D exchange, which proceeds via forma-

tion of an iminium ion, much more efficiently than mono-

amines. The observation of secondary amine isopropylation

is important in view of the past usage of acetone/NaBH4 as a

method of determining primary amine content of PEI.10


Basicity of PEI-D-NH 2HC

The acid-base properties of PEI-D-NH2-HC1 were deter-

mined by potentiometric titration (see Chapter IV for

experimental details). The values of a, fraction of nonpro-

tonated polymeric amine, after each addition of titrant, were

given by the concentration of added titrant divided by the

concentration of polymeric amine units, CH2CH2N. The values

of pKa,a at each value of a was calculated from the

Henderson-Hasselbalch equation9 (Equation 11-3, see

Appendix).



Equation II-3


i-a
pK = pH + log-
a,app a







From the plot of pKa,app vs a (Figure II-1) it can be

seen that the basicity, pKa,app, is dependent on the fraction

of polymeric amines protonated, a. A number of factors which

might well be dependent on a probably have an influence on

the PKa,app such as conformation, ion pairing, etc. However,

the important fact here is that the pKa,app is dependent on a

and therefore pH. The dependence of the polymeric pK,app on

a is no surprise if short chain polyamines are used as

models. The pKa values for ethylenediamine, diethylenetri-

amine and triethylenetetramine20 can be determined discretely

for each stage of ionization. The values of these pKa's are

plotted against a in Figure II-. The pKa's of these systems

follow qualitatively the same a dependence trend as the

polymer. Since the polymer is not crosslinked and the

PKa,app dependence on a is similar to the pKa dependence on a

for low molecular weight model compounds, conformations of

the polymer exposing all basic sites are assumed to be in

rapid equilibrium.


Kinetic Analysis

Rates of disappearance of substrate, p-nitrophenyl

acetate (PNPA) or p-nitrophenyl caproate (PNPC), were

monitored spectrophotometrically by observing the increase in

p-nitrophenoxide absorbance. These reactions were studied

under pseudo first order conditions (excess polymer) at pH's

within the range 6.5-9.0. The second order rate constants

(k2 M- min-) were calculated, after subracting background

rate (see Chapter IV), by division by concentration of

primary amine unless otherwise indicated.14






































SI I I I


0.2 0.4


Figure II-l.


0.6 0.8


Comparison of pKa ap dependence on a for
PEI-D-NH2-HC1 witn ow molecular weight
poly-amines.20 (0) PEI-D-NH2-HC1, (A)ethylene-
diamine, (o)diethylenetriamine, (O) triethyl-
enetetramine.





28

Two different polymeric esterolysis processes may

contribute to the pH dependence of p-nitrophenoxide release,

aminolysis and/or general base catalyzed hydrolysis. The

general base catalyzed hydrolysis and nucleophilic or

aminolysis reaction produce different products. The general

base reaction generates the carboxylate ion (Equation 1I-4),

whereas the aminolysis reaction produces the amide21

(Equation 11-5). Both reactions release the species followed

kinetically, p-nitrophenoxide. The general base route was

ruled out as a significant contribution to the rate.

First, the contribution of the general base component

(cf Equation II-4) can be shown to be minimal. Isopropyla-

tion of PEI-D (forming PEI-D-Ip-HC1) converts virtually all

the primary amines to secondary amines. Rate constants

from this system should estimate an upper limit for the

general base component. In order to compare the rate data

for these two systems the second order rate constants (k2GB

must be calculated based on the concentration of basic

sites, CH2CH2N units (Figure 11-2). In comparing the rates


Equation II-4



H
I 0- 0
O- O
O-H
R--C --PNP > R -- C + PNP

PNP p 0O
OH




















75-












50-




E


2 --
















0

6.5 7.5 8.5

pH

Figure 11-2. Comparison of esterolysis rates for (o)PEI-D-
NH2-HC1 and (D)PEI-D-Ip-HCi.






30

for the PNPA reaction it can be seen that the maximum

contribution of the general base reaction is less than 3%.

Secondly, the reaction products can be shown to be not

those of the general base reaction. The p-nitrophenoxide

anion inhibits the reaction of PNPA with PEI-D-NH -HC1

(Figure 11-3). This inhibition is not due to a normal salt

effect (ionic strength) as the ionic strength is 0.1, whereas

the maximum p-nitrophenoxide concentration is 1.5 x 10-4 a

variation of less than 0.2%. Nor is this effect peculiar to

the polyion since the same reaction with ethylenediamine

hydrochloride is also depressed to a similar extent. This

effect then can be compared to the common ion effect of

carbonium ion chemistry. The p-nitrophenoxide ion is acting

as a nucleophile on either an intermediate or product,

reforming starting material. This is not possible in the

case of general base hydrolysis, as neither the tetrahedral

intermediate nor the product, carboxylate ion, is susceptible

to nucleophilic attack (cf Equation 11-4). However the

result of direct attack of the amine on the ester generates

the amide which is susceptible to attack by the p-nitrophen-

oxide ion (Equation II-5).


Equation II-5

0 0- O
11 II
R--C--PNP + P-NH R--C--PNP R--C-NH-P
2 /
P-NH +
PNP













































[PNP ]b X 103


Figure 11-3.


Inhibition of () ethylenediamine, (A)PEI-D-NH2-
HC1 (pH=7.00) and (o)PEI-D-NH2-HC1 (pH=8.82)
esterolysis of PNPA by p-nitrophenol anion
(PNP).

a % Reduction refers to the % the rate constant
is reduced in the presence of PNP from the
rate constant at very low [PNP-].
The PNP concentration is based on p-nitro-
phenyl pK =7.14.22
a





32

Inhibition by p-nitrophenoxide ion has been observed by

Jencks and Gilchrest in 4-methylpyridine catalyzed hydroly-

sis of PNPA. In this example p-nitrophenoxide attacks the

reactive acetylpyridimium intermediate. However, p-nitro-

phenoxide attack on the much less reactive amides described

here was surprising.

Based on the above results some statements can be made




CH3 C
CH3

Acetyl-Pyridinium Intermediate


about the mechanism of the aminolysis reaction. The reaction

proceeds by nucleophilic attack of primary amine on the

ester. The tetrahedral intermediate is assumed but not

required. The p-nitrophenoxide anion is lost in a reversible

step to form the amide. However, the question of pH depen-

dence remains unresolved.

In small molecule systems pH dependencies can often be

explained on the basis of ionization. A protonated amine

cannot act as a nucleophile, and thus the rate is dependent

on the fraction of free amines.23 In the PEI-D-NH2-HC

case correction for ionization does not remove the

dependency. The pH reate profiles for PEI-D-NH2-HC1 (k2

based on primary amine and divided by a) for both PNPA and

PNPC still exhibit pH dependency after correction for

ionization (Figure II-4).













4.0-,


3.0-





.21




0





1.0~


Figure 11-4. Plot of log(k2/a) against pH for PEI-D-NH2-HC1
esterolysis of (o)PNPC and (D)PNPA.





34

The question of pH dependence in polymeric systems has

been explained in several ways. One explanation involves pH

dependent exclusion. As the charge density of the polymer

increases hydrophobic interactions decrease, excluding the

hydrophobic substrate.24 Restated, the binding increases

with increasing a thereby increasing the rate.

A second explanation relies on a bifunctional mechanism.

Overberger et al. have explained rate dependencies on a

(which in turn is dependent on pH) in poly-4(5)-vinylimida-

zole reactions with PNPA as being due to interaction of

adjacent imidazoles. These interactions presumably increase

the reaction rate (Chapter I).

A third explanation invokes the tenet relating nucleo-

philicity to basicity. That is, as the pKa (basicity)

increases so does the nucleophilicity and thus the reactivi-

ty. This explanation was used by Letsinger and Saveride25

for a dependence of rate on a. The relationship between

rate and pKa will be explored quantitatively for the

PEI-D-NH2-HC1 system in the next section via the Bronsted

relationship.


Bronsted Relationship


The Bronsted relationship, relating pKa to rate
a
constants (Equation II-6), has been used extensively in

quantifying esterolysis kinetic data.26 This relationship


Equation II-6


log k/kO = 3 pKa + C
0a.






35

can be applied to a variety of nucleophilic addition and

displacement reactions. However, discussion here will be

limited to ester aminolysis reactions. The values of the

parameters, B and C, can be obtained by plotting the values

of log k2 (corrected for ionization) against pKa for a

series of structurally similar nucleophiles. The value of B

is taken as a measure of the contribution of basicity to

nucleophilicity. An example of such a plot can be seen in

Figure 11-5. In this case the 8 value for a series of

primary and secondary amines reacting with PNPA is 0.83. A

B value of this magnitude indicates considerable sensitivity

to base strength in the reactivity of the nucleophile. The

Bronsted 3 value is not particularly sensitive to reaction

conditions. In Figure II-6 Bronsted plots for imidazoles

and anilines reacting with PNPA at conditions significantly

different than the amine reactions (Figure II-6) still

exhibit a B value of approximately 0.8. The value of C, the

vertical juxtaposition of the parallel lines, is sensitive,

however, to nucleophile structure and to reaction conditions.

The dependence of the nucleophilicity of primary and

secondary amines on base strength raises the question, "Can

a Bronsted relationship be used to explain the pH dependence

of the PEI-D-NH2-HC1 system?" Conventional Bronsted plots

require data from a number of different nucleophiles since

monobasic systems have one pKa and therefore one Bronsted

point, dibasic systems require two pKa's, etc. Conversely,

the pKaa of the polybase PEI-D-NH2-HC1 is pH dependent
a,aFigure -7); therefore a well defined roasted plot can be
(Figure 11-7); therefore a well defined Bronsted plot can be































































Figure 11-5.


/ 9 11

pK a
a


Bronsted plot for the esterolysis of PNPA by
simple primary and cyclic secondary amines.

aValues of k2 and pKa determined at 250C and
1.0 M ionic strength.22


















4-



2-




0-




O
-2



-4



-6 i -II
0 4 8 10 12
a
pK a








Figure II-6. Bronsted plot for the esterolysis of PNPA by
(o)imidazoles and () anilines.

a Values of k2 and pK determined at 300C in
in 28.5% ethanol. 7a















































I
2


I

4


1 I I 8
6 8


pH









Figure 11-7. Plot of PKa,app vs. pH for PEI-D-NH2-HC1.
a/app 2


101


' '





39

constructed from the one system. The Bronsted plots for the

reactions of PEI-D-NH2-HC1 with PNPA and PNPC are shown in

Figure 11-8. Included on the same coordinate system is the

reference plot for the simple primary and secondary amines

discussed earlier. It can be seen from the values of for

PNPA with PEI-D-NH2-HC1, 0.81, and reference, 0.83, that the

difference in pKa dependence is insignificant for the two

systems. Therefore the pH dependence for the PEI-D-NH2-HC1

reaction with PNPA is completely accounted for by the

dependence of nucleophilicity on base strength. However the

value of 3 for the reaction of PEI-D-NH2-HC1 with PNPC, 1.06,

is significantly different from the reference value.

The dissection of the polymeric rate can be facilitated

by assuming that k /a can be separated into two terms, k and
2 a
E (Equation 11-7). If the term E is taken to be the rate of


Equation II-7

log k2/a = log (ka'E) = log ka + log E = 9pKa + C


enhancement due to the polymeric environment, then the term

k can be estimated from the reference amines. That is, if
a
the Bronsted equation for the reference amines (Figure II-5)

is subtracted from the Bronsted equation for the polymer

systems, then E can be determined as a function of pK. and

thus a (Equation 11-8). The value of E then can be plotted

against pKa or a (Figure 11-9).

The binding aspects of the polymer can be examined more

closely via the PNPC reaction. The reactivity of the





















4.0-












2.0-












0.0 -
6.0


10.0


pK
a,app


Figure 11-8.


Bronsted plot for the esterolysis of () PNPA
PEI-D-NH2-HC1.

aThe Bronsted line from Figure II-5 is
included as a reference.


















































'S VV 'S 'S '.4JU


7.0


I 7
7.4


7.8
7.8


8.2


[0.52] [0.60] [0.68]


[0.76] [0.84]


pK
Ka,app

[a]


Figure 11-9.


Plot of E vs. pK or [a] for the esterolysis
of (A)PNPC and (o)PNPA by PEI-D-NH2-HC1.


3000-











2000-


1000-










100-





42

Equation II-8

log k *E = B pK + C P = polymer system

log kR = RPKa + CR R = reference system

log E = E pKa + CE E = polymeric enhancement


reference amines with PNPA is used to dissect the rates of

the PEI-D-NH2-HCI reaction with PNPC. In the case of the

PNPC aminolysis the dependence of rate on PKa,app is not

entirely resolved by the subtraction of the reference

equation as in the case of PNPA. The caproate ester and

acetate ester aminolyses must proceed by the same mechanism

since the only structural change is replacement of a methyl

group by a n-pentyl group. Although this change would not

be expected to effect the reaction mechanism, the hydrophobic

binding properties are greatly affected. Increased hydro-

phobic binding with reduced charge density then, must be the

explanation for the additional pH dependence in the case of

PNPC reaction.

The above discussion adequately accounts for the pH

dependencies of the reactions of both PNPA and PNPC with

PEI-D-NH2-HC1. However, a significant pH independent

enhancement remains unaccounted for. The rate constant for

the PEI-D-NH2-HCI reaction with PNPA at any given pKa
2 a
within the range studied is approximately 160 fold larger

than a reference amine of the same pKa. There are several

possible explanations for this enhancement, for example, a

solvent effect or pH dependent binding. Alternatively, the





43

apparent enhancement may be due to error in concentration of

primary amine and/or in the value of C. Unfortunately there

is insufficient data to determine the cause of this enhance-

ment. In spite of insufficient data, this ambiguity does

not diminish the importance of the pK a,app-rate correlation.


Summary


The reaction of PEI-D-NH2-HCI with PNPA and PNPC has

been shown to proceed via nucleophilic attack of polymeric

amine on ester. The rate of reaction is a function of the

state of protonation of the polymer which in turn is a

function of pH. Correction of the rate constant for fraction

of nucleophiles protonated does not lift the pH dependency.

However, the pH dependency can be quantitatively dissected by

use of the Bronsted relation. The dependence of the PNPA

reaction was found to be due to increased nucleophilicity of

the polymeric amine with increasing pH. The PNPC reaction

exhibits increased binding as well as increased nucleophili-

city with increasing pH. The Bronsted relation has not been

used directly before to quantitate the electrostatic

influence in polyionic nucleophiles. The utility of this

approach will be explored further in the next chapter.













CHAPTER III


HYDROLYSIS OF P-NITROPHENYL ESTERS
CATALYZED BY POLYMER BOUND HETEROCYCLES



Introduction


In this chapter hydrolysis of p-nitrophenyl esters

catalyzed by polymeric systems containing pendent nitrogenous

heterocycles will be discussed. As was pointed out in

Chapter I, primary amines and nitrogenous heterocycles

release p-nitrophenol from p-nitrophenyl esters by the same

mechanism. Therefore, the aminolysis reactions of Chapter II

are used as models for the discussion. The Bronsted

relationship will be employed and its applicability extended.


Polyethylenimine Systems


Preparation

The polymer systems were prepared from PEI-D (described

in Chapter II) and the appropriate derivatized heterocycle.

PEI-D-Im-HCI was prepared by alkylation of PEI-D with

4(5)-chloromethyl imidazole.i0 PEI-D-Pyr-HCl and PEI-D-APyr-

HC1 were prepared by reductive amination of PEI-D using,

respectively, 4-pyridinecarboxaldehyde and N-(2-pyridyl)-3-

aminopropionaldehyde with NaBH4.1 The fraction of the






45

polymer derivatized was determined spectrally by NMR and by

the C/N ratio from elemental analysis (Table III-1).




Table III-1


% PEI UNITS BOUND TO HETEROCYCLES


Polymer % CH2CH2N Units Bound to Heterocycles

Elemental Analysisa NMRb

PEI-D-Im-HC1 264 263

PEI-D-Pyr-HCl 174 122

PEI-D-APyr-HC1 4625 255

aTaken from Table IV-1.
bTaken from Table IV-2.




Primary Amine Analysis

As shown in Figure 1-5, the reaction rates of PEI-D-

NH2-HCl with PNPA are significantly higher than those of

the heterocyclic systems PEI-D-Im-HCl, PEI-D-Pyr-HCI, and

PEI-D-APyr-HC1 with the same substrate. This observation

adds an additional subject for consideration in the inter-

pretation of the kinetic studies for the PEI bound hetero-

cyclic systems. If the PEI bound heterocyclic systems

contained residual nucleophilic primary amines, the kinetic

data might easily be misinterpreted. That is to say, a

small highly reactive fraction of residual primary amines

might be the dominant factor in the esterolysis rates measured.

Such data might lead one to assign spuriously high rates to





46

the PEI bound heterocycle esterolysis reaction. It is clear

that careful consideration must be given to the primary

amine content of PEI bound heterocyclic systems.


Equation III-1

H H

N N N


O NO2 N

NO2 NO2 02


---- L -- N, p
O N N2




The first choice for primary amine detection in the PEI

bound heterocyclic systems might be the trinitrobenzenesul-

fonate reagent as applied to PEI-D-NH2-HC1 in Chapter II..

However the presence of the heterocyclic substituents

presents a possible ambiguity in the use of this reagent.

Imidazole28 and pyridines2 can serve as acyl transfer

catalysts in acylation of amines and alcohols. If these

neterocycles were to serve a similar function as aryl

transfer catalysts30 (Equation III-1) to secondary amines,

the selectivity of the TNBS reagent might well be reduced.

Preliminary experiments gave data consistent with this

prediction. Based on these observations of TNBS in the

PEI-D-Im-HCi systems and the observed secondary amine iso-

propylation in PEI-D-NH2HC1 (Chapter II), no further attempts

were made to find suitable chemical tests. It was decided






47

that it was unlikely that any reagents would give unambiguous

results for the fraction of residual primary amines present

in the PEI bound heterocyclic systems.

Having ruled out chemical methods of primary amine

detection,kinetic methods were attempted. In Chapter II the

exhaustive isopropylation of PEI-D-NH,-HC1 using Acetone/NaBH4

to convert primary amines to secondary amines was discussed.

Such a method, although not providing the number of primary

amines present, would at least provide a PEI bound hetero-

cyclic system free of primary amines. In Figure III-1 the

effects of repetitive isopropylation on the pH rate profile

for the PEI-D-Im-HCI esterolysis reaction with PNPA can be

seen. PEI-D-Im-HC1 was exhaustively isopropylated to PEI-D-

Im-Ip-HC1 with Acetone/NaBH 4 Similar treatment of PEI-D-Im-

Ip-HCI produced PEI-D-Im-Ip2-HC1. The results shown in

Figure III-1 illustrate two important features of the PEI-D-

Im-HC1. The esterolysis rate is very sensitive to isopropy-

lation. The rate and the pH dependence decrease with

increased isopropylation. Further, the rate reduction

decreases with isopropylation from 40% reduction for the

first isopropylation to 15% reduction for the second isopro-

pylation (at pH=8.00). It follows that a third isopropyla-

tion would decrease the rate of reaction by less than 15%

(at pH=8.00). A rate reduction of less than 15% is

considered insignificant, and the pH rate profile of PEI-D-

Im-Ip2-HC1 is assumed to be devoid of a primary amine

component.





















































Figure III-!.


Effects of repetitive isopropylation of
PEI-D-Im-HC1 on PNPA esterolysis reactivity.
(o)PEI-D-Im-HC1, (C)PEI-D-Im-Ip-HC1,
(A)PEI-D-Im-Ip2-HC1.


1








Further Kinetic Analysis

The PEI-D-Im-Ip2-HCI will be considered representative

of the PEI bound heterocyclic systems. It is clear that the

PEI bound pyridine systems show a pH dependency in the pH

rate profile just as was shown for the PEI bound imidazole

system. Furthermore there is very little difference between

the pH rate profiles for the two pyridine containing polymers

(Figure 1-5). Therefore, the two systems must be reacting

by the same mechanism precluding any significant bifunctional

effects for PEI-D-APyr-HCl with PNPA.

As in the PEI-D-NH2-HC1 system (Chapter II), the effect of

the possible PEI backbone reaction with PNPA must be deter-

mined in the reaction of PEI-D-Im-Ip2-HC1 with PNPA. The

rate constants k-GB, which are based on concentration of

backbone amine for PEI-D-Im-Ip2-HC1 and PEI-D-Ip-HC1, are

compared in Figure III-2. The rate constants of PEI-D-Ip-

HCI are only 10% of the PEI-D-Im-Ip2-HC1 reaction, at most.

Thus the backbone reaction can be ignored for the PEI-D-Im-

Ip~-HC1 system. However if the rate of the heterocycle

containing polymer was much slower, the backbone reaction

would be a serious complication.

The state of ionization of the imidazole in PEI-D-Im-

Ip2-HCI cannot be determined directly by spectrophotometric

methods because the amine backbone interferes with the

imidazole absorbance. However, the PEI-D-Pyr-HCI system

does not suffer such interference, and the pendent pyridine

was found to be completely nonprotonated in the pH range

studied. Therefore it is reasonable to assume that the


__




















50-













25
H





-
r-


Figure III-2. Comparison of PNPA esterolysis rates for
(o)PEI-D-HC1 and (D)PEI-D-Ip-HC1.





51

imidazole in PEI-D-Im-HCl is unionized throughout the large

part of the range.

Since the imidazole is virtually nonprotonated in the

region of interest, its pKa is undefined in this region

(Chapter I). Without pKa data a Bronsted plot cannot be

constructed. Even though the pKa data are not available it

is valid to assume that basicity increases with pH just as

it does in the PEI-D-NH2-HC1 system (Chapter II). It was

found in Chapter II that the nucleophilicity, and hence the

reactivity, of the PEI-D-NH2-HCl system was dependent upon

the pKa or basicity of the polymer. By analogy one would

then expect that the reactivity of the PEI-D-Im-Ip2-HC1

would increase with pH as is observed in Figure III-1.


Summary

There are several important statements to be made

concerning the above observations and conclusions. The fact

that the rate of reaction is so dependent on isopropylation

indicates that there may well be a significant contribution

to the esterolysis rate by residual primary amines (backbone

reaction) in the PEI-D-Im-HCl polymer. The residual primary

amine content in the presence of imidazole is difficult to
12
determine as discussed earlier. Klotz et aL2 prepared a

system very similar to PEI-D-Im-HC1. Their polymer only

differed in imidazole content (15%) and degree of ionization.

They tested for primary amine content by the ninhydrin

method and found that it did not exhibit the characteristic

ninhydrin color. This result led to the conclusion that







there were no primary amines present. Such a result is not

surprising, as the parent system (PEI-D-NH2-HC1) which

contains 20% primary amine also does not exhibit the expected

"Ruheman's Purple" (a deep purple blue) of ninhydrin with
29
primary amines.29 The parent system does produce a dull

brown color, possibly an oxidation product, but it in no

way resembles the true ninhydrin color. Further, it would be

expected that their polymer would contain even more primary

amines than PEI-D-Im-HC1 due to a smaller fraction of the

backbone alkylated with methyleneimidazole units, 15% as

opposed to 29%. The presence of primary amines in the polymer
12
studied by Klotz et al.12 might indicate a significant

primary amine component in the presumed imidazole acylation

rate.

Another important point is the ability to qualitatively

predict a pH dependence for the PEI-D-Im-Ip2-HC1 system

based on the Bronsted correlation found for PEI-D-NH2-HC1.

This example begins to show the usefulness of this applica-

tion of the Bronsted relationship. The application to non

PEI systems follows.


Non Polyethylenimine Homopolymer Systems


Unlike the PEI bound imidazole above, there are some

examples of polymer bound heterocycles for which Bronsted

plots can be constructed. Letsinger and Saveride25 studied

poly-4-vinylpyridine catalyzed 2,4-dinitrophenyl acetate

(DNPA) hydrolysis. They found that although 4-methylpyridine

produced the expected linear, rate vs a plot, the





53

poly-4-vinylpyridine exhibited a curved plot (Figure III-3).

They qualitatively attributed the curvature in the polymer

reaction to decreased nucleophilicity with increasing

protonation (decreasing a). Overberger et al.6 found very

similar results in poly-4(5)-vinylimidazole versus imidazole

catalysis of hydrolysis of PNPA (Figure I-la). They

attributed the curvature to several possible bifunctional

mechanisms as discussed in Chapter I. The rate vs a plot

for the reaction of PEI-D-NH2-HC1 with PNPA is shown in

Figure III-4. Similar to the plots for poly-4(5)-vinylimida-

zole (Figure I-la) and poly-4-vinylpyridine (Figure III-3)

the rate vs a plot for PEI-D-NH2-HC1 is curved. Since the

Bronsted relationship was able to quantitatively account for

the PEI-D-NH2-HC1 pH dependence of PNPA esterolysis, it is

reasonable to apply the relationship to the above reactions

of poly-4-vinylpyridine (Figure III-5) and poly-4(5)-vinyl-

imidazole (Figure III-6). As in the case of the PEI-D-NH -

HC1 reaction with PNPA (Figure II-8) the Bronsted plots are

linear. The plot for the poly-4-vinylpyridine system pro-

duces a large value of 3 (1.3); however, small molecule

pyridine 3 values are typically larger than 1.21 Further-

more, the slopes of the pH-rate profiles for the pyridine

containing PEI- systems, 0.69 and 0.74 (Figure I-3 and

Figure 1-2), were larger than the slope of the imidazole

system (0.56, Figure 1-4). The poly-4(5) vinylimidazole

Bronsted slope, 0.8, is precisely that expected from the

portion of the small molecule imidazole plot shown on the

same plot. Therefore the interpretation of Letsinger and













































1i 1 7
0.2 0.4 0.6 0.8 1.
a


Figure 111-3.


Plot of k against a for the esterolysis of
2-4-dinitrophenyl acetate by (0)poly-4-vinyl-
pyridine and (o)pyridine.

a values of k and a determined that 36.8C
in 50% aques ethanol.
in 50% aqueous ethanol.25


1. 0O-.


S0.50-
-








800-



400
/i /
~,~\ o----------------
-.f


0.50 0.60 0.70 0.80


Figure III-4. Plot of k2 against a for the esterolysis of
PNPA by PEI-D-NH2HC1.


























1.40 -


r-i
1.00-
0U

0

0.40


0.20




6.2 6.6 7.0 7.4 7.8
6.2 6.6 7.0 7.4 7.8


PKa, app


Figure III-5.


Bronsted plot for the esterolysis of PNPA by
(A)poly-4(5)-vinylimidazole and (o)imidazoles.

a Values of kcat and pKaa determined at

300C in 28.5% ethanol.'
From Figure 11-6.





























1.8-

1.7-

1.6 -

1.5-

1.4-


3.0 3.2 3.4 3.6 3.8
3.0 3.2 3.4 3.6 3.8


pKa
a,app


Figure III-6.


Bronsted plot for the esterolysis of
2,4-dinitrophenylacetate by poly-(4)-vinyl-
pyridine.

a Values of pKa and k2 determined at 36.8C
in 50% aqueous ethanol.2







Saveride25 would seem to be sufficient for the poly-4(5)-

vinylimidazole system. That is, the curvature in the rate

vs a plot is due to increasing pKa of the nucleophile. The

increase in pKa is in turn due to the reduction of the

positively charged electrostatic field as a increases. At

low values of a the highly charged polymer increases the

energy required for formation of additional positive charge.

As a increases the charge on the polymer decreases, thereby

reducing the energy required for formation of positive charge

and increasing pKa and reaction rate.


Conclusion


The correlation of PKa,app with reaction rate is quite

effective in quantifying pH dependencies in polymer ester-

olysis reactions for pH regions in which the nucleophile is

partially ionized. If the nucleophile is completely nonpro-

tonated the pKa,app cannot be determined. However, the

pKa,app and thus nucleophilicity are dependent on the state

of ionization of the polymer. Therefore it is reasonable

to predict that, if polymeric functions other than the

nucleophile undergo ionization with pH changes, the nucleo-

philicity (and thus rate) will be pH dependent. The pH

dependencies of the heterocycle containing PEI systems can

be explained as a pKa dependence. Bifunctional effects

were ruled out for the pyridine containing PEI systems due

to the similarity of the pH-rate profiles. It also seems

very unlikely that the imidazole containing PEI reaction

with PNPA has a significant bifunctional component in view





59

of the correlation of the pKa,app and rate for poly-4(5)-

vinylimidazole. As pointed out in Chapter I this conclusion

does not necessarily rule out the possibility of a bifunc-

tional mechanism for the polymer systems discussed. The

fault may be in the choice of the ester.

Our application of the Bronsted plot to quantitate pH

dependencies is a very important development in the field of

polymer esterolysis. It has been shown that polymers of

varying make up have different reactivities. For example,

Shimidzu et al.31 have plotted esterolysis rate constants

against the intrinsic pKa (see Appendix) for a series of

imidazole containing polymers. The work of Shimidzu shows

clearly that increasing acrylic acid content in a series of

acrylic acid and vinylimidazole copolymers increases the

imidazole pKa and thus the esterolysis rate. However, up

to now Bronsted correlations for polymer systems have been

for series of polymers. Such correlations cannot account

for pH dependencies as can Bronsted correlations for a single

polymer system plotting pKa,app vs rate constants. This type

of Bronsted correlation has proven valuable in explaining

both the pH dependencies generated by the research herein as

well as that of others. Perhaps the best test of a tool is

how well it works. The correlation of pKa,ap to esterolysis
a,app
rate constants has been shown here to work well for the

polymer systems tested.













CHAPTER IV


EXPERIMENTAL



Introduction



Melting points and boiling points are uncorrected unless

otherwise noted. Melting points were determined with a

Thomas Hoover Unimelt capillary melting point apparatus.

Boiling points were determined by conventional techniques.

Nuclear Magnetic Resonance spectra were recorded using a

Varian A60-A spectrometer. Chemical shift (6) data were

reported in parts per million (ppm) from the appropriate

internal reference, either tetramethylsilane (TMS), 2,2-di-

methyl-2-silapentane 5-sulfonic acid (DDS), or water d1

(HOD). Solvent evaporation was performed at reduced pressure

using a Buchler Instruments flash evaporator. Lyophilisation

or freeze drying was carried out using an apparatus and

technique similar to that described by Vogel.32 The aqueous

dialyses were carried out in a 3 liter filter flask connected

to a deionized water tap and equipped with a stir bar. The

polymer solutions to be dialyzed were transferred to dialysis

bags prepared from Union Carbide (36 100 ft dialysis membrane)

tubing according to the method of Gabbay et al.33 and

generously supplied by Shau-Fong Yen. In the case of






61
Dialysis against a stream of deionized water, the tap was

opened to allow a moderate flow of water through the flask

while stirring. In the case of dialysis against 0.1 M HC1,

the tap was closed, and 30 ml concentrated hydrochloric acid

was added. When dialyzing against ethanolic solvents the bag

was stirred in a 500 ml flask with the solvent. Temperature

control was provided by a Lauda K-2/R thermostat. All UV-

visible spectrophotometric measurements were made on a

Cary 17D spectrophotometer. The pH measurements were made

with a Beckman Research pH meter in conjunction with a

Radiometer GK 2321C electrode.


Syntheses


Dodecyl Polyethylenimine PEI-D

Dow PEI 600 polyethylenimine solution (128 g) was freeze

dried for 8 hr and then allowed to warm to room temperature

while maintaining vacuum (0.05 torr) for an additional 36 hr.

The gelatinous material (49 g, 1.2 mol) was dissolved in 470 ml

argon saturated anhydrous ethanol. Dodecyl iodide (35 g,

0.12 mol) in 5.0 ml of anhydrous ethanol was added to the

polymer solution. The reaction flask was sealed and main-

tained at 450C for 120 hr. The reaction mixture was quantita-

tively transferred to a one liter volumetric flask and diluted

to the mark with argon saturated anhydrous ethanol. This so-

lution was stored under argon and used as a stock solution in

the following preparations (1.2 M in ethylenimine units).





62

Dodecyl Polyethylenimine Hydrochloride PEI-D-NH -HCI

A solution of dodecyl polyethylenimine was prepared by

diluting PEI-D (25 ml, 0.030 mol) up to 150 ml with absolute

ethanol. This solution was added carefully to a 500 ml

beaker containing 5 ml concentrated hydrochloric acid in

50 ml absolute ethanol at 0C. The resultant salt was

filtered and washed with absolute ethanol.


Dodecyl-4-methylenepyridine-polyethylenimine Hydrochloride
PEI-D-Pyr-HC1

Freshly distilled 4-pyridinecarboxaldehyde (BP 191-192C

under N2, 0.96 g, 0.0090 mol) was dissolved in 25 ml absolute

ethanol in a 125 ml erlenmeyer flask equipped with a stir bar.

PEI-D (25 ml, 0.030 mol) was added to the flask. The cloudy

suspension was stirred for 2 hr under N2. A 50 ml solution of

NaBH (0.13 g, 0.0022 mol) was added to the flask at 0C. The

reaction was stirred for 2 hr. The reaction mixture was trans-

ferred to 4 tubes. One milliliter concentrated hydrochloric

acid was added to each tube to form the polymer salt, and the

tubes were then centrifuged. The precipitated polymer salt was

washed twice with absolute ethanol. The polymer salt was dia-

lyzed against 500 ml absolute ethanol twice. The ethanolic dia-

lysate was passed through a sephadex LH-20 column. The eluant

was transferred into 4 tubes. The salt was reprecipitated by

addition of one milliliter concentrated hydrochloric acid to

each tube. The tubes were centrifuged, and the salt was

washed twice with absolute ethanol. The polymer salt was

dryed producing a white powder (1.12 g, 0.00966 mol, 32%).








Dodecyl-N-(2-pyridyl)-3-propylamine-polyethylenimine Hydro-
chloride PEI-D-APyr-HCl

Ten milliliters of an absolute ethanol solution of N-(2-pyr-

idyl)-3-aminopropionaldehyde hydrochloride (0.65 g, 0.0031 mol)

was prepared in a 125 ml erlenmeyer flask with gentle heating.

A stir bar and PEI-D (10 ml, 0.012 mol) was added to the flask

which was stirred for one hour under N2. A 10 ml absolute eth-

anol solution of NaBH4 (0.092 g, 0.0024 mol) was added to the

reaction mixture. After stirring for an hour, the reaction

mixture was transferred to a tube. The polymer salt was preci-

pitated by addition of one milliliter concentrated hydrochloric

acid. After centrifuging, the polymer salt was washed twice

with absolute ethanol. The polymer salt was dialyzed against

a stream of deionized water for 24 hr. The polymer salt was

then dialyzed against 500 ml absolute ethanol overnight. The

dialysate was passed through a sephadex LH-20 column. The

polymer salt was reprecipitated by the addition of one milli-

liter of concentrated hydrochloric acid. The polymer salt was

dried producing an off-white solid (0.35 g, 0.0023 mole, 19%).


Dodecyl-4(5)-methylenimidazole-polyethylenimine Hydrochloride
PEI-D-Im-HCI

To a solution of PEI-D (25 ml, 0.030 mol) in a sealable tube

was added 4(5)-chloromethylimidazole hydrochloride (1.59 g,

0.10 mol), triethylamine (2.5 g, 0.025 mol) and a stir bar.

The tube was sealed and maintained at 650C in an oil bath for

36 hr. After cooling the tube was opened and the contents

transferred to two test tubes. One milliliter of concentrated

hydrochloric acid was added to each tube forming a heavy








precipitate. The tubes were centrifuged and the supernatant

decanted. The precipitates were washed with absolute ethanol,

centrifuged, and the supernatant decanted. The precipitate

was dissolved in 20 ml deionized water and dialyzed against a

stream of deionized water for 72 hr. The polymer was then

dialyzed with 500 ml absolute ethanol three times. The poly-

mer solution was removed from the dialysis bag and passed

through a sephadex LH-20 column. The salt was reprecipitated

with 4 ml concentrated hydrochloric acid. The salt was washed

twice with absolute ethanol. The polymer salt was dried

producing an off-white solid (1.75 g, 0.0143 mol, 48%).


Dodecyl-4(5)-methylenimidazole-isopropyl-polyethylenimine
Hydrochloride PEI-D-Im-Ip-HC1

PEI-D-Im-HC1 (0.0300 g) was dissolved in 30 ml of water.

This solution was treated with solid NaBH4 to bring the pH up

to approximately 7. After freeze drying the polymer was sus-

pended in 50 ml absolute ethanol containing one gram acetone

by the dropwise addition of 6 N hydrochloric acid. After one

hour NaBH4 (0.4 g, 0.01 mol) in 10 ml absolute ethanol was added

causing precipitation. A routine of suspension of the polymer

with 6 N hydrochloric acid followed by addition of acetone

(5 g, 0.8 mol), then one hour later addition of NaBH4 (0.4 g,

0.01 mol) in 10 ml absolute ethanol was followed twice. The

polymer was reacidified with 6 N hydrochloric acid and allowed

to stir for 12 hr to complete hydrolysis of the residual

NaBH4. The solution was dialyzed for 24 hr. The polymer

solution was centrifuged then filtered. The filtrate was

freeze dried producing a white solid (0.281 g, 94%).






65

Dodecyl-4(5)-methylenimidazole-isopropyl-isopropyl-polyethyl-
enimine Hydrochloride PEI-D-Im-Ip2-HCl

PEI-D-Im-Ip-HC1 (0.098 g) was dissolved in 10 ml of

water. This solution was heated with solid NaBH4 to bring

the pH up to approximately 7. After freeze drying the

polymer was stirred in 20 ml absolute ethanol containing

one gram acetone for one hour. The polymer was suspended by

dropwise addition of 6 N hydrochloric acid and 5 ml absolute

ethanol containing one gram acetone was added. After one

hour NaBH4 (0.4 g, 0.01 mol) was added in 10 ml of absolute

ethanol followed by suspension of the polymer salt with

6 N hydrochloric acid one hour later. The reaction mixture

was stirred under a stream of N2 overnight. The polymer was

dialyzed against a stream of deionized water for 24 hr

followed by 0.1 N hydrochloric acid for 3 hr. The polymer

solution was centrifuged, filtered then freeze dried

producing a white solid (0.062 g, 63%).


4(5)-Chloromethylimidazole Hydrochloride

4(5)-Chloromethylimidazole hydrochloride was prepared

according to the procedure of Turner et al.3 from

4(5)-hydroxymethylimidazole and thionyl chloride (mp 138-

1400C, lit mp 138-1410C).


4(5)-Hydroxymethylimidazole Hydrochloride

In a mixture of 50 ml concentrated hydrochloric acid,

125 ml water and 250 ml benzene 4(5)-hydroxymethylimidazole

picrate (69 g, 0.21 mol available from Eastman) was

dissolved with heating. The benzene which developed a






66

yellow color was decanted. The aqueous solution was

continuously extracted with 125 ml benzene layer for 9 hr.

After decanting the benzene layer, the aqueous layer was

evaporated under vacuum at 700C. The yellow brown residue

was recrystalyzed from ethanol ether (23 g, 81% mp 103-1060C,

lit mp 107-1090C).


3-Bromopropanal Dimethyl Acetal

3-Bromopropanal dimethyl acetal was prepared according

to the procedure of Pineau:35 bp 520C/10 mm Hg, lit35 bp

590C/12 mm Hg; nmr CDC13 6 2.12 doublett of triplets, J=5 Hz,

J=8 Hz, 2H), 3.36 (singlet, 6H), 3.43 (triplet, J=8 Hz, 2H),

4.53 (triplet, J=5 Hz, 1H).


N-(2-pyridyl)-3-aminopropanal Dimethyl Acetal

The synthesis of N-(2-pyridyl)-3-aminopropanal dimethyl

acetal was carried out according to the procedure described

by Reynaud et al.36 The product was an amber oil:

bp 1540C/10 mm Hg, lit bp 1530C/12 mm Hg; nmr (CDC31) 6

1.92 (quartet, 2H), 3.32 (singlet, 6H), 3.35 (quartet, 2H),

4.50 (triplet, 1H) 4.75-5.20 (mult., 1H) 6.25-6.75 (mult, 2H),

7.20-7.55 (mult., 1H), 8.05 doublett of doublets, 1H).


N-(2-pyridyl)-3-aminopropanal Hydrochloride

N-(2-pyridyl)-3-aminopropanal dimethyl acetal (1.3 g,

0.0064 mol) was dissolved in 5 ml concentrated hydrochloric

acid and heated on the steam cone for one hour. The solvent

was evaporated and the brown residue was dried under vacuum

(0.40 g, 0.0018 mol, 28%); mp 172-172.50C.






67

p-Nitrophenyl Acetate PNPA

P-nitrophenyl acetate was prepared according to the

procedure of Bender and Nakamura.37 The solid p-nitrophenyl

acetate was recrystalyzed twice from ethanol; mp 78-78.5C,

lit mp37 77.5-780C.


p-Nitrophenol Caproate PNPC

P-nitrophenol (14 g, 0.1 mol) was dissolved in 25 ml dry

pyridine. The pyridine solution was added to caproylchloride

(13 g, 0.1 mol) in a 100 ml round bottom flask equipped with

a reflux-condenser and stir bar. The reaction was refluxed

for 15 hr. After cooling, the contents of the reaction

flask were added to 40 ml of ice water. The resulting oily

layer was separated and the aqueous layer extracted twice

with 25 ml diethyl ether. The combined oil and ethereal

extract was washed 3 times with 50 ml water, 5 times with

50 ml 5% aqueous hydrochloric acid, and 6 times with 5%

aqueous sodium bicarbonate (each washing had to be salted

out). The ethereal layer was dried and stripped producing

17 g of crude material. The crude product was distilled

producing a viscous oil (13 g, 0.055 mol, 55%); bp 1500C/2 mm
38
Hg, lit38 bp 1450C/1 mm Hg.


Composition of PEI Derivatives by Elemental Analysis

Elemental analysis data were used to calculate the

values of D, (fraction CH2CH2N units dodecylated), z (frac-

tion of CH2CH2N units bearing a heterocycle), mole unit

weight (average weight of a mole of CH2CH2N units after








derivitization), and a initial (in the case of PEI-D-NH2HC1).

The C/N ratio was used to calculate both D and Z. The carbon

to nitrogen weight ratio is equal to the C/N ratio from

elemental analysis. Therefore the number of carbons per unit

(2 from the CH2CH2N unit, 12D from the dodecyl group, and

ZX from the heterocycle, X is the number of carbons in the

heterocycle) divided by the number of nitrogens per unit

(one from the CH2CH2N unit and ZY from the heterocycle, Y is

the number of nitrogens in the heterocycle) multiplied by

the C/N atomic weight ratio is equal to the C/N ratio from

elemental analysis (Equation IV-1). This equation is

readily rearranged to solve for either D or Z. In the case

of PEI-D-NH2-HCI the heterocycle terms drop out and the

equation is in a single unknown, D. For the PEI derivatives

containing heterocycles the value of D calculated for PEI-D-

NH2-HCI was used. The error limits in these calculations

are based on an assumed error of 0.2 in the elemental

analysis. Values of C/N were calculated using C+0.2/N-0.2

and C-0.2/N+0.2 which were substituted in the appropriate

equations to calculate error limits. The mole unit weight

was calculated from the relationship that % of an element in

a compound is equal to the weight of that element in the

compound divided by the formula (units) weight of the

compounds multiplied by 100 (Equation IV-2). The elemental

analysis values for carbon and nitrogen together with the

calculated values for D and Z were used for these calcula-

tions (Equation IV-30). The error calculation for the unit






69
mole weight was based on the difference between the average

from the carbon and nitrogen calculations and the highest

possible value from the carbon based calculation and lowest

possible value for the nitrogen based calculation consistent

with the assumed error of 0.2 in elemental analysis data.

The value of a initial was calculated for the PEI-D-NH2-HC1

titration based on the assumption that all analyzed chlorine

was present as the hydrochloride salt. The ratio of gram

atoms of chlorine to gram atoms of nitrogen in the polymer

represents the fraction of CH2CH2N units protonated, 1-a

initial (Equation IV-4). The error calculation for a initial

was made assuming an error of 0.02 in elemental analysis data

as in previous calculations (Table IV-1).


Equation IV-1



C/N = 2 + 12D + ZX 12.01
1 + zY 14.001


D = C/N -1. (1 + ZY) 2 + 2X 1-
142.01 12


D = C/N 12.01 2 1
12.01 2 -



C/N 1201 2 120
Z = 12.
14.01
S- C/N 14.01
12.01



Equation IV-2


% element = weight of element X 100
formula weight (mole unit weight


























































































0-
0

U 0
OO


4-i

SC

U

r-
0
0
C


>1

IU


H
r4


r 4


0


I +I




r~





N
'1


+I











0
N




C
0

0
I I


ON










0
C,,


H
H

CU7

H


04
H
I r-1


;a4
H
a







N



U
IC
au


H
A
i.


a 3

0

a t
I
H m




O I



U H

c
04 H>
0 01
-I >
U z z




> 1 i





04-;44r-
0 0 (
( u









4000
i4 0 m





0 0 -H
'0

m rU U






4-0


00 0-


H c ai
3 u

(U ^1 ^
C 41 41
rH 3 Tl U


'C
cc
-i-
I)
C
4-

r0d
Q

n3 m

a) >il
6
a)E
m
H <


'0


H10



d 5
U r





U


rl
0








H N
U
QI .


I c'J






S*H *>1


0I Q
H










- 00
a
I 0-I





1 U J






MO
O
0-







>Oa
'0 0





0a
4-1 (U S-
o~ u en












Sr N n n CN
I m' 0 N 0 N 0

> H o0 en 0 0o
I I I I I I .
I N 0 ,0 0 H. 4J
O- o en n x 10
H 0 0 C r 0 0 >1
W i


04



S 44
4-3 rn
a u u00



I N 0 a, o 4-


I * c .
r H o 0 0. o 0 tq-.
















M m H "
S I o ern e- 0 n N3m
S n no CN















O) 0
E-I 4 0 r d a
S* *















H 0(0
HH- o en o aen 0
I *
e -HOC





















-1 0
S 4 H I- I o I I I o w0
a0N0 0C H
IU 0 0 N 4

C\ O N >

0 Em)





H 5 0








H W014
I 0 0 r 0 (D
S N * * W U
or 0 0' d0)


0Z < 4 P 3>
H rH rl rH rH 0 U-lH1O



1 01 -

'0) U

< 0 0) U l u

0 01 0 0 U -H m U 4 r0) 0



a; < cn ; a f h a: aja






73

Equation IV-3



(2 + 12D + ZX) 12.01
mole unit weight = (2 120
%C X 0.01



mole unit weight = (1 + ZY) 14.01
%N X 0.01



Equation IV-4



1 = %Cl/35.45
%N/14.01


Composition of PEI Derivatives by NMR Spectral Analysis

The NMR spectra and peak areas (see Figure IV-1) of the

polymer salts were recorded (Table IV-2). The integrated

areas per hydrogen (for example, area of imidazole peak

divided by 2) was used to calculate the extent of substitu-

tion on the polymer. In the case of PEI-D-NH2-HC1 the frac-

tion of CH2CH2N units alkylated was determined by the ratio

of dodecyl area per hydrogen to the CH2CH2N area per hydrogen

(corrected for the N terminal methylene of the dodecyl

group). In the heterocycle containing systems the dodecyl

peak was used as a reference instead of the CH2CH N peak.


Determination of Primary Content of PEI Derivatives Using
Trinitrobenzene Sulfonate (TNBS)

The primary amine content of PEI derivatives was

determined using trinitrobenzenesulfonate (TNBS) by a method

similar to that used by Johnson and Klotz.10 To a 25 ml





74

volumetric flask was added 0.100 ml of an aqueous PEI solu-

tion (0.0001 units mol), 5 ml 4% aqueous sodium bicarbonate

solution, and one milliliter 0.1% aqueous TNBS solution.

The flask was heated 0.5 hr at 370C, cooled to room

temperature, filled to the mark with galatial acetic acid,

and the absorbance measured at 340 nm. The percentage of

residual primary amine was calculated based on the extinction

coefficient used by Johnson and Klotz,0 12000 M1

(Table IV-3).


TABLE IV-3


PEI

PEI

PEI

Bla


PRIMARY AMINE CONTENT OF PEI DERIVATIVES USING TNBS

Unit Mole % Primary
Derivative Absorbance Concentration of PEI Amine

-DNH2-HC1 0.9720.013 4.08 x 10-" 19.80.4

-D-Ip-HC1 0.0580.001 3.44 x 10-4 1.40.2

Ink 0.0070.007


Primary Amine Detection by Ninhydrin

The efficacy of primary amine detection by ninhydrin was

compared for PEI-D, ethylenediamine, and glycine according to

the procedure suggested by Pasto and Johnson.39 A strip

of filter paper and a strip of chromatogram sheet (silica

gel) was spotted with 0.1 M aqueous solutions of the amine

systems to be compared. After oven drying the strips were






75

sprayed with 0.25% ninhydrin solution and developed for five

minutes at 1000C. Additionally 0.1 M solutions of the amines

to be compared were treated with one milliliter of 0.25%

ninhydrin solution and gently warmed for 0.5 hr. The glycine

results were those expected: blue solution39 and purple blue

spots.40 Ethylenediamine did not react with the ninhydrin

and produced no color changes; similar results have been

previously reported.41 The PEI-D spots were brown and the

solution was yellow.


Relative Concentrations of Imidazole Containing Polymer
Solutions

The relative concentrations of stock solutions of the

three imidazole containing polymers PEI-D-Im-HC1, PEI-D-Im-

HC1, and PEI-D-Im-Ip2-HC1 used for kinetics were determined

spectrophotometrically. A series of 7 solutions were

prepared from the PEI-D-Im-HC1 stock solution in 1.0 M

hydrochloric acid. The absorbencies of these solutions

ranging in concentration from 2.24 x 10-5 to 28.0 x 10-5

molar in imidazole units were measured at 210 nm and

25.00.30C. A Beer's law plot was constructed from this

data. The least squares line passed through 0 (within

experimental error r=0.99998) and had a slope, extinction

coefficient,of 6.79 x 10 M-1. Solutions (2 each) of the

isopropylated imidazole containing polymers were prepared

from the respective stock solutions in the same way

(ca 105 M) and their absorbancies measured. The concentra-

tions of the stock solutions were calculated from the

extinction coefficient of the PEI-D-Im-HCl system. The


1






76

concentrations of PEI-D-Im-Ip-HC1 and PEI-D-Im-Ip2-HC1

solutions then were determined relative to the PEI-D-Im-HC1.


Potentiometric Titration of PEI-D-NH2-HC1

The PEI-D-NH2-HC1 (0.0931 g) was dissolved in 50.0 ml

of 0.100 M KC1 after drying under vacuum for 10 days. The

solution (0.0192 M in CH2CH2N units) was titrated with CO2

free KOH (0.09977 N) at 25.00.30C under a nitrogen

atmosphere. After each addition of titrant the pH was

recorded upon equilibration.

The values of pKa,app and a were calculated from the
a,app
titration data. The value of a at each pH was determined

by a modification of the method suggested by Albert and

Serjeant42 to calculate the ratio of nonprotonated species

(Equation IV-5). The value of KOH refers to the initial


Equation IV-5


[KOH] + int [CH2CH N] 10-H + 10pH-14.00

[CH2CH2N] + 10-pH 10 pH-14.00



concentration of potassium hydroxide. That is, the total

moles of titrant added to bring the pH to a given value

divided by the volume of the solution. The value of a.
int
is the fraction of nonprotonated CH2CH2N units in PEI-D-NH2-

HC1 as prepared and determined from the Cl/N ratio. The

exponential terms in pH serve to correct the hydroxide ion

concentration for water ionization. The value of pKapp
a,app
was calculated at each value of pH by use of the Henderson-

Hasslebalch equation9 (Equation 11-3, Table IV-4).








TABLE IV-4


POTENTIOMETRIC TITRATION DATA FOR PEI-D-NH2-HC1


Titrant (ml)a


0.000
0.500
1.000
1.640
2.000
2.500
3.000
3.500
4.000
4.500
5.000
5.500
6.000
6.500
7.000
7.500
8.000
8.500


3.00
3.19
3.47
4.14
4.64
5.45
6.05
6.62
7.17
7.63
8.00
8.32
8.64
8.96
9.34
9.75
10.25
10.73


0.199
0.233
0.269
0.321
0.356
0.407
0.459
0.510
0.562
0.614
0.666
0.718
0.770
0.821
0.873
0.922
0.967
0.996


a 0.09977-0.09% CO2 free


pKapp
a,app


3.61
3.71
3.91
4.46
4.89
5.62
6.12
6.60
7.06
7.43
7.70
7.91
8.12
8.30
8.50
8.67
8.78
8.33


aqueous KOH.






78

Equation II-3


1-n
pK = pH + log



Preparation and Standardization of CO2 Free 0.1 N KOH Titrant

An ampule of J. T. Baker Dilut-it was quantitatively

transferred to a 1000 ml volumetric flask as per enclosed

instructions. The contents of the volumetric flask were

diluted to the mark with "boiled out" water. The titrant

was standardized with oven dried (1200C, one hour) primary

standard pottasium hydrogen phthalate (KHP). The standardi-

zation was repeated seven times (N=0.099770.09%).


Preparation of Ester Solutions Used for Kinetic Studies

Either p-nitrophenyl acetate (0.010 g, 5.5 x 105 mol)

or p-nitrophenyl caproat (0.020 g, 8.4 x 10 ) was added to

a 10 ml volumetric flask. The flask was filled to the mark

with acetonitrile dried by refluxing 3 hr over P 25 followed

by distillation. The ester solutions were stored in tightly

sealed brown bottles.


Preparation of Polymer Solutions for Kinetic Studies

Into a dry tared 10 ml volumetric flask was weighed

0.1 g of the polymer salt. The flask and contents were

dried for three days under vacuum. After drying the flask

was reweighed to determine the dry polymer weight. The

flasks were then partly filled with glass distilled water

and shaken to dissolve the polymer. After standing overnight





79

to complete the dissolution as well as to allow foam to

dissipate, the flasks were filled to the mark. The solutions

were shaken and stored in tightly closed brown bottles.


Buffer Solutions

Tris(hydroxymethyl)aminomethane, tris, was used as the

buffering agent in the kinetic experiments. Two stock

buffer solutions were prepared both 0.1 M in tris and ionic

strength. A solution of tris HC1 was prepared by adding

primary standard tris (12.14 g, 0.1000 mol) and the contents

of an acculate 1/10 N hydrochloric acid ampule to a 1000 ml

volumetric flask and diluting to the mark with glass

distilled water. A solution of tris was prepared by adding

primary standard tris (12.14 g, 0.1000 mol) and certified

ACS potassium chloride (7.46 g, 0.100 mol) to a 1000 ml

volumetric flask and diluting to the mark with glass

distilled water. These solutions (both 0.100 M in buffer

and ionic strength) were combined to form solutions of the

correct pH.


Kinetic Method

To a 4 ml cuvette was added polymer or ethyenediamine

stock solution, water and acetonitrile. Amounts of water

and acetonitrile were added such that the total volume of

polymer stock solution and water was 0.100 ml and the total

volume of acetonitrilic ester solution (added later) and

acetonitrile was 0.050 ml. Buffer solution (3.00 MI, 0.100 M

tris, I=0.100 M) was pipetted into the cuvette. The pH of

the solution was recorded at room temperature, 2520C. The








cuvette was placed in a spectrophotometer and equilibrated

to 25.0=0.3C. The ester solution was placed on a glass rod

with a flattened tip which was used to add the ester and stir
14
the solution.14 The rate of increase in p-nitrophenoxide

anion concentration was observed at 400 nm until no change

in absorbance, A, was observable to determine A-. The

observed rate constants were calculated from the values of

A -At and 6 (time) using a least squares routine of a TI-58

programable calculator. Only the initial portion of the

reaction was used to determine the rate constants due to

p-nitrophenol inhibition. The rate constants were then

corrected for the background rate by subtraction of the value

of the background rate constant at the appropriate pH. These

corrected pseudo first order rates were then divided by

concentration of CH2CH2N units, nucleophile concentration or

protonated nucleophile concentration (PEI-D-NH2-HC1) to

determine kGB, k2 or k /a, respectively. In those cases

where the effect of p-nitrophenol was to be observed, the

p-nitrophenol in acetonitrile was added before the buffer,

but the total acetonitrile content remained 0.050 ml or 1.6%

(Tables IV-5 through IV-12).


Background Rate

The rate of p-nitrophenyl acetate esterolysis without

polymer was measured at 7 pH values in the region of interest.

The background rate constants (k BG) were found to fit Equa-
43
tion IV-6 (Figure IV-2). The value of k

(7.0 x 102 M1 min-), k44 (5.70 x 102 M min ) and k43
SOH






81

Equation IV-6


kb[trisil
k k b triTotal + k (1014-PH) + k [H 0]
kBG + 108.08-pH kOH10 + 20



(6 x 1-2 M min ) were available from the literature.

The values from this equation were then used for the back-

ground rate.


Pyridine Ionization in PEI-D-Pyr-HC1

The pyridine in PEI-D-Pyr-HC1 exhibits a pH dependent

absorbance at 259 nm. In 0.1 N hydrochloric acid solution

the extinction coefficient based on 2 measurements was

(5800400). In 0.1 M tris buffer solution at pH's 6.71,

7.28, and 7.75 the extinction coefficients were found to

be 3020, 3030, and 2970, respectively. These values

indicate that the state of ionization of pyridine is not

changing over pH 6.71 under these conditions.

































































Figure IV-2.


Plot of background PNPA esterolysis rate. The
line was calculated from Equation IV-6.






83

TABLE IV-5


ESTEROLYSIS RATES FOR PEI-D-NH2-HC1


pH a k2 M min- PEI-D-NH -HCl]a [PNP]b [PNPA]c


6.94
7.38
7.44
7.68
8.28
8.47
8.48
8.48
8.49
8.49
8.55
8.52
8.80
8.48
8.48
8.66
6.96
7.68
8.30
8.80
7.00
6.85
7.38
8.32
8.82
8.82
7.00
7.39
7.53
7.70


% Reac-
tion
[PNPC]c Followed


41
84
88
150
430
460
440
490
480
490
500
450
750
460
400
530
32
130
320
780
32
25
68
250
440
490
710
1600
2200
3300


a Primary amine concentration
PEI-D-NH2-HC1 (x 104).


based on 2% primary amine content in


b Initial p-nitrophenol concentration (x 105).
c Ester concentration (x 105).


6.8
13
6.8
6.8
6.8
6.7
13
13
13
6.8
6.8
6.8
6.8
6.8
6.8
6.8
6.8
6.8
6.8
6.8
6.8
6.8
6.8
6.8
6.8
2.8
6.9
6.9
6.9
6.9








TABLE IV-6


ESTEROLYSIS RATES FOR PEI-D-Ip-HCl


pH kGB M-min


7.40 .59 2.7
7.70 .96 2.7
8.34 2.6 2.7
8.82 5.2 2.7

a Concentration of CH2CH2N units (x 103)
b Ester concentration (x 10 ).


S% Reaction
PNPA] Followed

2.0 36
2.0 20
2.0 7
2.0 27


TABLE IV-7


ESTEROLYSIS RATES FOR PEI-D-Im-HC1


[PEI-D-Im-HCl]a

8.9
8.9
8.9
8.9
8.9
8.9
8.9
8.9
8.9


[PNPA]b

2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0


% Reaction
Followed

20
81
93
15
93
12
93
19
97


a Concentration of imidazole based on 28%
b Ester concentration (x 105).


imidazole (x 103).


pH

6.92
6.97
7.36
7.65
7.98
8.30
8.58
8.81
8.81


k M- min-

37
35
65
96
160
290
290
400
380


[


[PEI-D-Ip-HCl]a








TABLE IV-8


ESTEROLYSIS RATES FOR PEI-D-Im-Ip-HCI

-1 -1 % Reaction
pH 2 n [PEI-D-Im-Ip-HCl]a [PNPA] Followed


6.90
7.32
7.96
8.59
8.80


a Concentration of imidazole based on 28% imidazole (x 10).
b Ester concentration (x 10 )


TABLE IV-9


ESTEROLYSIS RATES FOR PEI-D-Im-Ip -HC1



pH k2 M min-1 [PEI-D-Im-Ip2-HCl]a [PNPA]b


6.72
7.58
8.25
8.54
8.75


% Reaction
Followed

94
81
96
95
83


a Concentration of imidazole based on 28% imidazole (x 10).
b Ester concentration (x 10 )







TABLE IV-10


ESTEROLYSIS RATES FOR PEI-D-APyr-HC1


pH k2 M min
pH 2


7.00
7.26

7.28
7.32
7.34
7.36
7.36
7.38
7.40
8.02
8.42
8.45


[PEI-D-APyr-HCl1]a

7.2
13

8.9
4.5
17
8.9
4.5
13
7.2
7.2
17
13.1


S% Reaction
[PNPA] Followed


a Concentration of 2-aminopyridine
dine in PEI-D-APyr-HC1.
b Ester concentration (x 105)


based on 28% 2-aminopyri-








TABLE IV-11


ESTEROLYSIS RATES FOR PEI-D-Pyr-HCl


H k2 M -min-

6.83 11
7.22 22
7.26 22
7.32 24
7.35 26
7.60 52
7.94 74
7.96 81
7.98 66
8.02 87
8.38 160
8.44 170
8.48 170
8.78 280


a Concentration of pyridine
PEI-D-Pyr-HC1.


b % Reaction
[PNPA] Followed


[PEI-D-Pyr-HCl]a

9.0
14
9.2
4.7
9.2
9.2
14
9.2
9.2
4.7
14
9.2
4.7
9.1


based on 17% pyridine in


b Ester concentration (x 105).


TABLE IV-12


ESTEROLYSIS RATES FOR ETHYLENEDIAMINE HC1



H k2 M-in-1 [H2NCH2CH2NH2HC] [PNPA] [PNPd Followed
pH% Reaction


6.65
6.66
6.66
6.66


5.99
5.75
5.55
5.15


7.86
7.86
7.86
7.86


0
3.56
7.13
11.8


-1 -1 45
a Literature value 5.6 M- minm (no added PNP).45
b Concentrat on of mono hydrochloride (x 103) based on
pK =7.14.i
a2 5
c Ester concentration (x 10 )
d Initial p-nitrophenol concentration (x 104).














APPENDIX


Definition of pK, in Polymer Systems

The multiplicity of pKa definitions for polymer systems

in the literature creates some ambiguity as to the meaning

of polymer pKa. Kunitake and Shinkai46 have defined three

different polymer pK values.


pK = pH + n'Logi
a a


pK = p + Log-a
a,app a


pK. = pK 0.43Gel
a,int a,app RT


The following passages quoted from Morawetz47 helps

clarify the issue:


With a polymer carrying a large number of
ionizable groups, it is obviously impracticable to
specify the successive ionization constants.
Instead of this, we define the apparent ionization
constant Kapp of an average ionizable group carried
by the polylon in the usual manner by

(H+)a /(l-a,) = K
app
where K will, of course, vary with the degree of
ionizatibh since the charged polymer will interact
with the hydrogen ions. With polymeric acids the
polyanion will attract the hydrogen ions and
. /a1<0; with polymeric bases, on the other
hanH, the hydrogen ions will be repelled by the
polycation and the acid strength of the polymer will
increase with its charge density. If the required
electrostatic free energy for the removal of an

88





89

equivalent.of protons at a given degree of ioniza-
tion is AGe(a) then

papp = pK0-0.43LAi ()/R

where K is characteristic of the ionizing group
under conditions where electrostatic interactions
with other ionizing groups are absent.



The investigation of base strength-reactivity effects

requires an accurate measure of basicity. The base strength

of a poly-functional base varies with a. Therefore the

value of pKa used in the correlation of rate with reactivity

must take into account the effect of a on basicity. The

intrinsic pK (pKa, pKa,int, or pK) is devoid of a depen-

dence. The value of pK on the other hand, is a
a., app
function of a and is determined readily from pH and a data.

The use of pK obviates the necessity of determining
a,app
the value of AGel as would be required if intrinsic pKa's

were used (see above). Therefore pK p is the parameter
of choice for the base strength-reactivity correlation.
of choice for the base strength-reactivity correlation.




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EFYFG4ZKJ_CK8DOL INGEST_TIME 2017-07-17T20:21:29Z PACKAGE UF00099114_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

BASE STRENGTH-REACTIVITY EFFECTS IN POLYETHYLENIMINE ESTEROLYSIS REACTIONS BY CURTIS S. LEGE A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1979

PAGE 2

TO MY WIFE AND DAUGHTER JOE ANN AND SPRING

PAGE 3

ACKNOWLEDGMENTS The author wishes to express his appreciation to his research director, James A. Deyrup, for his thoughtful guidance and criticism as well as his encouragement of individual creativity in problem selection and pursuit. The author is further indebted to Professor John A. Zoltewicz for the many helpful discussions and use of equipment. The author wishes to thank the faculty and staff of the University of Florida and his fellow graduate students for an enjoyable and fruitful experience. Special thanks are extended to Tom Baugh and Eric Langenmeyer for their friendship and advice. The author is especially indebted to his wife who has provided not only encouragement throughout but also has made considerable contribution to the preparation of this dissertation.

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGMENTS _ _ m LIST OF TABLES vii LIST OF FIGURES viii ABSTRACT x CHAPTER I POLYMER ESTEROLYSIS REACTIONS: DISCUSSION AND PROPOSAL 1 Introduction \ Discussion 2 Mechanism of Polymer Esterolysis Reactions 2 Polymer Catalyzed Hydrolysis 8 Summary n Proposal 12 Introduction 12 Nucleophiles 12 Polymer 13 Ester 15 Results 16 CHAPTER II AMINOLYSIS OF P-NITROPHENYL ESTERS BY DODECYL POLYETHYLENIMINE 2 2 Results and Discussion 22 Characterization 22 Basicity of PEI-D-NH 2 -HC1 25 Kinetic Analysis....; 26 Eronsted Relationship 34 Summary 43 IV

PAGE 5

CHAPTER III HYDROLYSIS OF P-NITROPHENYL ESTERS CATALYZED BY POLYMER BOUND HETEROCYCLES . . 44 Introduction 44 Polyethylenimine Systems 44 Preparation 44 Primary Amine Analysis 45 Further Kinetic Analysis 49 Summary 51 Non Polyethylenimine Homopolymer Systems. 52 Conclusion 58 CHAPTER IV EXPERIMENTAL 60 Introduction 60 Syntheses 61 Dodecyl Polyethylenimine PEI-D 61 Dodecyl Polyethylenimine Hydrochloride PEI-D-NH2-HC1 61 Dodecyl4 -me thy lenepyridine-polyethylenimine Hydrochloride PEI-D-Pyr-HCl 62 Dodecyl-N(2-pyridyl) -3-propylaminepolyethylenimine Hydrochloride PEI-D-APyr-HCl. . . . 62 Dodecyl-4 (5) -methylenimidazole-polyethylenimine Hydrochloride PEI-D-Im-HCl 63 Dodecyl-4 (5) -methylenimidazoleisopropyl-polyethylenimine Hydrochloride PEI-D-Im-Ip-HCl . . . . 64 Dodecyl-4 (5) -methylenimidazole-isopropyl-isopropyl-polyethylinimine ^Hydrochloride PEI-D-Im-Ip2-HCl. . . 65 4(d) -Chloromethylimidazole Hydrochloride 65 4(5) -Hydroxymethylimidazole Hydrochloride 65 3-Bromopropanal Dimethyl Acetal 66 N(2-pyridyl) -3-aminopropanal Dimethyl Acetal 6 6 N(2-pyridyl) -3-aminopropanal Hydrochloride 66 p-Nitrophenyl Acetate PNPA 67 p-Nitrophenol Caproate PNPC 67 Composition of PEI Derivatives by Elemental Analysis 6 7 Composition of PEI Derivatives by NMR Spectral Analysis 7 3 Determination of Primary Content of PEI Derivatives Using Trinitrobenzene Sulfonate (TNBS) 73

PAGE 6

APPENDIX. Primary Amine Detection by Ninhydrin 7 4 Relative Concentrations of Imidazole Containing Polymer Solutions 75 Potentiometric Titration of PEI-DNH 2 -HC1 76 Preparation and Standardization of C0 2 Free 0.1 N KOH Titrant 7 8 Preparation of Ester Solutions Used for Kinetic Studies 78 Preparation of Polymer Solutions for Kinetic Studies 78 Buffer Solutions 79 Kinetic Method 79 Background Rate 8 Pyridine Ionization in PEI-D-Pyr-HCl 81 33 REFERENCES % 9 BIOGRAPHICAL SKETCH 94 VI

PAGE 7

LIST OF TABLES Table Page II-l PRIMARY AMINE AND ISOPROPYL CONTENT OF PEI-D-NH 2 -HC1 AND PEI-D-IpHCl 2 3 III-l % PEI UNITS BOUND TO HETEROCYCLES 45 IV-1 COMPOSITION OF PEI DERIVATIVES FROM ELEMENTAL ANALYSIS DATA , 70 IV2 FRACTION OF PEI UNITS ALKYLATED FROM NMR SPECTRAL DATA 72 IV3 PRIMARY AMINE CONTENT OF PEI DERIVATIVES USING TNBS 74 IV-4 POTENTIOMETRIC TITRATION DATA FOR PEI-D-NH HC1 2 77 IV5 ESTEROLYSIS RATES FOR PEI-D-NH -HC1 83 IV-6 ESTEROLYSIS RATES FOR PEI-D-Ip-HCl 84 IV7 ESTEROLYSIS RATES FOR PEI-D-Im-HCl 84 IV8 ESTEROLYSIS RATES FOR PEI-D-Im-Ip-HCl 85 IV9 ESTEROLYSIS RATES FOR PEI-D-Im-Ip 2 -HCl 85 IV-10 ESTEROLYSIS RATES FOR PEI-D-APyr-HCl 86 IV11 ESTEROLYSIS RATES FOR PEI-D-Pyr-HCl 87 IV12 ESTEROLYSIS RATES FOR ETHYLENEDIAMINE • HC1 . . 87 Vll

PAGE 8

LIST OF FIGURES Figure p age I-la Solvolysis of PNPA catalyzed by Poly4 (5) -vinylimadazole and imidazole 10 I-lb Bifunctional mechanisms proposed for poly-4 (5) -vinylimidazole catalyzed hydrolysis of PNPA 10 1-2 Esterolysis of PNPA by PEI-D-Pyr-KCl 18 1-3 Esterolysis of PNPA by PEI-D-APyr-HCl . . . . 19 1-4 Esterolysis of PNPA by PEI-D-Im-HCl 20 1-5 Esterolysis of PNPA by PEI-D-NH--HC1 compared to esterolysis of PNPA by PEI bound heterocyclic systems 21 II-l Comparison of pK a aDp dependence on a for PEI-D-NH -HC1 with low molecular weight polyamines 27 II-2 Comparison of esterolysis rates for PEI-D-NH 2 -KC1 and PEI-D-Ip-HCl 29 II-3 Inhibition of ethylenediamine, PEI-D-NH2HC1 (pH=7.00) and PEI-D-NH^-HCl (pH=8.82) esterolysis of PNPA by p-nitrophenol anion (PNP) 31 II-4 Plot of log(k /a) against pH for PEI-DNH 2 -HC1 esterolysis of PNPC and PNPA 33 II-5 Bronsted plot for the esterolysis of PNPA by simple primary and cyclic secondary amines 36 II-6 Bronsted plot for the esterolysis of PNPA by imidazoles and anilines 37

PAGE 9

II-7 Plot of pK vs pH for PEI-D-NH„-HC1 38 a , app 2. II-8 Bronsted plot for the esterolysis of PNPA and PNPC by PEI-D-NH -HCl 4 II-9 plot of e vs pK or [a] for the esterolysis of PNPC and PNPA by PEI-D-NH -HCl 41 III-l Effects of repetitive isopropylation of PEI-DIm-HCl on PNPA esterolysis reactivity 48 III-2 Comparison of PNPA esterolysis rates for PEI-D-HC1 and PEI-D-Ip-HCl 50 III-3 Plot of k obs against a for the esterolysis of 2, 4-dinitrophenyl acetate by poly-4-vinylpyridine and pyridine 54 III-4 Plot of k„ against a for the esterolysis of PNPA by PEI-D-NH 2 -HC1 55 III-5 Bronsted plot for the esterolysis of PNPA by poly-4 (5) -vinylimidazole and imidazoles 56 III-6 Bronsted plot for the esterolysis of 2,4-dinitrophenylacetate by poly(4) -vinylpyridine. . . . 5 7 IV-1 Typical spectrum of a PEI-D derivative (PEI-DAPyr-HCl) 71

PAGE 10

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy BASE STRENGTH-REACTIVITY EFFECTS IN POLYETHYLENIMINE ESTEROLYSIS REACTIONS BY CURTIS S. LEGE December 1979 Chairman: James A. Deyrup Major Department: Chemistry The esterolysis of p-nitrophenyl acetate (PNP) and p-nitrophenyl caproate (PNPC) by polyethylenimine (PEI) derivatives was studied as a function of pH . The reactivity of partly dodecylated PEI containing primary amines (PEI-D-NH 2 -HC1) , imidazole (PEI-D-Im-HCl) , pyridine (PEI-D-Pyr-HCl) , and 2-aminopyridine (PEI-D-APyr-HCl) was found to be pH dependent without exception. The PEI-D-NH 2 ~HC1 system was found to react with PNPA predominantly by nucleophilic attack of polymeric primary amine. Potentiometric titration showed the pK or PEI-D-NH„-HC1 a , app 2 also to be dependent. A Bronsted-like plot of log k for polymeric free primary amine aaainst pK for the estera , app olysis of PNPA with PEI-D-NH -HC1 had a slope (3 value) of 0.81. A Bronsted plot of a series of low molecular weight amines with PNPA taken from the literature had a slope of 0.83. The Bronsted-like plot of log k with pK , an a , a p p

PAGE 11

approach unknown prviously in polymer systems, quantitatively accounted for the pH dependence of the reaction of PEI-D-NH 2 -HC1 with PNPA. The reactivity of PEI-D-NH -HC1 toward PNPA is dependent on the pK of the nucleophile as in the case of low molecular weight amine nucleophiles . The slope of the plot of log k versus pK a ^ for the esterolysis of PNPC by PEI-D-NH 2 -HC1 had a slope of 1.06. This slope was dissected into a pH dependent polymeric enhancement component (E) and a pH dependent nucleophilic component. The pK a , app of the heterocycles in the heterocycle containing polymers was unobtainable in the pH region of interest. However, the pHdependencies were assigned to nucleophilic effects by analogy. This approach proved useful in reviewing literature reports of other polymeric esterolytic systems. Data for PNPA esterolysis by poly-4 ( 5) vinylimidazole and for 2,4-dinitrophenyl acetate esterolysis were plotted as described above with slopes of 0.8 and 1.3, respectively.

PAGE 12

CHAPTER I POLYMER ESTEROLYSIS REACTIONS: DISCUSSION AND PROPOSAL Introduction Esterolysis by polymer systems has received considerable attention since the early days of polymer research. This area of research was spurred as the mechanism of enzyme reactions became better understood. It was hoped that synthetic polymer systems might offer a method of modeling or duplicating the action of the naturally occurring polymers (enzymes) . The aspects of enzyme action of greatest interest in this field are high selectivity and high efficiency. To this end synthetic polymer systems with particular binding properties and high reactivity have been sought. One of the aspects of polymer esterolysis reactions which has received a great deal of attention is cooperativity, i.e., two or more functional groups in the same polymer molecule participating in ester cleavage. The original goal of the research to be discussed herein was to find and study possible cooperative effects in nitrogenous polymer systems. Chapter I will first discuss the mechanism of polymer esterolysis and aspects of possible bifunctional or cooperative mechanisms. Secondly, the design of a system

PAGE 13

2 in which cooperative effects may be found will be discussed. The remainder of this dissertation is a discussion and interpretation of the results of the study and the impact of this interpretation on the field of polymer esterolysis. Discussion Mechanism of Polymer Esterolysis Reactions One of the features present in polymer reactions that is generally absent in monomer reactions is the possibility of binding preassociation. A simplified scheme for such a reaction is shown in Equation 1-1. Equation 1-1 P + S PS Reaction > Products t 6 1 The substrate may (path a) or may not (path b) bind, depending on the structure of both the polymer, P, and the substrate, S. Binding can greatly increase reaction rates by increasing the concentration of substrate in the presence of the polymer. There are at least two modes of binding, hydrophobic and electrostatic, for synthetic polymer systems. Hydrophobic binding is the mole of interest here; it requires polymers and substrates of hydrophobic nature. Such substrates frequently contain long aliphatic chains which are usually located on the carboxyl end of the ester. The electrostatic effects cf concern are those produced by a partly ionized polymer and their consequence on hydrophobic binding and nucleophilicity . These effects are quite

PAGE 14

3 important, but until now they have been impossible to quantitate. In esterolytic reactions of polymers, the binding can certainly increase the reaction rate, but the cleavage portion of the esterolysis mechanism is considered to be independent of binding. That is to say that the ester is cleaved by the polymer in the same way, whether it is bound to the polymer in a previous step or not (neglecting buffer reactions, etc.). This is a reasonable supposition in that nonspecific hydrophobic binding should not result in any significant change in the details of the reaction mechanism. This assumption, which is usually not stated, is very important in understanding polymer esterolysis reactions as it allows comparison to nonpolymeric reactions. The specific types of esterolysis reactions of interest to this study are those of nucleophilic amines with p-nitrophenyl esters in aqueous media. Satterthwait and Jencks have provided the mechanism shown in Equation 1-2 for esterolysis reactions which are first order both in p-nitrophenyl ester and nonpolymeric amine nucleophile. The tetrahedral intermediate T~ is formed in a rapid and reversible step. The rate determining step is the breakdown Equation 1-2 H 0~ R — NH,+Me — C — OR' , R — N — C — OR' K ~ i Products k-1 | | H Me

PAGE 15

4 of T . That is, T is partitioned between starting materials (k-1) and products. The partitioning is governed by the relative strengths of the C-N and C-OR bonds. The bond strengths can be related to the pK ' s of the two possible leaving groups. The greater the pK of the nucleophile, the greater the percentage of tetrahedral intermediates proceeding on to products. It is clear then that the pK of a the nucleophilic amine is a very important measure of its reactivity in aminolysis reactions. The correlation of reaction rate with pK , known as the Bronsted relation, is very important to the arguments put forth later and will be discussed in more detail in Chapter II. The breakdown of the tetrahedral intermediate may be catalyzed by general acid or general base mechanisms. The mechanism of general base catalyzed breakdown of the tetrahedral intermediate T~ is shown in Equation 1-3. This pathway transforms the protonated amine group into a nonprotonated amine. The amine pK of importance for the anionic Equation 1-3 H 0" 0" .+ kb[B] v R — N — C — OR' ^ — R — N — C — OR Products , , k-b[BH ] H Me h Me T" intermediate, T , is the conversion of the amine to the amide anion. This pK is very much greater than that of the alcohol. The intermediate T~ is thus prevented from breaking down to starting materials. This general base catalyzed

PAGE 16

5 pathway promotes favorable partitioning of T~ toward products. The third order pathway contributes more in cases where the amine pK is significantly lower than that of the alcohol moiety. The general acid pathway generates the T~ intermediate via a more intricate set of proton transfers. However this pathway includes proton transfers, which for reactions of moderately basic amines, are thermodynamically unfavorable for p-nitrophenyl ester reactions (proton donation from a weak acid to a weaker base) . The general acid pathway is of higher energy than the general base and the 4 uncatalyzed pathways. Therefore, general acid catalyzed reactions will not be considered further. The above discussion was largely directed at primary and secondary amines; however, it also applies to nucleo3 phi lie tertiary amines. The heterocyclic tertiary amines, pyridine and imidazole, are of particular interest. Most pyridines are incapable of losing a proton in the T x intermediate ,and the T intermediate is therefore not accessible (Equation 1-3) . Nucleophilic reactions of pyridines with p-nitrophenyl esters must pass through the second order H \\4 N Pyridine Imidazole uncatalyzed pathway. Imidazole on the other hand is not restrained from access to the T~ intermediate. The imidazole may lose a proton from the T~ intermediate to form T~. This pathway is analogous to the pathway proposed by Satterthwait

PAGE 17

3 and Jencks for the general base catalyzed aminolysis of esters by primary and secondary amines. Satterthv/ait and Equation 1-4 oP" ^ I Kb[B], ^ I H -N'+N — C — OR ^ +_, N N C — OR ^Products \— / | K-b[BH ] \ == J , CH 3 CH 3 Jencks did not discuss the scheme in Equation 1-4. However, this mechanism has been proposed for nonaqueous systems. 5 This discussion is not meant to imply that there is a general base catalyzed imidazole acylation reaction with p-nitrophenyl esters in aqueous systems, but for the purposes of this discussion any general base catalysis will be assumed to follow this mechanism. The general base catalyzed reaction discussed above favors partitioning of the intermediate T~ to products by formation of T~ (Equation 1-3 or 1-4) . If the pK of the a amine nucleophile is not greatly larger than the alcohol leaving group then, the second order uncatalyzed pathway (Equation 1-2) would not partition as efficiently as the third order general base catalyzed route. In terms of activation parameters, the third order pathway for phenyl esters in general is lower in activation enthalpy than the second order pathway. However, third order reactions are of lower activation entropy than second order reactions. The decrease in activation entropy tends to overcome the decrease

PAGE 18

7 in activation enthalpy. As a result the second order pathway is frequently of lower energy than the third order 4 pathway. It is clear from the above discussion that if the problem of activation entropy reduction could be resolved, esterolysis reaction rates could be enhanced via the general base catalyzed route (Equation 1-3 or 1-4) in some cases. If the general base and nucleophile were part of the same molecule, then the general base pathway would no longer be third order. The entropy of activation reduction would be resolved largely. Addition of more general base species to the nucleophile containing molecule should reduce the free energy of activation for the general base reaction further. A polymer containing nucleophiles and species capable of serving the general base function should lower the free energy of activation for the general base catalyzed reaction below that of the uncatalyzed reaction. Such a bifunctional or cooperative interaction should enhance the reaction rate. The extent of the enhancement would depend on the nucleophile-ester pair studied. As long as the formation of the intermediate T" (Equation 1-2) is rapid and reversible, the greater the pK^ of the alcohol leaving group relative to the pK a of the nucleo P h il e ' the more effective is the enhancement. Conversely, if the pK of the alcohol leaving group is low relative to the pK of the nucleophile, then the partitioning of the intermediate T~ would be very efficient. In the latter case the general base catalyzed reaction would

PAGE 19

not be significantly lower, and no enhancement should be observed. Polymer Catalyzed Hydrolysis The initially formed products of the ester aminolysis reactions are amides. However, in some cases these reactions can be viewed as the first step of ester hydrolysis reactions. An amine may serve as a nucleophilic catalyst if the initially formed amide is hydrolyzed at a reasonable rate (Equation 1-5) . In actual practice nucleophilic tertiary amines, pyridines or imidazoles are usually used for these reactions. These terriary amines have a reasonable acyiation rate as well as deacylation rate. However, the first step of the reaction, the aminolysis step, is the point of concern here. The actual fate of the ester after release of alcohol moiety is of little consequence. Thus the structure of the amine nucleophile relative to its deacylation is also of little consequence. Therefore, the aminolysis reaction by primary amines (Chapter II) will be used as a model for Equation 1-5

PAGE 20

Overberger et al. have published data that they interpret as being consistent with a bifunctional mechanism. They studied the hydrolysis of p-nitrophenyl acetate catalyzed by poly-4 (5) -vinylimidazole and imidazole as a function of the fraction of unprotonated imidazole units 6 (Figure I-la) . They found that although the imidazole esterolysis rate increases linearly with a, the polymer rate increases nonlinearly. They concluded that neighboring nonprotonated imidazoles interact cooperatively, enhancing the rate of reaction. At low values of a few nonprotonated imidazoles are neighbors, and the reaction presumably proceeds through a noncooperative pathway. As the value of a increases the cooperative interaction becomes more and more the predominant pathway, and the reaction rate increases accordingly. The mechanism of Jencks and Satterthwait (Equation 1-2, 1-4) although different from those proposed by Overberger et al. (Figure I-lb) is also consistent with their interpretation. In the terms discussed previously, the reaction would proceed via the uncatalyzed (Equation 1-2) path as a increases. It should be pointed out that the bifunctional interpretation of Overberger et al. has not been without criticism. Kunitake and okahata notes that later work on poly-4 (5)vinylimidazole shows considerable conformation change, with changes in solvent polarity. When the polymer is in its more compact conformation, it provides a more efficient catalysis of long chain phenyl esters. He suggests that (p 178)

PAGE 21

10 cat 5030100.60 0.80 1.00 Figure I-la. Solvolysis of PNPA catalyzed by poly-4 (5) vinylimidazole (A) and imidazole (O). 6 28.5 v/v% EtOH-H 2 0, u=0.02, 30°C. i! R-C —OR 1 \ «-/7 r;» / R-C -OR' o R_ r _0R' L> 'igure I-lb. Bifunctional mechanisms proposed for poly-4 (5) vinylimidazole catalyzed hydrolysis of PNPA. 6

PAGE 22

11 "the increased catalytic efficiency observed at higher pH ' s may similarly be explained on a microenvironmental effect. " lb Summary In summary, there are two p-nitrophenyl ester aminolysis reactions of concern. These are the uncatalyzed reaction (Equation 1-2) and the general base catalyzed reaction (Equation 1-3, 1-4). The general base catalyzed reaction would be faster than the uncatalyzed reaction (for nucleophile pK ' s not much greater than alcohol pK 's) if it were a a not third order. The general base reaction can be made second order by including the general base in the same molecule as the nucleophile. This effect should be increased by surrounding the nucleophile with general base functions as might be found in a polymer system. The reaction of the appropriate nucleophile ester pair should then be enhanced in the appropriate polymer due to contribution of the general base catalyzed reaction. Therefore it is reasonable to look for the existence of bifunctional or cooperative effects in polymeric esterolysis reactions based on the above discussion, if suitable reactants are studied. Failure to find a bifunctional effect could be due to choice of the wrong reactants. It is only possible then to attempt to choose a system in which the polymeric bifunctional reaction may be unambiguosly observed.

PAGE 23

12 Proposal Introduction There are several variables involved in the selection of the proper system in which to look for bifunctional catalysis. First one must choose the nucleophiles . Nucleophiles both capable and incapable of bifunctional catalysis are needed in order to distinguish the two pathways. Second, the polymer must contain general bases capable of removing + a proton from the T intermediate. Third, the appropriate Equation 1-6 k,[B]

PAGE 24

13 for this nucleophile in poly-4 (5) -vinylimidazole. It was decided, then, to compare imidazole to other similar heterocycles to determine if it had any special properties in esterolysis reactions. The similar structure 2-aminopyridine was also chosen. The T intermediate formed from 2-aminopyridine is structurally capable of losing a proton (Equation 1-6) just as the T r intermediate formed from imidazole (Equation 1-4) . The pyridine nucleophile was chosen as a nonbifunctional comparison. Unlike 2-aminopyridine and imidazole, pyridine is structurally incapable of losing a + _ proton from T to form T . Pyridine then cannot react via the general base pathway but only through the uncatalyzed pathway. Polymer Instead of studying a homopolymer as did Overberger et 6 ai . , a polymer system was chosen in which the basic sites remained constant but the nucleophiles could be varied. What was needed then was a polymer containing basic units to which the nucleophilic sites could be attached. The different nucleophiles then could be studied in the presence of a general base of constant strength. Polyethylenimine (PEI as furnished by Dow) seemed to be the ideal system. Dow PEI 600 (Mw=40 , 000-60 , 000) is derived from cationic ring opening polymerization of aziridine o (ethylenimine) as shown in Equation 1-7. The structure of the polymer can best be understood via a discussion of its formation. If the reaction proceeded as in Equation 1-7

PAGE 25

14 the polymer would consist of only secondary amines (ignoring end groups) . However the polymer has been shown to contain primary, secondary and tertiary amines. This observation is readily incorporated into the mechanism by allowing the aziridinium ion to react with both primary and secondary amines. This mechanism would provide a branched polymer. Every branch produces a tertiary and a primary amino group. Thus the ratio of primary to tertiary amines is 1:1. The end groups comprising less than 1% of the total amine functions are ignored in this calculation. ' The tertiary amine content has been determined by titration after exhaustive benzoylation to be 25%. Simple arithmetic reveals the primary amine content to be 25% and the secondary amine content to be 50%. Thus the ratio of primary : secondary : tertiary amines is 1:2:1. Klotz has published extensively on the properties of PEI and its derivatives. It has been shown that the primary Equation 1-7 NH + N + N H. + N H. amines of PEI are very reactive toward p-nitrophenyl esters. This is especially true for partially dodecylated and dodecanoylated derivatives reacting with apolar substrates. 11 A great deal of this enhanced reactivity is apparently due to binding. Of particular interest is a PEI 600 derivative of

PAGE 26

15 which 10% of its nitrogens have been dodecylated, and 15% of its nitrogens have been alkylated with 4 (5) -chloromethylimi12 dazole. " This "synsyme" is reported to approach a chymotrypsim in catalytic activity. The acylation rate (excess nucleo-phile) is 2700 M~ 1 min -1 for the synzyme with p-nitrophenyl caproate compared to 10,000 M _1 min -1 for a chymotrypsin with p-nitrophenyl acetate. The synzyme system of Klotz was chosen as the polymer to study. Replacement of the imidazole group with pyridine or 2-amincpyridine is readily carried out owing to the synthetic versatility of PEL Klotz et al . have shown that the polymer may be derivatizec 12 13 and reductive amination. ' 1 2 12 the polymer may be derivatized by alkylation, ^ acylation, Ester P-nitrophenyl acetate and caproate were chosen as esters to study. The p-nitrophenyl esters are convenient for kinetic analysis. The reactions are readily followed 14 spectropnotcmetricaily . Furthermore the great majority of polymer esterolysis studies have been carried out on p-nitrophenyl esters. The caproate ester binds very strongly to the dodecylated PEI systems and therefore proceeds through path a of Eguation 1-1. On the other hand, the acetate does not bind nearly so well and must proceed largely through path b. The use of these two esters provides an insight into the binding effects of the polymer as well as being comparable to the other studies in the area.

PAGE 27

16 Results It was reasoned that possible bifunctional effects would become apparent in pH rate profiles of the polymer esterclysis reactions. With increasing pH a larger and larger fraction of the backbone nitrogens should be able to act as general bases to convert the intermediate T" into T~ enhancing the reaction rate. If such were the case, the reaction rate should be pH dependent. If such were not the case there should be no pH dependence unless polymer charge density has an effect on rate. Overberger and Salamone, le though noting some controversy, states ". . .it appears that the varying charge density does not significantly alter the reactivity of a catalytically active polyion toward a neutral substrate" (p 218) . The pH rate profiles for the aminoiysis reactions by the various polymer reactions with p-nitrophenyl acetate are shown in Figures 1-2, 1-3, 1-4 and 1-5. It is clear that the PEI-D-Pyr-HCl (pyridine system) is pH dependent. As will be pointed out in Chapter III, the pyridine in PEI-D-Pyr-HCl remains in the nonprotonated nucleophilic form throughout the pH range. Furthermore the T" intermediate formed from pyridine does not have access to T~. Therefore the pK dependence for the PEI-D-Pyr-KCl system cannot be due either to ionization of the pyridine or a bifunctional mechanism. In view of Overberger ' s statement in this regard and its possible influence on polymer reactions, the nature of this pH dependency is very

PAGE 28

17 important. The remainder of this dissertation is an effort to understand and quantitate this pH dependency.

PAGE 29

18 3.0-1 2.01.0' P H Figure 1-2. Esterolysis of PNPA by PEI-D-Pyr-HCl (slope 0.74). See Chapter IV for experimental details .

PAGE 30

19 3.0* 2.CL cn o H 1.0P H Figure 1-3. Esterolysis of PNPA by PEI-D-APyr-HCl (slope = 0.69). See Chapter IV for experimental details.

PAGE 31

20 3.0 2.0 1.0 PH Figure 1-4. Esterolysis of PNPA by PEI-D-Im-HCl (slope = 0.56). See Chapter IV for experimental details

PAGE 32

21 3.0—1 2.0 1.0 / ' PH Figure 1-5. Esterolysis of PNPA by PEI-D-NH 2 -HC1 compared to esterolysis of PEI bound heterocyclic systems. See Chapter IV for experimental details. (a) PEI-D-Im-HCl (b) PEI-D-Pyr-HCl (c) PEI-D-APyr-HCl

PAGE 33

CHAPTER II AMINCLYSIS OF P-NITROPHENYL ESTERS BY DODECYL POLYETHYLENIMINE Results and Discussion Characterization A large quantity of PEI-600 was alkylated with dodecyl 10 iodide. The PEI-D from this preparation was used to prepare all the systems studied. This precaution served to maintain a constant level of dodecylation from system to system. The simplest system studied, PEI-D-NH--HC1 , was the HC1 salt of PEI-D. The high precision of the duplicate elemental analysis shows that PEI-D-NH -HC1 is homogenous (this was true for all systems Table IV-1) . The C/N ratic from elemental analysis can be used to calculate the extent of dodecylation as described in detail for Table IV-1. The fraction of backbone nitrogens dodecylated in PEI-D-NH -HC1 as determined from the C/N ratio, 11%, is in good agreement with the value from NMR spectral analysis, 10% (Table IV-2) . The other polymer system to be discussed in this chapter is PEI-D-Ip-HCl . This system was prepared by exhaustive isopropylation of PEI-D with acetone and NaBH, 22

PAGE 34

23 (Equation II-l) . NMR spectral analysis of the extent of Equation II-l P P H / ^ / -H " m ,N h 2+ J|^ ^ J! £S^ ^ ~I 3 CH 3 +H 2 CH 3 isopropylation was not possible due to interference of the dodecyl group. However the extent of isopropylation can be determined from elemental analysis. The C/N ratio for PEI-D-Ip-HCl gives a value of 0.32 for the fraction of nitrogens isopropylated. Primary amine determinations were made using trinitrobenzene sulfonate in a modification of the procedure developed by Satake et al. Reportedly this method is specific for primary amines in the presence of secondary amines. Primary amine analysis of PEI-D-NH_-HC1 provides 0.20 as the fraction of primary amines, whereas analysis of PEI-D-Ip-HCl gives a value of 0.01 (Table II-l). TABLE II-l PRIMARY AMINE AND ISOPROPYL CONTENT OF PEI-D-NH -HC1 AND PEI-D-Ip-HCl Polymer % Primary Amines % Isopropylation PEI-D-NH 2 -HC1 20 .... PEI-D-NH 2 -Ip-HCl 1 32 a Values rounded to the nearest whole % ; see Chapter IV for experimental details.

PAGE 35

24. The above result may be surprising as it requires isopropylation of secondary amines with NaBH,. Onlv 20% of 4 the backbone nitrogens of PEI-D, the starting material, were primary amines. However after isopropylation 32% of the total backbone amines were isopropylated. Therefore a significant number of secondary amines must have been isopropylated. It has been well established that unhindered ketones can be aminated with secondary amines using the less reactive NaCNBH., . However examples of this reaction with NaBH, are 17 few. The reason for this selectivity seems to involve the competition for hydride between the ketone (I) and the much more reactive but less abundant iminium ion (II) Equation II-2 / CH 3 K R 2 NH + 0=^ ^ I CH 3 R \ + / CH 3 N" + HO / A CH V NaBH CH. HO — (H CH. NaCNBH., or NaBH 4 R CH. R CH. (Equation II-2) . The NaCNBH reagent does not reduce the ketone to a significant extent, but the more reactive NaBH , 16 _ . . does. Iminium ion reduction must occur to a greater

PAGE 36

25 extent under the conditions used for preparation of PEI-DIp-HCl than usual. This can be explained in part by the use of a huge excess of both NaBH 4 and acetone. In addition the polymer may very well promote formation of the iminium ion as well. This possibility is suggested by the study of acetone H-D exchange, catalyzed by PEI, carried out by 1 8 Hine. They have shown that PEI (as well as certain diamines) catalyze H-D exchange, which proceeds via formation of an iminium ion, much more efficiently than monoamines. The observation of secondary amine isopropylation is important in view of the past usage of acetone/NaBH as a method of determining primary amine content of PEI 10 Basicity of PEI-D-NH 2 -HC1 The acid-base properties of PEI-D-NH 2 -KC1 were determined by potentiometric titration (see Chapter IV for experimental details) . The values of a, fraction of nonprotonated polymeric amine, after each addition of titrant, were given by the concentration of added titrant divided by the concentration of polymeric amine units, CH„CH„N. The values of P K a ,apo at each val ue of a was calculated from the Henderson-Hasselbalch equation (Equation II-3, see Appendix) . Equation II-3 pK = pH + log i-2-

PAGE 37

26 From the plot of pK vs a (Figure II-l) it can be d , app seen that the basicity, pK , is dependent on the fraction of polymeric amines protonated, a. A number of factors which might well be dependent on a probably have an influence on the pK such as conformation, ion pairing, etc. However, a , app the important fact here is that the pK is dependent on a a, app and therefore pH. The dependence of the polymeric pK on a, app a is no surprise if short chain polyamines are used as models. The pK values for ethylenediamine, diethylenetri20 amine and triethylenetetramine can be determined discretely for each stage of ionization. The values of these dK 's are a plotted against a in Figure II-l. The pK 's of these systems follow qualitatively the same a dependence trend as the polymer. Since the polymer is not crosslinked and the P K = =«r, dependence on a is similar to the pK dependence on a a, app a for low molecular weight model compounds, conformations of the polymer exposing all basic sites are assumed to be in rapid equilibrium. Kinetic Analysis Rates of disappearance of substrate, p-nitrophenyl acetate (PNPA) or p-nitrophenyl caproate (PNPC) , were monitored spectrophotometrically by observing the increase in p-nitrophenoxide absorbance. These reactions were studied under pseudo first order conditions (excess polymer) at pH's within the range 6.5-9.0. The second order rate constants (k 2 M min ) were calculated, after subracting background rate (see Chapter IV) , by division by concentration of primary amine unless otherwise indicated.

PAGE 38

27 ion 642 0.2 0.4 0.6 0.8 -I 1.0 Figure II-l. Comparison of pK a a dependence on a for PEI-D-NH 2 -HC1 witn low molecular weight poly-amines. 20 (0) PEI-D-NH^-HCl, (A) ethylenediamine, (o) diethylenetriamine, (0) triethylenetetr amine .

PAGE 39

28 Two different polymeric esterolysis processes may contribute to the pH dependence of p-nitrophenoxide release, aminolysis and/or general base catalyzed hydrolysis. The general base catalyzed hydrolysis and nucleophilic or aminolysis reaction produce different products. The general base reaction generates the carboxylate ion (Equation II-4) , 21 whereas the aminolysis reaction produces the amide (Equation II-5) . Both reactions release the species followed kinetically, p-nitrophenoxide. The general base route was ruled out as a significant contribution to the rate. First, the contribution of the general base component (cf Equation II-4) can be shown to be minimal. Isopropylation of PEI-D (forming PEI-D-Ip-HCl) converts virtually all the primary amines to secondary amines. Rate constants from this system should estimate an upper limit for the general base component. In order to compare the rate data for these two systems the second order rate constants (k ) 2Gb must be calculated based on the concentration of basic sites, CK 2 CH 2 N units (Figure II-2) . In comparing the rates Equation II-4 oI // c PNP > R C + PNP I V OH

PAGE 40

29 75— 50— 6.5 Figure II-2. Comparison of esterolysis rates for (o)PEI-DNH 9 -KC1 and (D) PEI-D-Ip-HCl.

PAGE 41

30 for the PNPA reaction it can be seen that the maximum contribution of the general base reaction is less than 3%. Secondly, the reaction products can be shewn to be not those of the general base reaction. The p-nitrophenoxide anion inhibits the reaction of PNPA with PEI-D-NH -HC1 (Figure II-3) . This inhibition is not due to a normal salt effect (ionic strength) as the ionic strength is 0.1, whereas the maximum p-nitrophenoxide concentration is 1.5 x 10~ , a variation of less than 0.2%. Nor is this effect peculiar to the polyion since the same reaction with ethylenediamine hydrochloride is also depressed to a similar extent. This effect then can be compared to the common ion effect of carbonium ion chemistry. The p-nitrophenoxide ion is acting as a nucleophile on either an intermediate or product, reforming starting material. This is not possible in the case of general base hydrolysis, as neither the tetrahedral intermediate nor the product, carboxylate ion, is susceptible to nucleophilic attack (cf Equation II-4) . However the result of direct attack of the amine on the ester generates the amide which is susceptible to attack by the p-nitrophenoxide ion (Equation II-5) . Equation II-5 ?" II .1 II R C PNP + P-NH "==} R C PNP — ^ R C NH-P I P-NH + PNP~

PAGE 42

31 ioo6040— 20Q 1 I 5 10 [PNP"] b X 10 5 15 Figure II-3. Inhibition of () ethylenediamine , (A) PEI-D-NH,HC1 (pH=7.00) and (o) PEI-D-NH -HC1 (pH=8.82) esterolysis of PNPA by p-nitrophenol anion (PNP) . % Reduction refers to the % the rate constant is reduced in the presence of PNP from the rate constant at very low [PNP ] . b The PNP concentration is based on p-nitrophenyl pK =7.14. 22

PAGE 43

32 Inhibition by p-nitrophenoxide ion has been observed by Jencks and Gilchrest in 4-methylpyridine catalyzed hydrolysis of PNPA. In this example p-nitrophenoxide attacks the reactive acetylpyridimium intermediate. However, p-nitrophenoxide attack on the much less reactive amides described here was surprising. Based on the above results some statements can be made ==„ » ^•\J~ CZ \ CH 3 Acetyl-Pyridinium Intermediate about the mechanism of the aminolysis reaction. The reaction proceeds by nucleophilic attack of primary amine on the ester. The tetrahedral intermediate is assumed but not required. The p-nitrophenoxide anion is lost in a reversible step to form the amide. However, the question of pH dependence remains unresolved. In small molecule systems pH dependencies can often be explained on the basis of ionization. A protonated amine cannot act as a nucleophile, and thus the rate is dependent 2 3 on the fraction of free amines. In the PEI-D-NH„-HC1 case correction for ionization does not remove the dependency. The pH reate profiles for PEI-D-NH„-HC1 (k 2 2 based on primary amine and divided by a) for both PNPA and PNPC still exhibit pH dependency after correction for ionization (Figure II-4) .

PAGE 44

33 4.0 _ 1.0 7.0 8.0 9.0 PH Figure II-4. Plot of log (k /a) against pH for PEI-D-NH„-HC1 esterolysis of (o)PNPC and (D)PNPA.

PAGE 45

34 The question of pH dependence in polymeric systems has been explained in several ways. One explanation involves pH dependent exclusion. As the charge density of the polymer increases hydrophobic interactions decrease, excluding the 24 hydrophobic substrate. Restated, the binding increases with increasing a thereby increasing the rate. A second explanation relies on a bifunctional mechanism. Overberger et al. have explained rate dependencies on a (which in turn is dependent on pK) in poly-4 (5) -vinylimidazole reactions with PNPA as being due to interaction of adjacent imidazoles. These interactions presumably increase the reaction rate (Chapter I) . A third explanation invokes the tenet relating nucleophilicity to basicity. That is, as the pK (basicity) increases so does the nucleophilicity and thus the reactivi25 ty. This explanation was used by Letsinger and Saveride for a dependence of rate on a. The relationship between rate and pK will be explored quantitatively for the PEI-D-NH 2 -HC1 system in the next section via the Bronsted relationship. Bronsted Relationship The Bronsted relationship, relating pK to rate a constants (Equation II-6), has been used extensively in quantifying esterolysis kinetic data. This relationship Equation II-6 log k/k = $ pK + C

PAGE 46

35 can be applied to a variety of nucleophilic addition and displacement reactions. However, discussion here will be limited to ester aminolysis reactions. The values of the parameters, 5 and C, can be obtained by plotting the values of log k 2 (corrected for ionization) against pK for a series of structurally similar nucleophiles . The value of 6 is taken as a measure of the contribution of basicity to nucleophilicity. An example of such a plot can be seen in Figure II-5. In this case the 6 value for a series of primary and secondary amines reacting with PNPA is 0.83. A 3 value of this magnitude indicates considerable sensitivity to base strength in the reactivity of the nucleophile. The Bronsted 2 value is not particularly sensitive to reaction conditions. In Figure II-6 Bronsted plots for imidazoles and anilines reacting with PNPA at conditions significantly different than the amine reactions (Figure II-6) still exhibit a 3 value of approximately 0.8. The value of C, the vertical juxtaposition of the parallel lines, is sensitive, however, to nucleophile structure and to reaction conditions. The dependence of the nucleophilicity of primary and secondary amines on base strength raises the question, "Can a Bronsted relationship be used to explain the pH dependence of the PEI-D-NH 2 -HC1 system?" Conventional Bronsted plots require data from a number of different nucleophiles since monobasic systems have one pK and therefore one Bronsted point, dibasic systems require two pK 's, etc. Conversely, the PK & of the polybase PEI-D-NH -HC1 is pH dependent (Figure II-7) ; therefore a well defined Bronsted plot can be

PAGE 47

36 4 2 id -2 — « PK. Figure II-5. Bronsted plot for the esterolysis of PNPA by simple primary and cyclic secondary amines. Values of k 2 and pK a determined at 25°C and 1.0 M ionic strength.' 2 '-

PAGE 48

37 '-] 20-2 -4 -6 t r i 10 12 PK. Figure II-6. Bronsted plot for the esterolysis of PNPA by (o) imidazoles and (D) anilines. Values of k and pK determined at 30°C in *7 3. in 28.5% ethanol . z '

PAGE 49

38 10 -T 64 2 1 10 P H Figure II-7. Plot of pK vs a , app pH for PEI-D-NH 2 -HC1

PAGE 50

39 constructed from the one system. The Bronsted plots for the reactions of PEI-D-NH 2 -HC1 with PNPA and PNPC are shown in Figure II-8. Included on the same coordinate system is the reference plot for the simple primary and secondary amines discussed earlier. It can be seen from the values of 6 for PNPA with PEI-D-NH -HC1, 0.81, and reference, 0.83, that the difference in pK dependence is insignificant for the two systems. Therefore the pH dependence for the PEI-D-NH 2 -HC1 reaction with PNPA is completely accounted for by the dependence of nucleophilicity on base strength. However the value of 3 for the reaction of PEI-D-NH 2 ~HC1 with PNPC, 1.06, is significantly different from the reference value. The dissection of the polymeric rate can be facilitated by assuming that k /a can be separated into two terms, k and E (Equation II-7) . If the term E is taken to be the rate of Equation II-7 log k /a = log (k «E) = log k + log E = BpK_ + C enhancement due to the polymeric environment, then the term k can be estimated from the reference amines. That is, if a the Bronsted equation for the reference amines (Figure II-5) is subtracted from the Bronsted equation for the polymer systems, then E can be determined as a function of pK and a thus a (Equation II-8) . The value of E then can be plotted against pK or a (Figure II-9) . The binding aspects of the polymer can be examined more closely via the PNPC reaction. The reactivity of the

PAGE 51

4 4.02.00.0" 6.0 3.0 10.0 pK a, app Figure II-8. Bronsted plot for the esterolysis of () PNPA PEI-D-NH 2 -HC1. The Bronsted line from Figure II-5 is included as a reference.

PAGE 52

41 3000_ 20001000100-O06.6

PAGE 53

42 Equation II-8 log k a -E = 6 p pK a + C p P = polymer system log k R = 3 R pK a + C R R = reference system log E B pK + C^ E = polymeric enhancement E 1 a E reference amines with PNPA is used to dissect the rates of the PEI-D-NH 2 -HC1 reaction with PNPC. In the case of the PNPC aminolysis the dependence of rate on pK is not a , app entirely resolved by the subtraction of the reference equation as in the case of PNPA. The caproate ester and acetate ester aminolyses must proceed by the same mechanism since the only structural change is replacement of a methyl group by a n-pentyl group. Although this change would not be expected to effect the reaction mechanism, the hydrophobic binding properties are greatly affected. Increased hydrophobic binding with reduced charge density then, must be the explanation for the additional pH dependence in the case of PNPC reaction. The above discussion adequately accounts for the pH dependencies of the reactions of both PNPA and PNPC with PEI-D~NH 2 -HC1. However, a significant pH independent enhancement remains unaccounted for. The rate constant for the PEI-D-NH 2 -HC1 reaction with PNPA at any given pK within the range studied is approximately 16 fold larger than a reference amine of the same pK . There are several possible explanations for this enhancement, for example, a solvent effect or pH dependent binding. Alternatively, the

PAGE 54

43 apparent enhancement may be due to error in concentration of primary amine and/or in the value of C. Unfortunately there is insufficient data to determine the cause of this enhancement. In spite of insufficient data, this ambiguity does not diminish the importance of the pK -rate correlation. a , app Summary The reaction of PEI-D-NH 2 -HC1 with PNPA and PNPC has been shown to proceed via nucleophilic attack of polymeric amine on ester. The rate of reaction is a function of the state of protonation of the polymer which in turn is a function of pH. Correction of the rate constant for fraction of nucleophiles protonated does not lift the pH dependency. However, the pH dependency can be quantitatively dissected by use of the Bronsted relation. The dependence of the PNPA reaction was found to be due to increased nucleophilicity of the polymeric amine with increasing pH. The PNPC reaction exhibits increased binding as well as increased nucleophilicity with increasing pH. The Bronsted relation has not been used directly before to quantitate the electrostatic influence in polyionic nucleophiles. The utility of this approach will be explored further in the next chapter.

PAGE 55

CHAPTER III HYDROLYSIS OF P-NITROPHENYL ESTERS CATALYZED BY POLYMER BOUND HETEROCYCLES Introduction In this chapter hydrolysis of p-nitrophenyl esters catalyzed by polymeric systems containing pendent nitrogenous heterocycles will be discussed. As was pointed out in Chapter I, primary amines and nitrogenous heterocycles release p-nitrophenol from p-nitrophenyl esters by the same mechanism. Therefore, the aminolysis reactions of Chapter II are used as models for the discussion. The Bronsted relationship will be employed and its applicability extended. Polyethylenimine Systems Preparation The polymer systems were prepared from PEI-D (described in Chapter II) and the appropriate derivatized heterocycle. PEI-D-Im-HCl was prepared by alkylation of PEI-D with 4 (5) -chloromethyl imidazole." 1 PEI-D-Pyr-HCl and PEI-D-APyrHC1 were prepared by reductive amination of PEI-D using, respectively, 4-pyridinecarboxaldehyde and N(2-pyridyl) -3aminopropionaldehyde with NaBH.. The fraction of the 44

PAGE 56

45 polymer derivatized was determined spectrally by NMR and by the C/N ratio from elemental analysis (Table III-l) . Polymer PEI-D-Im-HCl PEI-D-Pyr-HCl PEI-D-APyr-HCl Table III-l PEI UNITS BOUND TO HETEROCYCLES %__CH 2 CH 2 N Units Bound to Heterocycles Elemental Analysis 3 NMR b 26±4 26±3 17±4 12+2 46±25 25±5 Taken from Table IV-1. Taken from Table IV2. Primary Amine Analysis As shown in Figure 1-5, the reaction rates of PEI-DNH 2 -HC1 with PNPA are significantly higher than those of the heterocyclic systems PEI-D-Im-HCl, PEI-D-Pyr-HCl , and PEI-D-APyr-HCl with the same substrate. This observation adds an additional subject for consideration in the interpretation of the kinetic studies for the PEI bound heterocyclic systems. If the PEI bound heterocyclic systems contained residual nucleophilic primary amines, the kinetic data might easily be misinterpreted. That is to say, a small highly reactive fraction of residual primary amines might be the dominant factor in the esterolysis rates measured, Such data might lead one to assign spuriously high rates to

PAGE 57

46 the PEI bound heterocycle esterolysis reaction. It is clear that careful consideration must be given to the primary amine content of PEI bound heterocyclic systems. Equation III-l N

PAGE 58

47 that it was unlikely that any reagents would give unambiguous results for the fraction of residual primary amines present in the PEI bound heterocyclic systems. Having ruled out chemical methods of primarv amine detection, kinetic methods were attempted. In Chapter II the exhaustive isopropylation of PEI-D-NH--HC1 using Acetone/NaBH, z 4 to convert primary amines to secondary amines was discussed. Such a method, although not providing the number of primary amines present, would at least provide a PEI bound heterocyclic system free of primary amines. In Figure III-l the effects of repetitive isopropylation on the pH rate profile for the PEI-D-Im-HCl esterolysis reaction with PNPA can be seen. PEI-D-Im-HCl was exhaustively isopropylated to PEI-DIm-Ip-HCl with Acetone/NaBH 4 . Similar treatment of PEI-D-ImIp-KCl produced PEI-D-Im-Ip 2 -HCl . The results shown in Figure III-l illustrate two important features of the PEI-DIm-HCl. The esterolysis rate is very sensitive to isopropylation. The rate and the pH dependence decrease with increased isopropylation. Further, the rate reduction decreases with isopropylation from 40% reduction for the first isopropylation to 15% reduction for the second isopropylation (at pH=8.00). It follows that a third isopropylation would decrease the rate of reaction by less than 15% (at pH=8.00). A rate reduction of less than 15% is considered insignificant, and the pH rate profile of PEI-DIm-Ip -HC1 is assumed to be devoid of a primary amine component.

PAGE 59

48 3 _ 2 — P H Figure III-l. Effects of repetitive isopropylation of PEI-D-Im-HCl on PNPA esterolysis reactivity, (o) PEI-D-Im-HCl , (O) PEI-D-Im-Ip-HCl , (A)PEI-D-Im-Ip 2 -HCl.

PAGE 60

49 Further Kinetic Analysis 2 The PEI-D-Im-Ip -HC1 will be considered representative of the PEI bound heterocyclic systems. It is clear that the PEI bound pyridine systems show a pH dependency in the pH rate profile just as was shown for the PEI bound imidazole system. Furthermore there is very little difference between the pH rate profiles for the two pyridine containing polymers (Figure 1-5) . Therefore, the two systems must be reacting by the same mechanism precluding any significant bifunctional effects for PEI-D-APyr-HCl with PNPA. As in the PEI-D-NH 2 -HC1 system (Chapter II) , the effect of the possible PEI backbone reaction with PNPA must be deter2 mined in the reaction of PEI-D-Im-Ip -HC1 with PNPA. The rate constants k , which are based on concentration of 2. \j 3 2 backbone amine for PEI-D-Im-Ip -HC1 and PEI-D-Ip-HCl , are compared in Figure III-2. The rate constants of PEI-D-Ip2 HC1 are only 10% of the PEI-D-Im-Ip -HCl reaction, at most. Thus the backbone reaction can be ignored for the PEI-D-Im2 Ip -HCl system. However if the rate of the heterocycle containing polymer was much slower, the backbone reaction would be a serious complication. The state of ionization of the imidazole in PEI-D-Im2 Ip -HCl cannot be determined directly by spectrophotometric methods because the amine backbone interferes with the imidazole absorbance. However, the PEI-D-Pyr-HCl system does not suffer such interference, and the pendent pyridine was found to be completely nonprotonated in the pH range studied. Therefore it is reasonable to assume that the

PAGE 61

50 50 — .5 25 — s -e7.5 8.5 P H Figure III-2. Comparison of PNPA esterolysis rates for (o)PEI-D-HCl and (D) PEI-D-Ip-HCl .

PAGE 62

51 imidazole in PEI-D-Im-HCl is unionized throughout the large part of the range. Since the imidazole is virtually nonprotonated in the region of interest, its pK is undefined in this reqion (Chapter I) . Without pK data a Bronsted plot cannot be constructed. Even though the pK data are not available it is valid to assume that basicity increases with pH just as it does in the PEI-D-NH 2 -HC1 system (Chapter II) . It was found in Chapter II that the nucleophilicity , and hence the reactivity, of the PEI-D-NH 2 ~HC1 system was dependent upon the pK & or basicity of the polymer. By analogy one would then expect that the reactivity of the PEI-D-Im-Ip 2 -HCl would increase with pH as is observed in Figure III-l. Summary There are several important statements to be made concerning the above observations and conclusions. The fact that the rate of reaction is so dependent on isopropylation indicates that there may well be a significant contribution to the esterolysis rate by residual primary amines (backbone reaction) in the PEI-D-Im-HCl polymer. The residual primary amine content in the presence of imidazole is difficult to determine as discussed earlier. Klotz et aL 12 prepared a system very similar to PEI-D-Im-HCl . Their polymer only differed in imidazole content (15%) and degree of ionization. They tested for primary amine content by the ninhydrin method and found that it did not exhibit the characteristic ninhydrin color. This result led to the conclusion that

PAGE 63

52 there were no primary amines present. Such a result is not surprising, as the parent system (PEI-D-NH--HC1) which contains 20% primary amine also does not exhibit the expected "Ruheman's Purple" (a deep purple blue) of ninhydrin with 29 primary amines. " The parent system does produce a dull brown color, possibly an oxidation product, but it in no way resembles the true ninhydrin color. Further, it would be expected that their polymer would contain even more primary amines than PEI-D-Im-HCl due to a smaller fraction of the backbone alkylated with methyleneimidazole units, 15% as opposed to 29%. The presence of primary amines in the polymer 12 studied by Klotz et al . might indicate a significant primary amine component in the presumed imidazole acylation rate. Another important point is the ability to qualitatively predict a pH dependence for the PEI-D-Im-Ip -HC1 system based on the Bronsted correlation found for PEI-D-NH--HC1 . This example begins to show the usefulness of this application of the Bronsted relationship. The application to non PEI systems follows. Non Polyethylenimine Homopolymer Systems Unlike the PEI bound imidazole above, there are some examples of polymer bound heterocycles for which Bronsted plots can be constructed. Letsinger and Saveride 25 studied poly-4-vinylpyridine catalyzed 2 , 4-dinitrophenyl acetate (DNPA) hydrolysis. They found that although 4-methylpyridine produced the expected linear, rate vs a plot, the

PAGE 64

53 poly-4-vinylpyridine exhibited a curved plot (Figure III-3) . They qualitatively attributed the curvature in the polymer reaction to decreased nucleophilicity with increasing protonation (decreasing a). Overberger et al. found very similar results in poly-4 (5) -vinylimidazole versus imidazole catalysis of hydrolysis of PNPA (Figure I-la) . They attributed the curvature to several possible bifunctional mechanisms as discussed in Chapter I. The rate vs a plot for the reaction of PEI-D-NH -HC1 with PNPA is shown in Figure III-4. Similar to the plots for poly-4 (5) -vinylimidazole (Figure I-la) and poly-4-vinylpyridine (Figure III-3) the rate vs a plot for PEI-D-NH 2 -HC1 is curved. Since the Bronsted relationship was able to quantitatively account for the PSI-D-NH 2 -HC1 pH dependence of PNPA esterolysis, it is reasonable to apply the relationship to the above reactions of poly-4-vinylpyridine (Figure III-5) and poly-4 (5) -vinylimidazole (Figure III-6) . As in the case of the PEI-D-NH-HC1 reaction with PNPA (Figure II-8) the Bronsted plots are linear. The plot for the poly-4-vinyipyridine system produces a large value of 3 (1.3); however, small molecule pyridine 5 values are typically larger than 1. Furthermore., the slopes of the pH-rate profiles for the pyridine containing PEIsystems, 0.69 and 0.74 (Figure 1-3 and Figure 1-2) , were larger than the slope of the imidazole system (0.56, Figure 1-4). The poly-4 (5) vinylimidazole Bronsted slope, 0.8, is precisely that expected from the portion of the small molecule imidazole plot shown on the same plot. Therefore the interpretation of Letsinger and

PAGE 65

54 'igure III-3. Plot of k , against a for the esterolysis of , obs 2-4-dmitrophenyl acetate by (D) poly-4-vmylpyridine and (o) pyridine. Values of k and a determined that 3 6 in 50% aqueoul ethanol.25

PAGE 66

55 800^ 400 — | 0.50 0.60 0.70 0.80 Figure III-4. Plot of k„ against a for the esterolysis of PNPA by PEI-D-NH-HCl.

PAGE 67

56 1. 80— t 1.40 *> 1.00 0.40 0.20 _ 6.2 6.6 7.0 7.4 7.8 pK a, app Figure III-5. Bronsted plot for the esterolysis of PNPA by (A)poly-4 (5) -vinylimidazole and (o) imidazoles Values of k cat and pK a determined at 30°C in 28.5% ethanol . ^ ' *7 From Figure II-6

PAGE 68

57 1.8 — 1.7 1.6 — 1.5 — 1.4 — 1.3 3.0 3.2 3.4 3.6 3.8 pK a , app Figure III-6. Bronsted plot for the esterolysis of 2, 4-dinitrophenylacetate by poly(4) -vinylpyridine. Values of pK a and k 2 determined at 36.8°C in 50% aqueous ethanol. 5

PAGE 69

58 25 Saveride would seem to be sufficient for the poly-4(5)vinylimidazole system. That is, the curvature in the rate vs a plot is due to increasing pK of the nucleophile. The increase in pK is in turn due to the reduction of the positively charged electrostatic field as a increases. At low values of a the highly charged polymer increases the energy required for formation of additional positive charge. As a increases the charge on the polymer decreases, thereby reducing the energy required for formation of positive charge and increasing pK and reaction rate. 3. Conclusion The correlation of pK with reaction rate is quite effective in quantifying pH dependencies in polymer esterolysis reactions for pH regions in which the nucleophile is partially ionized. If the nucleophile is completely nonprotonated the pK cannot be determined. However, the a , a.pp ' pK a,app and thus nucleo Philicity are dependent on the state of ionization of the polymer. Therefore it is reasonable to predict that, if polymeric functions other than the nucleophile undergo ionization with pH changes, the nucleophilicity (and thus rate) will be pH dependent. The pH dependencies of the heterocycle containing PEI systems can be explained as a pK dependence. Bifunctional effects were ruled out for the pyridine containing PEI systems due to the similarity of the pH-rate profiles. It also seems very unlikely that the imidazole containing PEI reaction with PNPA has a significant bifunctional component in view

PAGE 70

59 of the correlation of the pK and rate for poly-4(5)a , app sr jl w/ vinylimidazole. As pointed out in Chapter I this conclusion does not necessarily rule out the possibility of a Afunctional mechanism for the polymer systems discussed. The fault may be in the choice of the ester. Our application of the Bronsted plot to quantitate pH dependencies is a very important development in the field of polymer esterolysis. It has been shown that polymers of varying make up have different reactivities. For example, 31 Shxmidzu et al. have plotted esterolysis rate constants against the intrinsic pK (see Appendix) for a series of imidazole containing polymers. The work of Shimidzu shows clearly that increasing acrylic acid content in a series of acrylic acid and vinylimidazole copolymers increases the imidazole pK and thus the esterolysis rate. However, up to now Bronsted correlations for polymer systems have been for series of polymers. Such correlations cannot account for pH dependencies as can Bronsted correlations for a single polymer system plotting pK vs rate constants. This type a , app 2 c of Bronsted correlation has proven valuable in explaining both the pH dependencies generated by the research herein as well as that of others. Perhaps the best test of a tool is how well it works. The correlation of pK to esterolysis a , app J rate constants has been shown here to work well for the polymer systems tested.

PAGE 71

CHAPTER IV EXPERIMENTAL Introduction Melting points and boiling points are uncorrected unless otherwise noted. Melting points were determined with a Thomas Hoover Unimelt capillary melting point apparatus. Boiling points were determined by conventional techniques. Nuclear Magnetic Resonance spectra were recorded using a Varian A6 0-A spectrometer. Chemical shift (6) data were reported in parts per million (ppm) from the appropriate internal reference, either tetramethylsilane (TMS), 2,2-dimethyl-2-silapentane 5-suifonic acid (DDS) , or water d, (HOD) . Solvent evaporation was performed at reduced pressure using a Buchler Instruments flash evaporator. Lyophilisation or freeze drying was carried out using an apparatus and 3 2 technique similar to that described by Vogel . The aqueous dialyses were carried out in a 3 liter filter flask connected to a deionized water tap and equipped with a stir bar. The polymer solutions to be dialyzed were transferred to dialysis bags prepared from Union Carbide (36 100 ft dialysis membrane) tubing according to the method of Gabbay et al. JJ and generously supplied by Shau-Fong Yen. In the case of 6

PAGE 72

61 Dialysis against a stream of deionized water, the tap was opened to allow a moderate flow of water through the flask while stirring. In the case of dialysis against 0.1 M HC1 , the tap was closed, and 30 ml concentrated hydrochloric acid was added. When dialyzing against ethanolic solvents the bag was stirred in a 500 ml flask with the solvent. Temperature control was provided by a Lauda K-2/R thermostat. All UVvisible spectrophotometry measurements were made on a Cary 17D spectrophotometer. The pH measurements were made with a Beckman Research pH meter in conjunction with a Radiometer GK 2321C electrode. Syntheses Dodecyl Polyethylenimine PEI-D Dow PEI 600 polyethylenimine solution (128 g) was freeze dried for 8 hr and then allowed to warm to room temperature while maintaining vacuum (0.05 torr) for an additional 36 hr . The gelatinous material (49 g, 1.2 mol) was dissolved in 470 ml argon saturated anhydrous ethanol. Dodecyl iodide (35 g, 0.12 mol) in 5.0 ml of anhydrous ethanol was added to the polymer solution. The reaction flask was sealed and maintained at 45°C for 120 hr . The reaction mixture was quantitatively transferred to a one liter volumetric flask and diluted to the mark with argon saturated anhydrous ethanol. This solution was stored under argon and used as a stock solution in the following preparations (1.2 M in ethylenimine units).

PAGE 73

62 Dodecyl Polyethylenimine Hydrochloride PEI-D-NH ? -HC1 A solution of dodecyl polyethylenimine was prepared by diluting PEI-D (25 ml, 0.030 mol) up to 150 ml with absolute ethanol. This solution was added carefully to a 500 ml beaker containing 5 ml concentrated hydrochloric acid in 50 ml absolute ethanol at 0°C. The resultant salt was filtered and washed with absolute ethanol. Dodecyl-4-methylenepyridine-polvethylenimine Hydro chloride PEI-D-Pyr-HCl " " Freshly distilled 4-pyridinecarboxaldehyde (BP 191-192°C under N 2 , 0.96 g, 0.0090 mol) was dissolved in 25 ml absolute ethanol in a 125 ml erlenmeyer flask equipped with a stir bar. PEI-D (25 ml, 0.030 mol) was added to the flask. The cloudy suspension was stirred for 2 hr under N„ . A 50 ml solution of NaBH 4 (0.13 g, 0.0022 mol) was added to the flask at 0°C. The reaction was stirred for 2 hr. The reaction mixture was transferred to 4 tubes. One milliliter concentrated hydrochloric acid was added to each tube to form the polymer salt, and the tubes were then centrifuged. The precipitated polymer salt was washed twice with absolute ethanol. The polymer salt was dialyzed against 500 ml absolute ethanol twice. The ethanolic dialysate was passed through a sephadex LH-2 column. The eluant was transferred into 4 tubes. The salt was reprecipitated by addition of one milliliter concentrated hydrochloric acid to each tube. The tubes were centrifuged, and the salt was washed twice with absolute ethanol. The polymer salt was dryed producing a white powder (1.12 g, 0.00966 mol, 32%).

PAGE 74

63 Dodecyl-N(2-pyridyl) -3-propylamine-polyethylenimine Hydro chloride PEI-D-APyr-HCl ' ~ Ten milliliters of an absolute ethanol solution of N(2-pyridyl) -3-aminopropionaldehyde hydrochloride (0.65 g, 0.0031 mol) was prepared in a 125 ml erlenmeyer flask with gentle heating. A stir bar and PEI-D (10 ml, 0.012 mol) was added to the flask which was stirred for one hour under N„ . A 10 ml absolute ethanol solution of NaBH 4 (0.092 g, 0.0024 mol) was added to the reaction mixture. After stirring for an hour, the reaction mixture was transferred to a tube. The polymer salt was precipitated by addition of one milliliter concentrated hydrochloric acid. After centrifuging , the polymer salt was washed twice with absolute ethanol. The polymer salt was dialyzed against a stream of deionized water for 24 hr . The polymer salt was then dialyzed against 500 ml absolute ethanol overnight. The dialysate was passed through a sephadex LH-20 column. The polymer salt was reprecipitated by the addition of one milliliter of concentrated hydrochloric acid. The polymer salt was dried producing an off-white solid (0.35 g, 0.0023 mole, 19%). Dodecyl-4 (5) -methylenimidazole-polyethylen imine Hydrochloride PEI-D-Im-HCl ~ To a solution of PEI-D (25 ml, 0.030 mol) in a sealable tube was added 4 ( 5) -chloromethylimidazole hydrochloride (1.59 g, 0.10 mol), triethyiamine (2.5 g, 0.025 mol) and a stir bar. The tube was sealed and maintained at 6 5°C in an oil bath for 36 hr. After cooling the tube was opened and the contents transferred to two test tubes. One milliliter of concentrated hydrochloric acid was added to each tube forming a heavy

PAGE 75

64 precipitate. The tubes were centrifuged and the supernatant decanted. The precipitates were washed with absolute ethanol, centrifuged, and the supernatant decanted. The precipitate was dissolved in 20 ml deionized water and dialyzed against a stream of deionized water for 72 hr . The polymer was then dialyzed with 500 ml absolute ethanol three times. The polymer solution was removed from the dialysis bag and passed through a sephadex LH-20 column. The salt was reprecipitated with 4 ml concentrated hydrochloric acid. The salt was washed twice with absolute ethanol. The polymer salt was dried producing an off-white solid (1.75 g, 0.0143 mol, 48%). Dodecyl-4 (5) -methylenimidazole-isopropyl-polyethylenimine Hydrochloride PEI-D-Im-Ip-HCl ~ PEI-D-Im-HCl (0.0300 g) was dissolved in 30 ml of water. This solution was treated with solid NaBH. to bring the pH up to approximately 7. After freeze drying the polymer was suspended in 50 ml absolute ethanol containing one gram acetone by the dropwise addition of 6 N hydrochloric acid. After one hour NaBH 4 (0.4 g, 0.01 mol) in 10 ml absolute ethanol was added causing precipitation. A routine of suspension of the polymer with 6 N hydrochloric acid followed by addition of acetone (5 g, 0.8 mol), then one hour later addition of NaBH (0.4 g, 0.01 mol) in 10 ml absolute ethanol was followed twice. The polymer was reacidified with 6 N hydrochloric acid and allowed to stir for 12 hr to complete hydrolysis of the residual NaBH 4 . The solution was dialyzed for 24 hr. The polymer solution was centrifuged then filtered. The filtrate was freeze dried producing a white solid (0.281 g, 94%).

PAGE 76

65 Dodecyl-4 (5) -methylenimidazole-isopropyl-isopropyl-polyethyl enimine Hydrochloride PEI-D-Im-Ip 2 -HCl PEI-D-Im-Ip-HCl (0.098 g) was dissolved in 10 ml of water. This solution was heated wiuh solid NaBH. to bring the pH up to approximately 7. After freeze drying the polymer was stirred in 20 ml absolute ethanol containing one gram acetone for one hour. The polymer was suspended by dropwise addition of 6 N hydrochloric acid and 5 ml absolute ethanol containing one gram acetone was added. After one hour NaBH 4 (0.4 g, 0.01 mol) was added in 10 ml of absolute ethanol followed by suspension of the polymer salt with 6 N hydrochloric acid one hour later. The reaction mixture was stirred under a stream of N„ overnight. The polymer was dialyzed against a stream of deionized water for 24 hr followed by 0.1 N hydrochloric acid for 3 hr. The polymer solution was centrifuged, filtered then freeze dried producing a white solid (0.062 g, 63%). 4 (5) -Chloromethylimidazole Hydrochloride 4 (5) -Chloromethylimidazole hydrochloride was prepared according to the procedure of Turner et al . J from 4 (5) -hydroxymethylimidazole and thionyl chloride (mp 138140°C, lit mp 138-141°C) . 4(5) -Hydroxymethylimidazole Hydrochloride In a mixture of 50 ml concentrated hydrochloric acid, 125 ml water and 250 ml benzene 4 (5) -hydroxymethylimidazole picrate (69 g, 0.21 mol available from Eastman) was dissolved with heating. The benzene which developed a

PAGE 77

66 yellow color was decanted. The aqueous solution was continuously extracted with 125 ml benzene layer for 9 hr. After decanting the benzene layer, the aqueous layer was evaporated under vacuum at 70°C. The yellow brown residue was recrystalyzed from ethanol ether (23 g, 81% mp 103-106°C, lit mp 107-109°C) . 3-Bromopropanal Dimethyl Acetal 3-Bromopropanal dimethyl acetal was prepared according 35 IS to the procedure of Pineau: bp 52°C/10 mm Hg , lit bp 59°C/12 mm Hg; nmr CDC1 3 5 2.12 (doublet of triplets, J=5 Hz, J=8 Hz, 2H) , 3.36 (singlet, 6H) , 3.43 (triplet, J=8 Hz, 2H) , 4.53 (triplet, J=5 Hz, 1H) . N(2-pyridyl) -3-aminopropanal Dimethyl Acetal The synthesis of N(2-pyridyl) -3-aminopropanal dimethyl acetal was carried out according to the procedure described 3 6 by Reynaud et al . The product was an amber oil: bp 154°C/10 mm Hg , lit bp 153°C/12 mm Hg; nmr (CDC1 ) d 1.92 (quartet, 2H) , 3.32 (singlet, 6H) , 3.35 (quartet, 2H) , 4.50 (triplet, 1H) 4.75-5.20 (mult., 1H) 6.25-6.75 (mult, 2H) , 7.20-7.55 (mult., 1H) , 8.05 (doublet of doublets, 1H) . N(2-pyridyl ) -3-aminopropanal Hydrochloride N(2-pyridyl) -3-aminopropanal dimethyl acetal (1.3 g, 0.0064 mol) was dissolved in 5 ml concentrated hydrochloric acid and heated on the steam cone for one hour. The solvent was evaporated and the brown residue was dried under vacuum (0.40 g, 0.0018 mol, 28%); mp 172-172. 5°C.

PAGE 78

67 p-Nitrophenyl Acetate PNPA P-nitrophenyl acetate was prepared according to the 3 7 procedure of Bender and Nakamura. The solid p-nitrophenyl acetate was recrystalyzed twice from ethanol; mp 78-78. 5°C, lit mp 37 77.5-78°C. p-Nitrophenol Caproate PNPC P-nitrophenol (14 g, 0.1 mol) was dissolved in 25 ml dry pyridine. The pyridine solution was added to caproylchloride (13 g, 0.1 mol) in a 100 ml round bottom flask equipped with a reflux-condenser and stir bar. The reaction was refluxed for 15 hr. After cooling, the contents of the reaction flask were added to 40 ml of ice water. The resulting oily layer was separated and the aqueous layer extracted twice with 25 ml diethyl ether. The combined oil and ethereal extract was washed 3 times with 50 ml water, 5 times with 50 mi 5% aqueous hydrochloric acid, and 6 times with 5% aqueous sodium bicarbonate (each washing had to be salted out) . The ethereal layer was dried and stripped producing 17 g of crude material. The crude product was distilled producing a viscous oil (13 g, 0.055 mol, 55%); bp 150°C/2 mm Hg, lit 38 bp 145°C/1 mm Hg . Composition of PEI Derivatives by Elemental Analysis Elemental analysis data were used to calculate the values of D, (fraction CH 2 CH 2 N units dodecylated) , z (fraction of CH-CH^N units bearing a heterocycle) , mole unit weight (average weight of a mole of CH-CH-N units after

PAGE 79

68 derivitization) , and a initial (in the case of PEI-D-NH^HCl) . The C/N ratio was used to calculate both D and Z. The carbon to nitrogen weight ratio is equal to the C/N ratio from elemental analysis. Therefore the number of carbons per unit (2 from the CH 2 CH 2 N unit, 12D from the dodecyl group, and ZX from the heterocycle, X is the number of carbons in the heterocycle) divided by the number of nitrogens per unit (one from the CH 2 CH 2 N unit and ZY from the heterocycle, Y is the number of nitrogens in the heterocycle) multiplied by the C/N atomic weight ratio is equal to the C/N ratio from elemental analysis (Equation IV-1) . This equation is readily rearranged to solve for either D or Z. In the case of PEI-D-NH 2 -HC1 the heterocycle terms drop out and the equation is in a single unknown, D. For the PEI derivatives containing heterocycles the value of D calculated for PEI-DNH 2 ~HC1 was used. The error limits in these calculations are based on an assumed error of 0.2 in the elemental analysis. Values of C/N were calculated using C+0.2/N-0.2 and C-0.2/N+0.2 which were substituted in the appropriate equations to calculate error limits. The mole unit weight was calculated from the relationship that % of an element in a compound is equal to the weight of that element in the compound divided by the formula (units) weight of the compounds multiplied by 100 (Equation IV-2) . The elemental analysis values for carbon and nitrogen together with the calculated values for D and Z were used for these calculations (Equation IV-30). The error calculation for the unit

PAGE 80

69 mole weight was based on the difference between the average from the carbon and nitrogen calculations and the highest possible value from the carbon based calculation and lowest possible value for the nitrogen based calculation consistent with the assumed error of ±0.2 in elemental analysis data. The value of a initial was calculated for the PEI-D-NH -HC1 titration based on the assumption that all analyzed chlorine was present as the hydrochloride salt. The ratio of gram atoms of chlorine to gram atoms of nitrogen in the polymer represents the fraction of CH 2 CH 2 N units protonated, 1-a initial (Equation IV4) . The error calculation for a initial was made assuming an error of 0.02 in elemental analysis data as in previous calculations (Table IV1) . Equation IV1 fl2.01 C/N = 2 + 12D + ZX 1 + ZY 14.01 D IC/N i|^i (1 + ZY) 2 + 2x] ^ •' = v* && * h C/N TJfgT 2 12 ° x " C / N nfSr Y Equation IV2 element = wei 9 ht o f element X 100 ' ' formula weight (mole unit weight

PAGE 81

70 LD rH CN rf 3 O

PAGE 82

71 1

PAGE 83

72 En u a Cm CO « 2 O « Q Eh < U Cm I a i H W u X I M >i Cm < I a i H W Cm r-

PAGE 84

73 Equation IV3 mole unit weight ' 2 ^f^f "01 mole unit weight . (1 ^^^1 Equation IV4 , %Cl/35.45 a '" %N/14.01 Composition of PEI Derivatives by NMR Spectral Analysis The NMR spectra and peak areas (see Figure IV1) of the polymer salts were recorded (Table IV2) . The integrated areas per hydrogen (for example, area of imidazole peak divided by 2) was used to calculate the extent of substitution on the polymer. In the case of PEI-D-NH„-HC1 the fraction of CH-CH-N units alkylated was determined by the ratio of dodecyl area per hydrogen to the CH-CH^N area per hydrogen (corrected for the N terminal methylene of the dodecyl group) . In the heterocycle containing systems the dodecyl peak was used as a reference instead of the CH CH.N peak. Determination of Primary Content of PEI Derivatives Using Trinitrobenzene Sulfonate (TNBS) The primary amine content of PEI derivatives was determined using trinitrobenzenesulfonate (TNBS) by a method similar to that used by Johnson and Klotz . To a 25 ml 1

PAGE 85

74 volumetric flask was added 0.100 ml of an aqueous PEI solution (0.0001 units mol) , 5 ml 4% aqueous sodium bicarbonate solution, and one milliliter 0.1% aqueous TNBS solution. The flask was heated 0.5 hr at 3 7°C, cooled to room temperature, filled to the mark with galatial acetic acid, and the absorbance measured at 340 nm. The percentage of residual primary amine was calculated based on the extinction coefficient used by Johnson and Klotz, 12000 M -1 (Table IV-3) . TABLE IV-3 PRIMARY AMINE CONTENT OF PEI DERIVATIVES USING TNBS Unit Mole % Primary PEI Derivative Absorbance Concentration of PEI Amine PEI-DNH 2 -HC1 0.972±0.013 4.08 x 10~ 4 19.8+0.4 PEI-D-Ip-HCl 0.058±0.001 3.44 x 10~ 4 1.4±0.2 Blank 0.007+0.007 Primary Amine Detection by Ninhydrin The efficacy of primary amine detection by ninhydrin was compared for PEI-D, ethylenediamine, and glycine according to 39 the procedure suggested by Pasto and Johnson. A strip of filter paper and a strip of chroma togram sheet (silica gel) was spotted with 0.1 M aqueous solutions of the amine systems to be compared. After oven drying the strips were

PAGE 86

75 sprayed with 0.25% ninhydrin solution and developed for five minutes at 100 °C. Additionally 0.1 M solutions of the amines to be compared were treated with one milliliter of 0.25% ninhydrin solution and gently warmed for 0.5 hr. The glycine 39 results were those expected: blue solution and purple blue 40 . . spots. " Ethylenediamine did not react with the ninhydrin and produced no color changes; similar results have been 41 previously reported. The PEI-D spots were brown and the solution was yellow. Relative Concentrations of Imidazole Containing_Polymer Solutions The relative concentrations of stock solutions of the three imidazole containing polymers PEI-D-Im-HCl, PEI-D-Im2 HC1, and PEI-D-Im-Ip -KCl used for kinetics were determined spectrophotometrically . A series of 7 solutions were prepared from the PEI-D-Im-HCl stock solution in 1.0 M hydrochloric acid. The absorbencies of these solutions ranging in concentration from 2.24 x 10 to 28.0 x 10 molar in imidazole units were measured at 210 nm and 25.0±0.3°C. A Beer's law plot was constructed from this data. The least squares line passed through (within experimental error r=0. 99998) and had a slope, extinction 3 1 coefficient, of 6.79 x 10 M . Solutions (2 each) of the isopropylated imidazole containing polymers were prepared from the respective stock solutions in the same way (ca 10 ~ M) and their absorbancies measured. The concentrations of the stock solutions were calculated from the extinction coefficient of the PEI-D-Im-HCl system. The

PAGE 87

76 concentrations of PEI-D-Im-Ip-HCl and PEI-D-Im-Ip 2 -HCl solutions then were determined relative to the PEI-D-Im-HCl Potentiometric Titration of PEI-D-NH n -HC1 The PEI-D-NH 2 -HC1 (0.0931 g) was dissolved in 50.0 ml of 0.100 M KC1 after drying under vacuum for 10 days. The solution (0.0192 M in CH 2 CH 2 N units) was titrated with C0 2 free KOH (0.09977 N) at 25.0±0.3°C under a nitrogen atmosphere. After each addition of titrant the pH was recorded upon equilibration. The values of pK and a were calculated from the a , app titration data. The value of a at each pH was determined by a modification of the method suggested by Albert and 42 Serjeant to calculate the ratio of nonprotonated species (Equation IV-5) . The value of KOH refers to the initial Equation IV-5 [KOH] + a. nt [CH 2 CH 2 N] 10 _pH + 10 P H 14 00 [CH 2 CH 2 N] + 10" pH i P H 14 00 concentration of potassium hydroxide. That is, the total moles of titrant added to bring the pH to a given value divided by the volume of the solution. The value of a int is the fraction of nonprotonated CH CH N units in PEI-D-NH,,HC1 as prepared and determined from the Cl/N ratio. The exponential terms in pH serve to correct the hydroxide ion concentration for water ionization. The value of pK a , app was calculated at each value of pH by use of the Henderson19 Hasslebalch equation (Equation II-3, Table IV-4).

PAGE 88

77 TABLE IV4 POTENTIOMETRIC TITRATION DATA FOR PEI-D-NH 2 -HC1 Titrant (ml)

PAGE 89

78 Equation II-3 P K a = P H + log ilSL Preparation and Standardization of CO ,, Free 0.1 N KOH Titrant An ampule of J. T. Baker Dilut-it was quantitativelytransferred to a 1000 ml volumetric flask as per enclosed instructions. The contents of the volumetric flask were diluted to the mark with "boiled out" water. The titrant was standardized with oven dried (120°C, one hour) primary standard pottasium hydrogen phthalate (KHP) . The standardization was repeated seven times (N=0 . 09977±0 . 09%) . Preparation of Ester Solutions Used for Kinetic Studies Either p-nitrophenyl acetate (0.010 g, 5.5 x 10 mol) or p-nitrophenyl caproat (0.020 g, 8.4 x 10 ) was added to a 10 ml volumetric flask. The flask was filled to the mark with acetonitrile dried by refluxing 3 hr over P~0rfollowed by distillation. The ester solutions were stored in tightly sealed brown bottles. Preparation of Polymer Solutions for Kinetic Studies Into a dry tared 10 ml volumetric flask was weighed 0.1 g of the polymer salt. The flask and contents were dried for three days under vacuum. After drying the flask was reweighed to determine the dry polymer weight. The flasks were then partly filled with glass distilled water and shaken to dissolve the polymer. After standing overnight

PAGE 90

79 to complete the dissolution as well as to allow foam to dissipate, the flasks were filled to the mark. The solutions were shaken and stored in tightly closed brown bottles. Buffer Solutions Tris(hydroxymethyl)aminomethane, tris, was used as the buffering agent in the kinetic experiments. Two stock buffer solutions were prepared both 0.1 M in tris and ionic strength. A solution of tris HC1 was prepared by adding primary standard tris (12.14 g, 0.100 mol) and the contents of an acculate 1/10 N hydrochloric acid ampule to a 1000 ml volumetric flask and diluting to the mark with glass distilled water. A solution of tris was prepared by adding primary standard tris (12.14 g, 0.1000 mol) and certified ACS potassium chloride (7.46 g, 0.100 mol) to a 1000 ml volumetric flask and diluting to the mark with glass distilled water. These solutions (both 0.100 M in buffer and ionic strength) were combined to form solutions of the correct pH. Kinetic Method To a 4 ml cuvette was added polymer or ethyenediamine stock solution, water and acetonitrile. Amounts of water and acetonitrile were added such that the total volume of polymer stock solution and water was 0.100 ml and the total volume of acetonitrilic ester solution (added later) and acetonitrile was 0.050 ml. Buffer solution (3.00 Ml, 0.100 M tris, 1=0.100 M) was pipetted into the cuvette. The pH of the solution was recorded at room temperature, 25±2°C. The

PAGE 91

80 cuvette was placed in a spectrophotometer and equilibrated to 25.0r0.3°C. The ester solution was placed on a glass rod with a flattened tip which was used to add the ester and stir 14 the solution. The rate of increase in p-nitrophenoxide anion concentration was observed at 400 nm until no change in absorbance, A, was observable to determine A . The CO observed rate constants were calculated from the values of A oo~ A t and 6 (time) using a least squares routine of a TI-58 programable calculator. Only the initial portion of the reaction was used to determine the rate constants due to p-nitrophenol inhibition. The rate constants were then corrected for the background rate by subtraction of the value of the background rate constant at the appropriate pH . These corrected pseudo first order rates were then divided by concentration of CH 2 CH 2 N units, nucleophile concentration or protonated nucleophile concentration (PEI-D-NH^-HCl) to determine k G3 , k 2 or k /a, respectively. In those cases where the effect of p-nitrophenol was to be observed, the p-nitrophenol in acetonitrile was added before the buffer, but the total acetonitrile content remained 0.050 ml or 1.6% (Tables IV-5 through IV-12) . Background Rate The rate of p-nitrophenyl acetate esterolysis without polymer was measured at 7 pH values in the region of interest. The background rate constants (k ) were found to fit EguaBG ^ tion IV6 (Figure IV2) . The value of k 43 b (7.0 x 10~ 2 M" 1 min -1 ), k, 44 (5.70 x 10 2 M _1 min" 1 ) and k 43 Un W

PAGE 92

81 Equation IV6 k v [tris] k b ~" ' T ° tal + k (lO 14 ^") + k [H,0] BG . ,.8.08-pH OH' ' w L 2 (6 x 10 M min ) were available from the literature. The values from this equation were then used for the background rate. Pyridine Ionization in PEI-D-Pyr-HCl The pyridine in PEI-D-Pyr-HCl exhibits a pH dependent absorbance at 259 nm. In 0.1 N hydrochloric acid solution the extinction coefficient based on 2 measurements was (5800±400) . In 0.1 M tris buffer solution at pH • s 6.71, 7.28, and 7.75 the extinction coefficients were found to be 3020, 3030, and 2970, respectively. These values indicate that the state of ionization of pyridine is not changing over pH 6.71 under these conditions.

PAGE 93

82 6.6 — 3.3 — PH Figure IV-2. Plot of background PNPA esterolysis rate. The line was calculated from Equation IV6 .

PAGE 94

83 TABLE IV5 PH ESTEROLYSIS RATES FOR PEI-D-NH -HC1 % Reaction a K M min [PEI-D-NH ? -HCl] [PNP] [PNPA] C [PNPC]° Followed 6.94

PAGE 95

84 TABLE IV6 ESTEROLYSIS RATES FOR PEI-D-Ip-HCl P H

PAGE 96

85 TABLE IV8 ESTEROLYSIS RATES FOR PEI-D-Im-Ip-HCl k„ M mm £H_ 2 [PEI-D-Im-Ip-HCl]' 6.90

PAGE 97

86 TABLE IV10 ESTEROLYSIS RATES FOR PEI-D-APyr-HCl PH

PAGE 98

87 TABLE IV11 ESTEROLYSIS RATES FOR PEI-D-Pyr-HCl PH

PAGE 99

APPENDIX Definition of pK-, in Polymer Systems The multiplicity of pK definitions for polymer systems in the literature creates some ambiguity as to the meaning 4 fi of polymer pK . Kunitake and Shinkai have defined three different polymer pK values. pK = pK + n'Logi— ^ pK = pH + Log^^a , app c ^ a PK . _ = P K °43AG el a , xnt a , app RT The following passages quoted from Morawetz helps clarify the issue: With a polymer carrying a large number of ionizable groups, it is obviously impracticable to specify the successive ionization constants. Instead of this, we define the apparent ionization constant K aBp of an average ionizable group carried by the polyion in the usual manner by (H + )a/(l-a 1 ) = K a PP where K will, of course, vary with the degree of ionization since the charged polymer will interact with the hydrogen ions. With polymeric acids the polyanion will attract the hydrogen ions and ^ K app/^ a i <0 '" with polymeric bases, on the other hand", the hydrogen ions will be repelled by the polycation and the acid strength of the polymer will increase with its charge density. If the required electrostatic free energy for the removal of an

PAGE 100

89 equivalent .of protons at a given degree of ionization is {±0"" -. (a.) then el l ' pK = pK°-0.43A(?i 1 (a, )/RT r app l el i ' where K° is characteristic of the ionizing group under conditions where electrostatic interactions with other ionizing groups are absent. The investigation of base strength-reactivity effects requires an accurate measure of basicity. The base strength of a poly-functional base varies with a. Therefore the value of pK used in the correlation of rate with reactivity must take into account the effect of a on basicity. The intrinsic pK (pK, , pK ..or pK°) is devoid of a dependence. The value of pK , on the other hand, is a a , app function of a and is determined readily from pH and a data. The use of pK obviates the necessity of determininq a , app 3 the value of AG as would be required if intrinsic pK 's were used (see above) . Therefore pK is the paramete: a , app of choice for the base strength-reactivity correlation.

PAGE 101

REFERENCES For review see: a. T. Kunitake in "Bioorganic Chemistry," Vol. 1, E. E. van Tamelen, Ed. (Academic Press, New York, 1977) , pp 153-171. b. T. Kunitake and Y. Okahata in "Advances in Polymer Science," Vol. 20, H.-J. Cantow, Ed. (SpringerVerlag, Berlin, 1976), pp 159-221. c. C. G. Overberger, A. C. Guteryl, Jr., Y. Kawakami , L. J. Mathias, A. Meenakshi and T. Tomono, cure and Appl. Chem. , 50_, 309 (1978) . d. C. G. Overberger and T. W. Smith in "Reactions on Polymers," J. A. Moore, Ed. (D. Reidel , Boston, 1973) , pp 1-26. e. C. G. Overberger and J. C. Salamone, Ace. Chem. Res. 2, 217 (1969) . f. T. Shimidzu in "Advances in Polymer Science," Vol. 23, H.-J. Cantow, Ed. (Springer-Verlaa , Berlin, 1977) , pp 55-102. A. C. Satterthwait and W. P. Jencks , J, Amer. Chem. Soc. 96_, 7018 (1974) . W. P. Jencks, "Catalysis in Chemistry and Enzymology" (McGraw-Hill, New York, 1969), p 80. T. C. Bruice and S. J. Benkovic, J. Amer. Chem. Soc, 86, 418 (1964) . This pathway has been proposed for the reaction of imidazole and PNPA in toluene. F. M. Merger and A. C. Vitale, •/. Amer. Chem. Soc. , 9_5, 4931 (1975) . C. G. Overberger, St. Piere, N. Vorchheimer, J. Lee and S. Yaroslavsky, J. Amer. Chen. Soc. , 87_, 296 (1965) . C. G. Overberger and M. J. Morumoto, J. Amer. Chem. Soc. , 93_, 3222 (1971) . C. R. Dick and G. E. Ham, J. Macromol. Sci. Chem., A4 , 1301 (1970) . 90

PAGE 102

91 9. L. E. Davis, in "Water Soluble Resings," R. L. Davidson and M. L. Sittig, Eds. (Reinhold, New York, 1968) , p 219. 10. T. W. Johnson and I. M. Klotz , Maavomoleaules , 7, 618 (1974). 11. I. M. Klotz and V. H. Stryker, J. Amer. Chem, Soo. , 90, 2717 (1968); G. P. Royer and I. M. Klotz, J. Amer. Chem. Soo. , 91, 5885 (1969) . 12 I. M. Klotz, G. P. Royer and I. S. Searper, Proo. Eat. Aoad. of Sol. USA, 6_8, 263 (1971). 13. R. A. Pranis and I. M. Klotz, Biopolymers , 16, 299 (1977) . ' — 14. M. L. Bender, F. J. Kezdy and F. C. Wedler, J. Chem. Ed. , 44, 84 (1967) . 15. K. Satake, T. Okuyama , M. Ohashi and T. Shinoda, J. Bioohem. (Tokyo), 47_ 654 (1960). 16. C. F. Lane, Synthesis, 135 (1975). 17. K. A. Schellenberg, J. Org. Chem., 2_8, 3259 (1963). 18. J. Hine, Aoo. Chem. Res., 11, 1 (1978). 19. H. Morawetz, "Macromolecules in Solution" (Interscience, New York, 1965), p 350. 20. As compiled by W. J. Spentnagel and I. M. Klotz, J. Polym. Soi. Polym. Chem. Ed., 15, 621 (1977). A plot of P K 3 vs for bezyated and non benzylated PEI shows that hydropholic substituents reduce the pK . The 3. , app authors further state that the polymer does not" protonate below =0.4. V. S. Pshezhetskii , G. A. Murtazaeva and V. A. Kabanov, Europ . Polym. J. , 10, 5 81 (19 74) . 21. M. L. Bender, Chem. Rev. , 6_0 (1960) . 22. W. P. Jencks and M. Gilchrest, J. Amer. Chem. Soo., 90, 2622 (1968) . — 23. Ref. 2, cf p 580. 24. Ref. 19, p 416. 25. R. L. Letsinger and T. J. Saveride, J. Amer. Chem. Soo., 84, 3122 (1962) . 26. Ref. 2, pp 79-35.

PAGE 103

92 27. T. C. Bruice and R. Lapinski , J. Amer. Chem. Soc, 30, 2265 (1958) . 28. L. F. Fieser and M. Fieser, "Reagents for Organic Synthesis," Vol. 1 (John Wiley, New York, 1967), pp 493, 985. 29. D. J. McCaldin, Chem. Rev., 60_, 39 (1960). 30. Imidazole containing PEI has been found to transfer aryl groups to H 2 in the hydrolysis of nitrocatechol sulfate. H. C. Kiefer, W. J. Congdon, I. S. Scarpa and I. M. Klotz, Proa. Nat. Acad. Sci. USA. 69, 2155 (1972). 31. T. Shimidzu, H. Chiba, K. Yamazaki and T. Minato, Macromolecules , 9, 641 (1976). 32. A. Vogel, "Textbook of Practical Organic Chemistry," 4th Edition (Longmans, London, 1978), p 127. 33. E. J. Gabbay, K. Sanford, C. S. Baxter and L. Kapicak, Bioohem. , 12, 4022 (1973) . 34. R. A. Turner, C. F. Huebney and C. H. Scholz, J. Amer. Chem. Sec. , 11, 2801 (1949) . 35. R. Pineau, J. reoherches centre natl . recherche sci,,, Labs. Bellevue (Parris), 292 (1951), CA 46:417a. 36. P. Reynaud, T. Tupin and R. Delaby, Bull. Soc, Chem., France, 718-24 (1957). 37. M. L. Bender and K. Nakamura, J. Amer. Chem. Soc, 84, 2577 (1962) . 38. L. E. Romsted and E. H. Cordes, J. Amer. Chem. Soc, 90, 4404 (1968) . 39. D. J. Pasto and C. R. Johnson, "Organic Structure Determination" (Prentice-Hall Inc., Englewood Cliffs, New Jersey, 1969) . 40. Ref. 32, p 1084. 41. M. Yamaghichi and T. Yochida, J. Pham. Soc. , Japan, 7_4, 1075 (1954), as referenced in 29. 42. A. Albert and E. P. Serjeant, "The Determination of Ionization Constants," 2nd Edition (Chapman and Hall, New York, 1971), Chapter II. 43. W. P. Jencks and J. Carruilo, -J. Amer. Chem. Soc, 82, 1778 (1960) .

PAGE 104

93 44. W. P. Jencks and M. Gilchrist, J. Amer. Chem. Soa., 90, 2622 (1968) . — 45. T. Kunitake and S. Shinkai, J. Amer. Chem. So?., 93, 4256 (1971) . — 46. H. Morawetz, "Macromolecules in Solution" (Interscience, New York, 1965), pp 350, 354.

PAGE 105

BIOGRAPHICAL SKETCH Curtis Stanley Lege was born on Devember 27, 1946, in Princeton, Indiana. His family moved to Orlando, Florida, in 194 7. He graduated from Maynard Evans High School in 1964. Mr. Lege enlisted in the United States Air Force and received an Honorable Discharge in December 1969. Shortly after return to civilian life he married the former Joe Ann Purdue of Defuniak Springs, Florida, on February 21, 1970. He received an Associate of Arts degree from Valencia Community College in June of 1972. Later he received his Bachelor of Science degree from the University of West Florida in June 1974. He entered the doctoral program at the University of Florida in the fall of 1974 where he pursued a degree in organic chemistry. During this time the Lege family's first child, Spring Ann, was born on July 15, 1979. Upon graduation, he will begin work as a Research Chemist at the Westvaco Corporation's Charleston Research Center. 94

PAGE 106

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. James A. Deyrup, Chairman Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree Doctor of Philosophy. Merle A. Battiste Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree Doctor of Philosophy. George B. Butler Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree Doctor of Philosophy. Ben M. Dunn Professor of Biochemistry

PAGE 107

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Robert C. Stoufe/l Associate Professor of Chemistry This dissertation was submitted to the Graduate Faculty of the Department of Chemistry in the College of Liberal Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December 1979 Dean, Graduate School

PAGE 108

Hilpi