1 ¢

div-(.1

, . . ‘ o....r.oi«:..v .c
. , ..‘,c.flo t

 

.. .1. A
a... :0... vol!

1

09"
so. a 1:}. :9
'!II V:

9
.ltr}. .
. 9..) . {to
If e 3 I: u . . . to:
.525— ...2’0. 2"): 04 . ((1rd (Cf:
7:3. ,1 I 4...! L» r l
.i 5245'.- 1
pl:

.3,
p $.19.—

.l“ or.
. (rival to"!
low:

. l».
1.1....

 

3);! 3.. {I
a!a../..;o$t...
. . I. u. .
t. p
.. .
.

40a , ~..p.-.Ipo
ofci. (Ii-afar.aalo,l.
. lit-(f.¢<|:¢.rvv(. a)ol
. :Ilr . J.
‘ 41f.

 

r. :11). 7 :135? 5.
n;ta;t;t.r4{.. ..;,..:. r23
..l:.v.l.v. (,Jl:.!.!.i.
V V: arflnvavf Avis: .
.thylpr... I u . .
. ,4?! at. . . 4
‘ 41. . .

fa...
(Irv...

. :viclt

I )cl'.f!

. .r..,
.11 «11..
.1...»
a. I :47

TI.
.25!

'.?...:rc¢
.. :19:
v u.

 

it.
.. l.‘t¢..¥p.
V (to Rial: VI 3) 2.1
, '1': .
.
.

.«z......l. I
. .‘OI‘!tzf.4ol.:fr.(.
, ‘ V ‘ 1:414...
A :1
.

I. 1!.4'. 1;...DII. .
an l!l'!ol:illl.lc.
' I’llllvt

Illfi)
.r. I!
(vital! to
r' we 1 01.090tln‘..1rl
to! I:
.IAI¢XI.1 0:99.!
. , ‘

luioilovl .. .v
in]. ‘5. .v. ..

 

$th

"IflfiSSS

J llllllll\llll\lllllllllllllllllh

3 1293 00792 9098

This is to certify that the

thesis entitled
SYNTHESIS AND CHARACTERIZATION OF HYDROGEN
BONDED INTERPOLYMER COMPLEXES

presented by

Carl Lawrence Aronson

has been accepted towards fulfillment
of the requirements for

M.S. degree in Chem. Engr.

 

 

MM/ 3444;???

Major professor

Date u/n/qz.

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

_ _ ._.. ..._ _—...-

 

LIBRARY
lMlchlgen State
; Un‘lvorelty‘

?

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE ‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

J
‘ 7
‘JJ

 

 

 

 

 

 

 

Tm

MSU Is An Affirmative Action/Equal Opportunity Institution
cMcMunS-DJ

SYNTHESIS AND CHARACTERIZATION OF

HYDROGEN BONDED INTERPOLYMER COMPLEXES

By

Carl Lawrence Aronson

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Chemical Engineering

1992

ABSTRACT
Synthesis and Characterization of
Hydrogen Bonded Interpolymer Complexes
By

Carl Lawrence Aronson

Macromolecular complexes of poly(4-vinyl phenol) (PVPh) with poly(N,N-
dimcthylacrylamide) (PDMA) and poly(ethylene glycol) (PEG) with poly(methacrylic
acid) (PMAA) were synthesized and characterized. In the first system, PVPh is a proton
donor and PDMA is a proton acceptor and the two form relatively strong hydrogen

bonded complexes. The complex was isolated by precipitation from alcohol solutions.

The composition of the complex was determined, using high resolution 1H NMR
spectroscopy and elemental analysis, and was correlated with the homopolymer feed
ratio. The glass transition temperature of the complex was found to be substantially
higher than either of the two constituent homopolymers. A variety of solvents dissolved
both constituent homopolymers but would not dissolve the complex. However, strong
hydrogen bonding solvents, such as N,N-dimethylformamide, dimethyl sulfoxide and
pyridine, appear to break the complex upon dissolution. The complexation of PEG with
PMAA was studied in dilute aqueous solutions. Complexation was detected by NMR

relaxation experiments at lower PEG molecular weights than previously reported.

DEDICATION

to Cynthia Jean

My Wife and Inspiration

iii

ACKNOWLEDGMENTS

I would like to thank my advisor, Professor Dr. Alec Byron Scranton, for his
guidance, encouragement, and support. He was a constant source of knowledge and help

throughout this work.

I am grateful to the College of Engineering and the Department of Chemical
Engineering at Michigan State University for providing a fellowship for myself as well
as the fine facilities to carry out this work. I am grateful to the State of Michigan for
providing financial research support my in the form of grants through the Research
Excellence Funds at Michigan State University. I am also very grateful to the Michigan

Polymer Consortium for a graduate fellowship.

I am grateful to Hope College for the valuable knowledge that I gained in my
undergraduate years. I am also grateful for the research experience in chemistry that I

was afforded.

I wish to thank Lori A. Murphy, Kelley R. Copes, Cheryl R. Harned, Maryam
Hooshmand, Richard L. Bost and Krista J. Jenison who diligently worked in the lab as
undergraduate research assistants. 1 wish to thank the staff at the Max T. Rogers NMR
facility at Michigan State University for their encouragement and consultation. I
appreciate the help of Cynthia J. Aronson, my wife, and Dr. Jay A. Siege] of the
Department of Criminal Justice at Michigan State University for their help with
solubility, refractive index and FT-IR measurements. I appreciate the help of my wife,
Cynthia J. Aronson, in putting together this thesis. I also wish to thank Dr. Jay A. Siegel

for the use of his lab and equipment. I would like to thank Professors Eric A. Grulke and

V

R. Mark Worden of the Department of Chemical Engineering at Michigan State
University, for their encouragement. guidance and stimulating discussions about my

research and graduate education.

My sincere thanks are given to the many friends, old and new, I have at Michigan
State University and in East Lansing, Michigan. My appreciation goes to the members
of Dr. Scrantons’s laboratory including graduate students Scott Coons and Eric Nelson
for their input into my research. Special appreciation goes to Arvind Mohan Mathur, my
office mate and friend. He has been a great support to me during these past two years. I
surely enjoyed the discussions concerning science, engineering, research, religion and

marriage.

I wish to thank my wife Cynthia Jean (Schutt) for marrying me, putting up with
my idiosyncrasies and loving me so much. I wish to acknowledge my family, who have
helped much more than words can express. To them I give my deepest love and
gratitude. I wish to thank my parents-in-law for giving me their daughter, a great set of
extended relatives; and especially for their love, support and generosity. I particularly
wish to thank my father and mother, Dr. Lawrence D. Aronson and Mrs. Joan Kelly for
being my friends as well as my teachers. Thank you mom and dad for giving me the
desire to learn, the discipline to accomplish a task and the wisdom to know that God is in

charge of my life.

I wish to thank the Lord for giving me the mind, health and perseverance to finish

this work. Only He is able to keep us from all evil.

TABLE OF CONTENTS

Page
LIST OF TABLES ix
LIST OF FIGURES xi
CHAPTER 1 INTRODUCTION 1
CHAPTER 2 BACKGROUND 5
2.1 Polymer Complexation 5
2.1.1 Complexes with Poly(4-Vinyl Phenol) 6
2.1.2 Complexes of Poly(ethylene glycol) (PEG) with
Poly(methacrylic acid) (PMAA) 8
2.1.3 Applications of Poly(ethylene glycol) (PEG)/
Poly(methacrylic acid) (PMAA) Complexes 1]
2.1.4 Theoretical Models of Complexation 12
CHAPTER 3 OBJECTIVE 14

CHAPTER 4 COMPLEXES OF POLY(4-VINYL PHENOL) (PVPh)

WITH POLY(N,N-DIMETHYLACRYLAMIDE) (PDMA) 16

vi

vii

Page

4.1 Synthesis of the PVPh/PDMA Complexes 16
4.2 Characterization of the PVPh/PDMA Complexes 19
4.2.1 Elemental Analysis Studies 19
4.2.2 Nuclear Magnetic Resonance Studies 19
4.2.3 Refractive Index Studies 24
4.2.4 Fourier Transform-Infrared Studies 24
4.2.5 Solubility Studies 25
4.2.6 Characterization of Thermal Properties 25
4.3 Results and Discussion 26
4.3.1 Composition of the Complexes 26
4.3.2 Refractive Index of the Complexes 32
4.3.3 Fourier Transform—Infrared Spectra of the Complexes 37
4.3.4 Thermal Properties of the Complexes 44
4.3.5 Solubility of Complexes 47

CHAPTER 5 NMR RELAXATION STUDIES OF

POLY(ETHYLENE GLYCOL)/POLY(METHACRYLIC ACID) COMPLEXES 51

viii

5.1 Experimental Analysis of Hydrogen Bonded Complexes

5.2 NMR Sample Preparation

5.2.1 Synthesis and NMR Sample Preparation of

PEG/PMAA Complexes

5.3 Spin Echo NMR Experiments

5.4 Results and Discussion

5.4.1 Spin-Spin Relaxation Times of PEG Ethylene Protons

5.4.2 Spin Echo Experiments on Mixtures on

Mixtures of Different Molecular Weight PEG

5.4.3 Spin-Spin Relaxation Times of PEG/PMAA Complexes

CHAPTER 6 CONCLUSIONS AND RECOMMENDATIONS

6.1 PVPh/PDMA Complexes

6.2 NMR Analysis of PEG/PMAA Complexes

REFERENCES

APPENDIX

Page

51

52

52

55

57

57

60

66

69

69

73

75

82

LIST OF TABLES

Table

1.

Composition of the PVPh/PDMA Complexes

Comparison of Compositional Results for the PVPh/PDMA

Complexes as determined by Elemental Analysis and 1H NMR
Refractive Index of the PVPh/PDMA Complexes

using the Sodium C line

Refractive Index of the PVPh/PDMA Complexes

using the Sodium D line

Refractive Index of the PVPh/PDMA Complexes

using the sodium F line

Thermal Properties of the PVPH/PDMA Complexes

. Solubility of the PVPh/PDMA Complex, PVPh and PDMA

Enthalpies of Hydrogen Bond Formation Calculated from

Drago Parameters

NMR Spin-Spin Relaxation Times of the PEG Ethylene Protons

ix

Page

28

29

34

35

36

45

47

50

59

Table Page

10. Spin Echo Measurements for Mixtures of

Different Molecular Weight PEG 65

11. T2 Values for Complexed and Uncomplexed PEG 67

LIST OF FIGURES

Figure Page
1. Schematic representation of polymer complexation 2

2. Structure of Poly(4-vinyl phenol) (PVPh) 17
3. Structure of Poly(N,N-dimethylacrylamide) (PDMA) 18

4. High Resolution 1H NMR Spectrum of Poly(4-vinyl phenol)

(PVPh) in DMSO-d6 at 40°C 21
5. High Resolution 1H NMR Spectrum of Poly(N,N-dimethylacrylamide)

(PDMA) in DMSO-dg at 40°C 22
6. High Resolution 'H NMR Spectrum of PVPh/PDMA complex

in DMSO-d6 at 40°C 23

7 Composition of the PVPh/PDMA complexes as determined by

elemental analysis for nitrogen 31
8. Infrared spectra of PVPh at room temperature (20°C) 38
9. Infrared Spectra of PVPh at 60°C 39

xi

xii

Figure

10. Infrared spectra of PDMA at room temperature (20°C)

11.

12.

13.

14.

15.

16.

17.

Infrared spectra of PVPh/PDMA complex

(53 mol% VPh repeating unit in complex) at

room temperature (20°C)
Glass transition temperature of the PVPh/PDMA complex
over a composition range
Structure of poly(ethylene glycol) (PEG)

Structure of poly(methacrylic acid) (PMAA)
High resolution 1H NMR Spectra of
poly(ethylene glycol) (PEG) in D20 at 25°C

Ethylene protons of PEG (1H NMR) as a function

of total time for delay using spin echo pulse sequence

Ethylene protons of a mixture of 62 and 5,000,000

molecular weight PEG (1H NMR) as a function of

total time for delay using spin echo pulse sequence

Page

43

46

53

54

56

58

62

xiii

Figure Page

18. Data of composite integral versus total time for

delay for the spin echo pulse sequence on

mixtures of 62 and 5,000,000 molecular weight PEG 63

19. Graphical representation of equation 1 with

contribution from each molecular weight polymer in the mixture 64

CHAPTER I

INTRODUCTION

Polymer complexes are formed by the association of two or more complementary
polymers, and may arise from electrostatic forces, hydrophobic interactions, hydrogen
bonding, van der Waals forces or combinations of these interactions (1-4). Due to the
long-chain structure of the polymers, once one pair of complementary repeating units
associate to form a segmental complex, many other units may readily associate without a
significant loss of translational degrees of freedom. Therefore the complexation process
is cooperative, and stable polymer complexes may form even if the segmental interaction
energy is relatively small (1,4). A schematic representation of polymer complexation is
shown in Figure I. The formation of complexes may also strongly affect the polymer
solubility, rheology, conductivity and turbidity of polymer solutions. Similarly, the
mechanical properties, permeability and electrical conductivity of the polymeric systems
may be greatly affected by complexation.

The use of macromolecular complexation for the development of novel polymeric
materials is an area of unexplored potential. The specificity and reversibility of polymer
complexes make them useful for providing some control of a materials structure and
properties. Because macromolecular complexes often exhibit chemical and physical
properties that are drastically different than those of the individual constituent polymers,
complexation may be used to modify the properties of polymeric materials (1-2,4).
However few examples have been reported in the literature. Only recently, hydrogen
bonded molecular complexes have been exploited for the formation of environmentally

sensitive materials (5,6), while ionic complexes have been used for development of

. Complex in Polymer Mixture

(A)

(B)

 

Figure 1: Schematic representation of polymer complexation. (A) Two different

polymer species uncomplexed. (B) Segmental interactions to form polymer complexes.

3

materials which will contract upon an increase in temperature. or the application of

electrical potential (3).

Compared to complexes between complementary polyelectrolytes, complexes
formed by hydrogen bonding between complementary polymeric Lewis acids and bases
(14) typically dissociate under a wider variety of conditions. Hydrogen bonding occurs
between a Lewis acid containing an electron deficient proton and a Lewis base
containing a lone pair of electrons. Hydrogen bonds are distinctly directional and
specific, and are more localized than any other type of weak intermolecular interaction
(7). In associating polymer systems the equilibrium between complexed (hydrogen
bonded) and uncomplexed polymers in solution may be highly sensitive to surrounding
conditions, such as pH, temperature, solvent composition and polymer concentration.
Molecular structural parameters such as molecular weight and polymer architecture may

also influence the complexation equilibrium.

Many interesting properties of macromolecular complexes arise from the
cooperative nature of the complexation process (1-2,4). Unlike small molecules,
macromolecules can accommodate many energetically favorable segmental interactions
without undergoing a large change in entropy. For this reason, stable macromolecular
complexes may form even if the segmental interaction energy is relatively small, and the
complex stability is highly dependent upon the polymer chain lengths. In fact, many

systems exhibit a critical chain length below which complexation is not detected (l-2,4).

We have investigated the use of macromolecular complexation for developing
materials which exhibit an enhanced glass transition temperature and increased solvent
resistance. We have used a model system of poly(4—vinyl phenol) (PVPh) with
poly(N,N-dimethylacrylamide) (PDMA). In this system, PVPh is a proton donor,
PDMA is a proton acceptor, and the two form relatively strong hydrogen bonded

complexes. The PVPh/PDMA complex was isolated by precipitation from solution with

4
yields of over 95 mass percent under optimum conditions. The PVPh /PDMA was

characterized in solution and in solid polymer blends. The thermal, optical and solution

characteristics of the complex were correlated with composition.

We have also investigated the role of macromolecular complexation in responsive
polymers whose physical properties depend on external conditions such as temperature
and pH. The model system used was hydrogen bonded complexes of poly(methacrylic
acid) (PMAA) and poly(ethylene glycol) (PEG). Here, PMAA is a Lewis acid or proton
donor, and PEG is the Lewis base or proton acceptor. The system was studied using
nuclear magnetic resonance (NMR) spectroscopy to investigate whether spin-spin
relaxation studies could be used to detect complex formation. The spin-spin NMR
relaxation times of dilute poly(ethylene glycol) (PEG) solutions were determined using
spin-echo pulse sequences. PEG samples with higher molecular weight had shorter spin-
spin NMR relaxation times. Spin echo experiments were also done on mixtures of two
PEG species of different molecular weight. Regression analysis with two decaying
exponentials successfully characterized the mass fraction of each constituent. The
complexation of PEG with poly(methacrylic acid) was studied in dilute aqueous
solutions. Complexation in dilute solution was detected by NMR relaxation time

experiments at lower PEG molecular weights than previously reported.

CHAPTER 2

BACKGROUND

2.1 Polymer Complexation

Polymer complexation, interassociation or adduct formation has received
considerable attention in recent years. For example, several authors have reported that
the miscibility of two polymers is drastically affected by complex formation (8-13). A
number of modified Flory-Huggins thermodynamic models which incorporate strong
interactions such as hydrogen bonding have been proposed (8-11). In general, the
entropy of mixing for two polymers is small, therefore the enthalpy of mixing must be
negative for the two polymers to be miscible. For this reason, the overriding
thermodynamic requirement for polymer compatibility is a large enthalpic interaction to
produce a negative free energy of mixing. Various types of secondary intermolecular
binding forces have been investigated for producing miscible blends, including hydrogen
bonding (14-20), acid-base interactions (21), charge transfer (22,23), and ion-dipole
interactions (24). Hydrogen bonded complexes are particularly attractive for this
purpose because a variety of organic constituents may participate in hydrogen bonding.
Hydrogen bonding occurs between a proton donor group containing an electron deficient
proton, and a proton acceptor group containing an electron lone pair (e.g. O, N and S).
Hydrogen bonds are distinctly directional and specific, and are more localized than any
. other type of weak intermolecular interaction (7). Hydrogen bond energies are typically
in the range of 1-10 kcal/mol. Coleman et a1. (25) recently published a concise guide to
polymer miscibility enhanced by hydrogen bonding.

2.1.1 Complexes with Poly(4-vinyl phenol)

In the past few years, several authors have studied macromolecular complexes of
poly(4-vinyl phenol) (PVPh), also named poly(4-hydroxyl styrene) (PHOST), with a
variety of complementary polymers. Moskala and co-workers (26,27,28) used infrared
spectroscopy to investigate complexes of PVPh with poly(vinyl acetate), poly(lactones),
poly(vinyl alkyl ethers), poly(ethylene oxide) and poly(vinyl pyrrolidone). In all cases
the strength of the hydrogen bonding interaction directly correlated with the polymer-
polymer miscibility. Meftahi and Frechet (29) investigated the effect of complex
formation on polymer miscibility by studying the compatibility of poly(vinyl pyridine)
with copolymers of PVPh. They observed that a critical fraction of 4-vinyl phenol (VPh)
units copolymerized with styrene was necessary to achieve complete miscibility with

poly(vinyl pyridine). Qin et al. (30,31) investigated amorphous blends of PVPh with

poly(vinyl methyl ketone) and poly(ethylene glycol) using solid state 13C nuclear
magnetic resonance (NMR) and Fourier transform infrared spectroscopy (FT-IR). Again
physical properties such as miscibility, glass transition and melting temperature varied
directly with the degree of complexation. Zhang et a1. (32) have recently studied the

composition dependence and phase structure of blends between poly(4-vinyl phenol) and

poly(ethylene oxide) or poly(ethylene glycol). Carbon-13 (13C) resonance and proton
spin diffusion NMR results gave evidence for intermolecular hydrogen bonding phase
structure as well as domain size. Ting et a1. (33) have investigated hydrogen bonding
between copolymers of styrene and (4-vinyl phenol) with poly(ethylene oxide). An
enhanced glass transition temperature as well as infrared resonance effects gave
evidence of hydrogen bonded interaction equilibrium attributed primarily to the acidity

of the phenol group (33).

7
Other studies of PVPh complexes with complementary polymers have focussed

on polymer-polymer miscibility and glass transition temperatures. Serman et al. (34,35)
recently studied the hydrogen bonded complexation of PVPh with polyethers and
poly(n-alkyl methacrylate). These authors constructed spinodal polymer-polymer phase
diagrams and established association equilibrium constants using FT-IR. Jong et a1. (36)
reported limited miscibility in blends of poly(styrene-co-vinyl phenol) with poly(n-butyl
methacrylate) based on the amount of 4-vinyl phenol (VPh) monomer units in the
copolymer. Results of 13C solid state NMR, DSC, FT-IR and cloud point experiments
showed the domain size for homogeneity is also proportional to the number of VPh
monomer units. French and Machado (37-39) have investigated blends of poly(styrene-
co-vinyl phenol) with polyacetal. In the initial study of low molecular weight analogs,
hydrogen bonding was detected between 4-ethyl phenol and dirnethoxymethane in the
infrared spectrum (37). These authors obtained an interaction parameter by melting
point depression (38) that agreed well with interaction parameters obtained for the low
molecular weight analogs (39). The polymer blends exhibited a single glass transition
temperature over a wide composition range indicating miscibility (39). In addition, the
degree of crystallinity in the polyacetal phase decreased upon blending with the
poly(styrene-co-vinyl phenol) random copolymer (39). Coleman et a1. (40) used infrared
spectroscopy and thermal analysis to determine miscibility maps between copolymer-
copolymer blends of poly(styrene-co-4-vinyl phenol) and poly(ethylene-co-
methacrylate). These experimentally determined miscibility maps compared favorably
with those theoretically calculated using the Flory-lattice model for hydrogen bonded

polymer interactions.

Landry and Teegarden (41) measured the heats of mixing of small molecular
weight analogs of PVPh and poly(N,N-dimethylacrylamide) (PDMA). These studies

revealed that the hydrogen bonding interaction in this system is strong compared to

8
typical systems. Based upon differential scanning calorimetry (DSC) and FT-IR studies,

these researchers established complete miscibility for the PVPh/PDMA blend over the
entire composition range. Wang et al. (42) examined the effect of PVPh molecular
weight in complexes with PDMA, and investigated a variety of solvents for synthesizing
the blend. Here, all the blends showed a single Tg which was enhanced relative to either
constituent polymer. Wang et al. (42) characterized the PVPh/PDMA complex
stoichiometry through FT-IR and elemental analysis and found an azeotrope at 46 mole

percent PVPh in the feed irrespective of synthesis solvent (42). Suzuki et al. (43,44)

characterized the PVPh/PDMA blend using solid state 13C NMR and observed a shift of

3 ppm in the phenolic carbon resonance peak due to hydrogen bonding.

2.1.2 Complexes of Poly(ethylene glycol) (PEG) with
Poly(methacrylic acid) (PMAA)

The properties of dilute aqueous solutions of PEG and PMAA or PAA have been
studied by a number of investigators using viscometry, turbidimetry and potentiometric
titration (1-4,45-49). These studies have shown that the complexed polymer fraction
depends strongly upon the molecular weight of the complementary polymers. In
experiments with PAA or PMAA of high molecular weight, a critical PEG molecular
weight was identified below which no complexes were observed. A critical molecular
weight of about 2000 was reported for the PEG/PMAA system, while a value of around
6000 was reported for the PEG/PAA system (46-48). Several investigators (1—4,45-49)
found that both atactic PAA and PMAA complexed with PEG in a 1:1 repeating unit
molar ratio while isotactic PAA formed complexes with a 2:3 carboxylatezether molar
ratio. Finally, all investigators found that the complex stability increased with decreasing

pH, although there are discrepancies in the value of the critical pH for complexation. For

9
example, based upon turbidimetry Ikawa (49) reported a critical pH for PMAA/PEG

complexation of 3.0 in contrast to a value of 5.6 obtained previously using viscometry and

potentiometry.

Several authors have investigated the hydrophobic stabilization of PEG/PMAA
complexes. The earliest evidence for hydrophobic stabilization in aqueous media arose
from the temperature dependence of the complex stability. PMAA/PEG complexes
increase in strength with increasing temperature (45,46), whereas those of PAA/PEG do
not change appreciably. These trends suggest that the PEG/PMAA complexes are
hydrophobically stabilized, perhaps due to the a-methyl group which is present on
PMAA. Further evidence of hydrophobic stabilization of PMAA/PEG complexes was
reported by Ikawa et al. (49). For PMAA systems the critical chain length of PEG
increased significantly when the solvent was changed from water to a water/methanol
mixture; while it remained relatively constant for the PAA/PEG system. Similarly, Osada
and collaborators (48,50) found the temperature dependence of complex stability to be
opposite in ethanol/water mixtures from that in pure water, due to disruption of
hydrophobic interactions in the former case. These results were supported by
measurements made by Papisov et al. (51) who found that complexation of PMAA and
PEG in water is endothermic with a positive entropy change.

Recently the complexation of PAA and PMAA with PEG were studied using
fluorescent spectroscopic techniques (52-58). Morawetz and coworkers (56,57) studied
PAA association with PEG using PAA labeled with dansyl chromophores. These
chromophores exhibit a fluorescent intensity change and spectral shift when moved from
a hydrophilic to a hydrophobic environment. Large peak intensity changes upon addition
of PEG to dansylated PAA revealed that the chromophore ends up in an environment
largely devoid of water. Frank and collaborators (52-55) used fluorescence spectroscopy
to investigate the complexation of pyrene end-labeled PEG with PMAA and PAA.

Examination of the pyrene excimer to monomer ratio of sparsely tagged PEG allowed

IO

intramolecular end-to-end contact to be characterized, while experiments with fully
tagged PEG provided information about both intra- and intermolecular contacts. These
studies confirmed previous observations that complexation is highly dependent upon
chain length, facilitated by low pH, reduced by neutralization of the acid and reduced by
the addition of methanol. However, in contrast to earlier results, complexes of the PEG
with PMAA and PAA were detected for PEG molecular weights as low as 1850 and 4200,
respectively (52-55).

Investigations of the complex stoichiometry by fluorescence spectroscopy yielded
mixed results. Based upon studies using the previously described chromophores,
Morawetz and collaborators (56,57) and Frank and collaborators (52-55) found the extent
of complexation in the PEG/PAA system increased as the number of acid moieties was
increased past a 1:1 molar ratio. In fact, for low molecular weight PAA, the formation of
intermolecular excimers plateaued at the PAA/PEG repeating unit ratio of around 3:1
(55). In contrast, Heyward and Ghiggino (58) used fluorescence polarization studies of
acenaphthylene labeled PAA to show that complexation is maximized for the 1:1 molar
ratio of repeating units.

The affinity of PMAA and PAA for polymeric Lewis bases can depend quite
strongly upon the structure of the base. For example, PMAA will bind more strongly with
poly(vinyl pyrrolidone) (PVP) than with PEG (l-4,59). In fact, PVP will displace PEG
of similar molecular weight from a complex with PMAA. Interestingly, PEG of much
larger molecular weight than the PVP can displace PVP from a complex with PMAA.
Similarly the Lewis acid poly(itaconic acid monomethyl ester) will form a hydrogen-
bonded complex with PVP, but not with PEG. Other polymeric Lewis bases with which
poly(carboxylic acids) will form complexes include poly(acrylamide),
poly(dirnethoxyethylene), poly(vinyl methyl ether), poly(vinylbenzo-18-crown-6) and
poly(vinyl alcohol).

11

2.1.3 Applications of Poly(ethylene glycol) (PEG) /

Poly(methacrylic acid) (PMAA) Complexes

Complexes of PEG and PMAA have been used to design novel polymeric systems.
For example, PEG complexes with PMAA have been used to form membranes with
controlled permeability (60). Addition of PEG to water-swollen PMAA membranes at
low pH gave rise to large reversible shrinkage of the membranes. The shrinkage
increased with PEG concentration and with temperature. PEG with a molecular weight
as low as 2000 gave rise to shrinkage (61), while addition of alcohols or base to shrunken
membranes broke the complexes and resulted in swelling. Osada and coworkers (62,63)
used these membrane systems as chemical valves with controllable permeability. The
porous PMAA membranes were fixed in frames so that when shrinkage and swelling were
induced, the pores would open and close, thereby regulating solute flux. A similar
concept was studied by Nishi and Kotaka (64). End-linked PEG gels were swollen with
acrylic acid, which was subsequently polymerized. The resulting interpenetrating
networks showed pH responsive swelling and permeability similar to that described
above. The flux of solute macromolecules was varied by changing the pH of the

surrounding medium.

Polymer complexation has been exploited in our laboratory for the design of
responsive polymeric materials which exhibit sharp transitions in swelling and
permeability in response to small changes in pH, temperature or solvent composition
(6,65). This approach is based upon the fact that material properties under complex-
promoting conditions may be dramatically different than those under complex-impeding

conditions.

12

2.1.4 Theoretical Models of Polymer Complexation

The complexation of complementary polymers has been modeled theoretically by
Kabanov and collaborators (4,47,66). In this analysis the total free energy of
complexation was divided into two contributions: one arising from the specific
interactions between complexing functional groups, and a second arising from
configurational changes of the system upon complexation. The authors typically assumed
that in the complexed state, the polymer of the shortest chain length (the PEG) was
completely bound with all repeating units participating in hydrogen bonds. This
assumption greatly simplified the calculation of the number of system configurations in
the complexed state. According to these models the extent of complexation can change
very abruptly near critical values of polymer molecular weight or free energy of
complexation due to cooperative effects in the complexation process. Calculations based
on the model also agreed at least qualitatively with other experimentally observed trends.
For example, the equilibrium bound fraction depended strongly upon chain length and

stable complexes formed even with weak segmental interactions.

Several authors have modeled the association of biological polymers (67-70). The
various models differ in the manner in which the polymer conformation in the complexed
state is considered. One approach incorporates the assumption that a complexed polymer
has only one possible conformation, that in which all repeating units are bound to the
complementary polymer. Examples include models for the complexation of poly- and
oligo-nucleotide (67,68). In contrast, models of the double-stranded helix complexes of
long-chain nucleic acids have included the possibility of loops in the conformation of a
complexed chain (69,70). In these models, which are reviewed by Poland and Scheraga
(70), it is typically assumed that the loops are fairly long (> 20 units) when calculating the

loop entropy.

13

Mathematical techniques based on sequence generating functions have been
proposed for efficient formulation of the statistical mechanical partition functions of
polymeric systems. For example, Lifson (71) reported a procedure for evaluating the
canonical partition function of long-chain polymers with repeating units existing in two
or more distinct states, while Eichinger et al. (72) reported a generating function
technique for the evaluation of the partition functions of a long polymer chain absorbed
onto a planar surface. Both of these methods were applicable in the limit of infinite chain
length. Scranton et al. (73) used a similar generating function technique to describe the
complexation thermodynamics of free and graft oligomers with complementary
polymers. Since the long chain assumption was not required, this analysis was applicable
to short and intermediate chain lengths.

In summary, polymeric acids may form complexes with a number of polymeric
Lewis bases (including PEG) in water. The extent of complexation increases with
polymer concentration, polymer molecular weight and with reductions in pH. The extent
of complexation may increase, decrease or remain essentially unchanged with changing
temperature depending upon the contribution of hydrophobic interactions to the complex
stability. Recent spectroscopic measurements suggest that complexes may take place at
lower molecular weights than previously determined using viscometry, turbidimetry or
titration. In addition, some complexes previously thought to form with a 1:1 repeating
unit stoichiometry may not do so in every case. Certain detection techniques may be more
sensitive to complexation than others, perhaps accounting for the discrepancies in critical

chain length and complex stoichiometries.

CHAPTER 3

OBJECTIVE

Although significant advances in the area of polymer blending via complexation
have been made, analysis of the literature reveals relatively few contributions utilizing
polymer complexation for the purpose of creating materials with enhanced solvent
resistance and an enhanced glass transition. Most researchers have concentrated their

efforts on enhancing miscibility through secondary interactions.

The work with the poly(4-viny1 phenol) (PVPh)/poly(N,N-dimethylacrylamide)
(PDMA) system addresses important fundamental aspects of interpolymer hydrogen
bonded complexation and establishes composition-property relationships for a model

hydrogen bonded interpolymer system.

The first broad objective of this work was to synthesize hydrogen bonded
complexes of poly(4-vinyl phenol) with poly(N,N-dimethylacrylamide) with varying

initial feed composition. In particular, the following specific objectives were established:

(i) to optimize the synthetic yield of the PVPh/PDMA complexes with respect to feed

composition, and solvent;

(ii) to characterize the composition (stoichiometry), degree of hydrogen bonding and
thermal properties of the PVPh/PDMA complex using spectroscopic, thermal, optical

and elemental analysis techniques;

(iii) to test the solubility of the PVPh/PDMA complexes in a wide variety of solvents in

order to investigate enhanced solvent resistance due to complexation as well as to

14

15
establish a relationship between the solubility of the polymer complex and the energy of

the hydrogen bonded interaction; and

(iv) to investigate any enhancement of the glass transition temperature (Tg) without a

change in the melting or degradation temperatures with respect to the uncomplexed

homopolymers.

The second broad objective of this work was to investigate whether nuclear
magnetic relaxation could be used as a tool in detecting hydrogen bonding. Here, the
system used is complexes of poly(ethylene glycol) with poly(methacrylic acid). More

specifically, the major objectives of this work were:

(i) to investigate the correlation between spin-spin relaxation time (T2) and molecular

weight of monodisperse poly(ethylene glycol) samples;

(ii) to investigate the determination of mass fraction of a single polymeric species in the

midst of a mixture using nuclear magnetic relaxation spin echo techniques; and

(iii) to investigate the detection of a critical molecular weight for complexation through

nuclear magnetic relaxation spin echo experiments.

CHAPTER 4
COMPLEXATION OF POLY(4-VINYL PHENOL) (PVPh) WITH

POLY(NN-DIMETHYLACRYLAMIDE) (PDMA)

4.1 Synthesis of the PVPh/PDMA Complexes

Poly(4-vinyl phenol) (PVPh) and poly(N,N-dimethylacrylamide) (PDMA) were
used as received from the supplier (Polysciences, Inc., Warrington, PA). The structures
of PVPh and PDMA are shown in Figures 2 and 3 respectively. The nominal molecular
weight of the PVPh as reported by the supplier was 30,000 g/mol. The molecular weight
of the PDMA was measured by viscometry. The intrinsic viscosity of the PDMA at

25°C in methanol was 1.33 dL/g. Based upon the Mark-Houwink parameters reported
for PDMA by Trossarelli and Meirone (74), the weight average molecular weight was

determined to be 510,000 g/mol.

For complex synthesis, separate solutions of PVPh and PDMA were prepared by
dissolving appropriate amounts of the polymers in a solvent Acetone, methanol and
ethanol were investigated with polymer concentrations of l or 3 weight percent.
Precipitate formed immediately upon mixing solutions of the component polymers at

room temperature. The opaque complex precipitate was allowed to settle for 1 hour and
then the supernatant was decanted off. The complex precipitate was placed in a 65°C

oven for one week in order to evaporate off all residual solvent. After a week the red-

orange complex was weighed to determine the yield of the complex.

16

17

/\/\/\/\/\

H

I—O
12—0
0
.

Figure 2: Structure of Poly(4-vinyl phenol) (PVPh).

18

HSC‘ .. ’CHS Hac‘ .. [CH3 H30\ .N. ’CHS H30: .ICHa
i' i‘ I 'i
le=0 ?=0 $=0 ('3— =0

\/°”\/”\/°“\/°”\/

Figure 3: Structure of Poly(N,N-dimethylacrylamide) (PDMA).

19

4.2 Characterization of the PVPh/PDMA Complexes

4.2.1 Elemental Analysis Studies

The composition of the complex was determined by elemental analysis.
Elemental analysis was performed on a CHN analyzer. The mass percents of hydrogen,

carbon and nitrogen were determined for each sample.

4.2.2 Nuclear Magnetic Resonance Studies

Solution state 1H NMR spectroscopy was performed using a VXR-300
spectrometer (Varian, Palo Alto, CA). Deuterated dimethyl sulfoxide (DMSO—d6,

99.96% D, Isotec, Inc., Miamisburg, OH) was used as the solvent in all the experiments.

NMR samples had concentrations between 0.5 and 0.8 weight percent polymer. All

experiments were performed with a controlled temperature of 40°C, using a transmitter
frequency of 299.949 MHz. Experiments were conducted at an elevated temperature in
order to reduce the viscosity of the sample and obtain narrower linewidths. The 180
degree pulse was measured for each sample and varied between 34.4 and 35.2
microseconds. The delay between successive pulses was at least 30 seconds. At least
200 transients were collected before Fourier transformation. Peak assignments for the
proton spectrum of PVPh are as follows (Figure 4): ethylene protons showed in a broad
peak centered at 1.4 ppm; aromatic protons were displayed by a peak centered at 6.5
ppm and the hydroxy hydrogen was observed at approximately 9 ppm. For PDMA

(Figure 5), the peak for the backbone ethylene protons came at 1.5 ppm and the peaks

20
corresponding to the 6 methyl protons for the two CH3 groups attached to the nitrogen
were centered at 2.8 ppm. It should be noted that in the 1H NMR spectrum of PDMA,

the peak corresponding to the 6 methyl protons, of the two CH3 groups attached to the

nitrogen, are likely split because of hindered rotation around the N-CO bond due to the
presence of the carbonyl group. Rotation around the N-CO bond is likely slower than
the NMR timescale. In the spectrum of the PVPh/PDMA complex (Figure 6), the peak
due to the methyl groups attached to the nitrogen for PDMA and the peak due to the
aromatic protons meta and ortho to the hydroxyl group on the aromatic ring of PVPh
were integrated to determine composition (Figure 5). The hydroxyl peak of the PVPh

appeared at the same location for the complex and the pure PVPh.

21

 

 

 

 

 

 

/\©/\/\/\/\
5’
H H
A A JL-.JM____
I I I1 'rT""I"'I III III II I W
9 7 6 5 4 3 1 DD.

Figure 4: High resolution 1H NMR spectrum of poly(4-vinyl phenol) (PVPh) in DMSO-

d6 at 40°C.

22

.0 COOC O‘CO'C
“30.“.th HJC.N.CH3 “a .N. "a H: N H:

i-°i
\/

-o (Ar-o

.0 I
V” \/

 

 

 

 

Figure 5: High resolution 1H NMR spectrum of poly(N,N-dimethylacrylamide) (PDMA)

in mason6 at 40°C.

23

 

 

 

 

 

13 12 it 10 9 8 7 6 5 4‘ 3 2 1 .-0 pp-

Figure 6: High resolution 1H NMR spectrum of PVPh/PDMA complex in DMSO-d6 at

40°C.

24
4.2.3 Refractive Index Studies

Refractive index (ni) measurements were performed using the Becke line method.

The sample and refractive index standard liquid were heated using a Mettler—FP52 hot
stage and the Becke line was observed to fade at a specific temperature. Using light

filters the refractive indices relative to the sodium C (lmax=651.2 nm), sodium D
(Max:5880 nm) and sodium F (Max:4863 nm) lines were accurately measured. The

standard deviation for the refractive index measurements was less than 0.001. Blends of

PVPh with PDMA in methanol were prepared by film casting for refractive index
calibration. The solutions were dried for 7 days at 70°C to remove most of the solvent.

The samples were then placed in a vacuum oven at 70°C under a vacuum of 28 inches of
mercury for 12 hours to drive off residual solvent. The refractive indices of each sample

were measured relative to the sodium C, D, and F lines.

4.2.4 Fourier Transform-Infrared Studies

Samples for infrared analysis were prepared by mixing the dried complex with
potassium bromide (KBr) at a concentration less than 1 weight percent. The mixture was

pressed into pellets under pressure. Infrared spectra were acquired on a Nicolet 5DXB

Fourier Transform infrared spectrometer in transmittance at a resolution of 4 cm‘l. The

frequency scale was internally calibrated with a reference helium-neon laser to an

1. At least 200 scans were signal averaged and the spectra were

accuracy of 0.2 cm'
stored on a disk system. A Beckman-RIIC TEM 1C automatic temperature controlled
cell mounted in the spectrometer was used to obtain elevated temperature spectra. The

temperature of the cell was checked by an external temperature sensor and was accurate

to within +/-0.5°C.

25
4.2.5 Solubility Studies

The solubility of the complex was investigated by mixing equal amounts of
successively dilute homopolymer solutions of 1.0, 0.1 and 0.01 mass percent polymer in
methanol, and then observing for precipitate formation. The solubility of the complex as
compared to the individual homopolymers was also determined in a wide range of
solvents. The complex or one of the homopolymers was placed in a solvent at polymer
concentrations between 0.1 and 0.5 mass percent polymer. The resulting solutions were

vortex mixed and observed at room temperature for at least 14 days for dissolution of the

polymer.

4.2.6 Characterization of Thermal Properties

The thermal properties of the complexes were studied using differential scanning

calorimetry (DSC) and thermogravimetric analysis (TGA). DSC and TGA were

performed on a DuPont 9900 instrument, with a heating rate 10°C/minute in both cases.

The sample size was between 10 and 20 mg for both DSC and TGA. For the DSC

experiments, two runs were conducted on each sample. The initial run to 250°C was
performed to drive off any residual solvent. The glass transition temperature (Tg) was
taken as the inflection point in the change in heat capacity with temperature for the
second run. All complexes exhibited a single Tg. In the TGA experiments, the
temperature for the onset of degradation as well as the temperature of 25.00% weight
loss were measured for each sample heated under nitrogen purge. Both the DSC and

TGA measurements were repeated at least twice on samples from identical batches and

the variance between samples was less than 10°C for both the Tg and the 25.00%

degradation temperatures.

26

4.3 Results and Discussion

The complexes were synthesized by precipitation from solutions containing 1 or 3
weight percent of each of the constituent polymers. Alcohols such as methanol and
ethanol are particularly useful for this synthesis method because, although both
homopolymers are soluble in these liquids, the complex is not. Therefore, the complex

may be isolated in a purified form containing no uncomplexed homopolymers.

Ethanol and methanol were chosen as the optimum solvents for synthesizing the
PVPh/PDMA complexes. For synthesis runs containing 1:1 4-vinyl phenol (VPh)/N,N-
dimethylacrylarnide (DMA) repeating unit mole ratios, the mass percent yield of the
complex was consistently over 95% of the total polymer in the feed for both ethanol and
methanol. These alcohols were chosen as synthesis solvents largely because of their
volatility, allowing them to be evaporated off easily. Precipitation of the complex from

methanol was observed at concentrations as low as 0.01 weight percent polymer.

4.3.1 Composition of Complexes

Table 1 contains data for the composition of PVPh/PDMA complexes as a
function of the corresponding feed composition. These complexes were synthesized by
mixing solutions containing 1 weight percent of each homopolymer in ethanol. The
values in the first column of the table report the percentage of the polymer repeating
units in the synthesis mixture (both complexed and uncomplexed) which are vinyl
phenol (VPh) (the solvent is ignored for this calculation). The second column reports the
mole percentage of polymer repeating units in the complex which are vinyl phenol, as

determined by elemental analysis for nitrogen. The values in this column were checked

27

using high resolution, solution state 1H NMR and the agreement between the two
methods was very good above 30 mol% VPh in the complex as shown in Table 2. Below
30 mol% VPh in the complex, the two methods do not agree well with each other. This
is probably due to the insensitivity of the NMR integration for relatively broad peaks
with low intensity. Therefore, the elemental analysis data are probably more trustworthy

at these concentrations.

28

Table 1: Composition of the PVPh/PDMA Complexes

 

 

 

 

 

 

 

 

 

 

 

Overall mol% VPh Repeating mol% VPh Repeating Units in
Units Complex, Elemental Analysis
3.39 16.8

7.33 30.9

11.9 31.8

17.4 39.6

32.2 43.2

42.5 44.7

55.9 45.4

74.0 53.4

 

 

 

29

Table 2: Comparison of Compositional Results for the PVPh/PDMA Complexes as
determined by Elemental Analysis and 1H NMR

 

 

 

 

 

 

 

 

Overall mol% VPh mol% VPh Repeating mol% VPh Repeating
Repeating Units Units in Complex, Units in Complex, NMR
Elemental Analysis

3.39 16.8 2.30

7.33 30.9 14.1

11.9 31.8 29.1

17.4 39.6 39.6

32.2 43.2 43.0

42.5 44.7 45.0

55.9 45.4 48.5

74.0 53.4 52.6

 

 

 

 

 

30

Data from Table 2 are plotted in Figure 7. The diagonal line in the figure
corresponds to compositional azeotropes which would occur if the composition of the
complex were the same as that of the feed. The data appears to intersect the diagonal at
about 45 mol% VPh, suggesting that the system exhibits an azeotrope at this
composition. It is clear from Figure 6 that below the azeotrope, the complex is enriched
in VPh relative to the feed. As the mol% VPh in the feed is increased to 74 mol%, the
percentage of VPh in the complex approaches a plateau at around 53 mol percent. These

trends are in qualitative agreement with those reported by Wang et al. (42).

31

 

 

10 0 . ' , . ' {- - . Compositional Azeotropel
o 10 20 30 40 50 60 7° 30

mol% VPh repeating unit In told polymer tee-d

Figure 7: Composition of the complexes as determined by elemental analysis for

nitrogen. The diagonal line corresponds to compositional azeotropes.

32

The reason for this trend arises in part from the molecular weight differences
between the PDMA and PVPh used in this study. The PDMA, which has a molecular
weight of about 510,000 g/mol, contains over 5,000 repeating units per chain; while the
PVPh (M.W. 30,000) contains only about 250 repeating units per chain. Therefore, each
PDMA chain is capable of accommodating twenty fully bound PVPh chains. It could
accommodate even more if the PVPh chains were partially bound with dangling ends. It
is unclear how many PVPh chains must be bound to a PDMA chain before it becomes
insoluble and precipitates. For a system which is lean in PVPh, the complex is enriched
in PVPh relative to the solution suggesting that a threshold number of bound chains may
be required for precipitation. As the VPh/DMA repeating unit ratio increases, the
complex stoichiometry approaches a value near 1:1 with the VPh mol% in the complex
plateauing at 53%. In this regime, the solution contains excess PVPh chains, and the
complex contains PDMA chains completely covered with PVPh. The plateau clearly
implies that once a PDMA chain is consumed in the complex (fully covered with PVPh),
it cannot accommodate more PVPh. Therefore an increase in the PVPh concentration in

the solution does not lead to an increase in the PVPh content of the complex.

4.3.2 Refractive Index of Complexes

The refractive index measurements on the polymer blends prepared by film
casting indicated that the refractive index of the blends varied linearly with the molar
composition. Therefore, refractive index measurements were proposed to provide an
alternative method for determining the composition of the complexes. Data for the
refractive indices of the complex are shown in Table 3, 4 and 5 for the sodium C, D and

F lines respectively. The first column of each table represents the mole fraction of VPh

33

repeating units in the complexes as determined by 1H NMR. The second column in each
table represents the experimentally determined refractive index of the complex, while
the third column shows the value calculated as the pure component values multiplied by
the corresponding mole fractions. The excellent agreement between the experimental
and predicted values demonstrates that the refractive index varies linearly with the molar

composition and may be used to estimate the composition of the complexes.

34

Table 3: Refractive Index of the PVPh/PDMA Complexes using the Sodium C line

 

 

 

 

 

 

 

 

 

 

 

 

 

mol% VPh Repeating Experimental Refractive Calculated Refractive
Units in Complex Index Index
0: 1.513 1.513
2.30 1.514 1.515
14.1 1.522 1.527
29.1 1.532 1.543
41.9 1.544 1.555
43.0 1.551 1.557
45.0 1.568 1.559
48.5 1.557 1.562
52.6 1.561 1.566
100 1.615 1.615

 

 

 

 

35

Table 4: Refractive index of the PVPh/PDMA Complexes using the Sodium D line

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mol% VPh Repeating Experimental Refractive Calculated Refractive
Units in Complex Index Index
T:_=lfi 1.515
2.30 1.518 1.518
14.1 1.525 1.530
29.1 1.531 1.545
41.9 1.544 1.558
43.0 1.552 1.559
45.0 1.571 1.565
48.5 1.560 1.567
52.6 1.562 1.569
100 1.618 1.618

 

 

36

Table 5: Refractive Index of the PVPh/PDMA complex using the Sodium F line

 

 

 

 

 

 

 

 

 

 

 

 

 

mol% VPh Repeating Experimental Refractive Calculated Refractive
Units in Complex Index Index
0 1.523 1.523
2.30 1.523 1.525
14.1 1.533 1.538
29.1 1.539 1.555
41.9 1.558 1.568
43.0 1.563 1.570
45.0 1.579 1.575
48.5 1.573 1.577
52.6 1.574 1.580
100 1.632 1.632

 

 

 

 

37

4.3.3 Fourier fiansform-Infrared Spectra of the Complexes

The infrared (IR) spectra of PVPh, PDMA and the PVPh/PDMA complex were
measured at room temperature. The room temperature infrared spectra of PVPh as
shown in Figure 8 had several distinguishing features. The absorbances at 825, 1100,
1170, 1445, and 1602/1609 cm'1 were due to the aromatic ring of the pendant group.
The absorbance at around 870 cm'1 is due to the para substitution on the phenyl ring.
Absorbances from 1200-1400 cm'1 are the -OH deformation and C-0 stretching
vibrations mixed together to some degree. The absorbance band from 3800 to 3000cm'1
(maximum absorbance at 3381 cm'l) is due to ~OH stretching vibrations. The -OH

stretching region of the spectra (3800 to 3000 cm'l) of PVPh collected at 60°C is shown

in Figure 9. There is still not much resolution of the complexed and free -OH groups.

However, Moskala et a1. (26) reported that at elevated temperatures above 100°C, two

components contained in this band can be identified. In the -OH stretching region (3800

to 3000 cm'l), the relatively broad band can be attributed to hydrogen bonded -OH
groups. This broadening is due to the hydrogen bonds being associated in aggregates of
various sizes and shapes, producing a variety of different extents of interactions and
bond strengths. While the narrower band is due to free, non-hydrogen bonded hydroxyl
groups. Increasing the temperature causes the intensity of the absorbance due to free -
OH groups to increase relative to the hydrogen bonded absorbance (26). The IR spectral
assignments for PVPh presented for the data here are similar to those presented by

Moskala et al. (26) as well as Wang et al. (42).

38

/\/\/\/\/\

6352

‘P 1’

H H

0

C

0 PVFH I. Apr 0. 00:00:47
0

C

0

l

01>

Aboorlone.

9.000; 9.rno4 9.3.07 o.e41o

 

 

 

 

 

 

 

-0.000I

4 A A A A L k 4L A A g A

7 T V Y T’ T Y V f ‘7 V T V
0000.0 3000.0 8000.0 3000.0 0400.0 8000.0 {000.0 1000.0 1400.0 3000.0 1000.0 000.00 000.00 400.30
novonu-oor (ca-11

 

Figure 8: Infrared spectra of PVPh at room temperature (20°C).

39

/\g)/\E>p/\/\/\
i: i

00Lu4.vl'ut PrtI’NOL) At out. .1: Sq. 94 15 20 52

‘ w“ M‘ “A“ W~~ " M

’3
“I
(h
r'\

J ‘4 4;

U i———4 4 4 , i ‘ i 3 v
.H'It i .1844 7 .1004 7 .1544 7 3394 7 3144 7 .5094 7 2944 7 2794 7 2644 7

Novena-met (cm-I)

Figure 9: Infrared Spectra of PVPh at 60°C.

40

The IR spectra at room temperature of PDMA as shown in Figure 10 also showed

1, with a

characteristic group absorbances. The absorbances between 1759 and 1547 cm"
maximum at 1639, are due to carbonyl group (C=O) stretching. Yang et al. (14) have

shown that at elevated temperatures it is readily apparent that this C=O band contains
two components. The peak centered at 1639 cm’1 is due to free, non-hydrogen bonded

C=O groups and the peak centered at 1624 cm'1 is due to hydrogen bonded C=O groups.
The spectral assignments here for PDMA are similar to those given by Yang et al. (14).
Yang et al. (14) have shown that at elevated temperatures it is readily apparent that this

=0 band can be quantitatively resolved into both free and hydrogen bonded C=O

components.

41

me. Nets, n.6,- N.Ci-t; H16: -.CHa HaC: N.Cfl:

Juuuv

I. A.' 00 II I, 00

i

l.0l00 0.000! 0 7000

000.000“...
l.‘00’
.__.._...--- ~..+—___ __ .__._..._.‘.___ ._...+_ __~__‘

0.1070

—+——Q——-——.— é #— fi= 4' :7 4V t ; 4; A
1000.0 0000.0 0000.0 .000 0 0400.0 0000.0 |.00.0 I000». Il00 0 1.00.0 I000.0 000.00 000.00 000.00
'.V.~U‘.' 40.-..

Figure 10: Infrared spectra of PDMA at room temperature (20°C).

42

The room temperature IR spectrum of the PVPh/PDMA complexes contained

components from both constituent homopolymers as well as features due to interpolymer
complexation. The absorbance peak at 1630 cm'1 in Figure 11 is due to the C=O moiety

of PDMA. The absorbance from 3800 to 3000 cm"1 is due to the para substituted -OH
moiety of PVPh. According to Landry and Teegarden (41), absorbances at 1612 and

1592 cm'1 correspond to ring vibrations in PVPh. At room temperature the free
hydrogen bonded components of the C=O and -OH absorbances were not able to be
resolved. The spectral assignments for the PVPh/PDMA complex given here are similar
to those presented by Landry and Teegarden (41). These authors have shown that at
elevated temperatures, IR absorbances due to hydrogen bonding were visible (41).
These authors show that the absorbance of the non-hydrogen bonded (free) -OH as well

as both the peak due to the (C=O)free fraction and the absorbance of the (C=O)hydrogen
bonded fraction were able to be resolved (41). This is due to the decrease in hydrogen

bonding at higher temperatures and hence a narrowing of the corresponding peak.
Landry and Teegarden also have shown a change in the phenol intensity and

(C=O)hydrogen bonded fraction with a change in PVPh concentration in the complex (41).

Absorptions due to hydrogen bonding stretching are moved to longer wavelengths
(lower frequencies) accompanied by increased intensity and band broadening. Landry
and Teegarden used the linear relationship between the change in absorbance frequency
upon hydrogen bonding and the enthalpy of interaction to obtain an experimental

enthalpy of mixing for the PVPh/PDMA complex (41).

43

0
‘ IAMIO00 'V’fl. tone '00-! 010” I0 00" 00 II‘ 00. 00
i

 

 

 

 

r - -+—- —-+ r r :
ooe gnome rooo.e 000.00 000.00 000...

Figure 11: Infrared spectra of PVPh/PDMA complex (53 mol% VPh repeating unit in

the complex) at room temperature (20°C).

4.3.4 Thermal Properties of Complexes

Table 6 shows the glass transition and degradation temperatures of the complex
over a range of compositions as determined by elemental analysis for nitrogen. Each
complex exhibited a single glass transition, indicating that the complexes are

homogeneous. The data from the first two columns of Table 6 are shown in Figure 12.
The Tg value obtained for pure PDMA (108°C) is somewhat lower than that reported by

previous authors (118°C-122°C) (41,42,43,44). The Tg value obtained for PVPh is

156°C. The dotted line in Figure 11 represents a weighted average between the Tg
values for PDMA and PVPh. In general, the value of Tg increases as the mol% of 4-
vinyl phenol (VPh) increases from 0% to 45%. For systems containing more than 30
mol% VPh in the complex, the Tg of the complex is substantially above the weighted
average. In fact, for systems containing more than 40 mol% VPh in the complex, the Tg
of the complex is greater than that of either homopolymer. Similar trends have been
reported by previous investigators (41,42). Hydrogen bonding can lead to an enhanced
glass transition temperature because these interactions act as physical cross-links (42),
and additional energy must be provided to the hydrogen bonds for the individual
polymer chains to gain enough mobility to become rubbery. As illustrated in Table 6, the
degradation temperature steadily decreases as the PVPh content of the complex
increases. The degradation steadily decreases as the mol% VPh increases. We observed
no enhancement of the degradation temperature due to complexation. The TGA
experiments are void of the effects of hydrogen bonding since these experiments are

primarily sensitive to the breaking of covalent bonds.

Table 6: Thermal Properties of the PVPh/PDMA Complexes

45

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mol% VPh Degradation 25.00% Weight
Glass Transition

Repeating Units Onset Loss Temperature,

. Temperature, °C

in Complex Temperature, °C °C

0.00 108 421 431

16.8 108 404 430

30.9 147 392 412

3 1 .8 144 406 420

39.6 156 376 411

43.2 167 378 404

44.7 182 37 6 409

45.4 169 397 411

53.4 175 374 419

100.0 156 386 399

 

 

46

 

 

 

200T
180" ' .
.I
9,160" + -—=I
o
u I.
8‘40" WelghtedAveroge
120«— ’
+"' - I
roof , , . . . t e i . .
o 10 20 30 40 50 oo 70 80 90 100
mol% VPh In complex

Figure 12: Glass transition temperature of the PVPh/PDMA complex over a

composition range.

47

4.3.5 Solubility of Complexes

Table 7 contains data for the solubility of the complex and the constituent
homopolymers in a wide variety of solvents. These studies were performed using a
rather low concentration of 0.1 to 0.5 weight percent polymer in solvent to characterize
solubility. In the table a “+” denotes solubility while a “-” denotes insolubility. It is
interesting to note that eleven of the solvents investigated will dissolve both constituent
homopolymers but will not dissolve the complex. Complexation may drastically affect
polymer-solvent miscibility because in the complex, polar polymer functionalities are
interacting strongly with one another, and are unavailable to interact with the solvent. It
was hoped that nonpolar solvents would dissolve the complex while leaving the
hydrogen bonds intact. However, only three solvents were found which dissolved the
complex, and all three are strong hydrogen bonding solvents. Therefore, it is likely that

these solvents disrupted the hydrogen bonds and broke apart the complexes during

dissolution. The fact that the 1H NMR PVPh hydroxyl peak occurred in the same
location for the complex and the pure PVPh indicates that DMSO indeed disrupts the

polymer-polymer complex.

48

Table 7: Solubility of the PVPh/PDMA Complex, PVPh and PDMA

 

Solvent Complex PVPh PDMA

 

 

Methanol -

 

 

l-Propanol -

 

+
Ethanol - +
+
+

2-Propanol -

 

Water - _

 

HCl/HZO pH 4.0 - -

 

I-ICl/HZO pH 2.0 - -

 

1-1CllI-120 pH 1.5 - -

 

Acetone - +

 

++++++++++

Methyl ethyl ketone - +

 

Formaldehyde - -

+

 

Hexanes - - -

 

Cyclohexane - -

 

Fonnamide - +

 

Dimethylfonnamide + +

 

Tetrahydrofuran - +

 

++++

Dimethyl sulfoxide + +

 

Carbon tetrachloride - -

 

Chloroform - -

+

 

Methylene chloride .. - +

 

Benzene - _ ,

 

Toluene -

 

Phenol -

 

+ +

Pyridine +

 

Dioxane -

+

 

 

++++

Acetonitrile -

+

 

 

 

 

 

49
Drago et al. (75) have developed a double-scale, four-parameter equation for

predicting enthalpies of hydrogen bond formation. This equation can be used to
correlate the enthalpies of interaction in donor-acceptor chemical systems. We have
calculated enthalpies of complex formation from the Drago parameters between eligible
donor (acid) and acceptor (base) constituent group pairs for the homopolymers and
various solvents used in Table 7. The predicted enthalpies of hydrogen bond formation

and the associated standard deviations are shown in Table 8.

The segmental enthalpy of hydrogen bonding for the PVPh/PDMA system can be
modeled by considering interactions between the low molecular weight analogs of the
polymer pendant groups on PVPh and PDMA, phenol and N,N-dimethylformamide
respectively. The value of the enthalpy of hydrogen bond formation between phenol and
N,N-dimethylformamide using Drago parameters is -6.42 +/- 0.24 kcal/mol. In
comparison to the phenol/N,N-dimethylformamide system only the solvents that
dissolved the PVPh/PDMA complex, pyridine and dimethyl sulfoxide, showed a more
negative or more favorable enthalpy of interaction with the phenol group than N,N-
dimethylforrnamide of the solvents listed in Table 8. These calculations suggest that the

PVPh/PDMA complex is probably broken during dissolution and are in agreement with

the 1H NMR results using dimethyl sulfoxide as the solvent. In order to dissolve the
complex the acceptor group has to displace the N,N-dimethylformamide pendant groups
hydrogen bonded to the PVPh. This PVPh/solvent complex would then displace the
interpolymer PVPh/PDMA complex during the dissolution process. The other acceptor
groups (Lewis base solvents) listed in Table 8 had a less favorable or higher enthalpy of
hydrogen bonding, using Drago’s parameters, than N,N-dimethylformamide with
phenol. These solvents in actuality did not dissolve the PVPh/PDMA complex. Thus,
there is a correlation between the interaction enthalpy as calculated by Drago parameters

and the solubility of the complex. These results also suggest that the solubilities of the

50
PVPh/PDMA complex in a solvent can be correlated to Drago’s parameters. Thus, the

Drago parameters could be used to try and predict solubility trends of interpolymer

complexes.

Table 8: Enthalpies of Hydrogen Bond Formation Calculated from Drago Parameters

 

 

 

 

 

 

 

 

 

 

 

Acid / Base -AH, kcal/mol +/- standard deviation
phenol / N,N-dimethylformamide 6.42 +/- 0.24
phenol / pyridine 7.89 +/- 0.14
phenol / dimethyl sulfoxide 7.06 +/- 0.13
phenol / tetrahydrofuran 6.12 +/- 0.13
phenol / acetone 5.30 +/- 0.12
chloroform / dirnethylformamide 4.44 +/- 0.25
phenol / acetonitrile 4.43 +/- 0.09
phenol / benzene 2.42 +/- 0.12

 

 

 

CHAPTER 5

NMR RELAXATION STUDIES OF POLY(ETHYLENE GLYCOL) (PEG)/
POLY(METHACRYLIC ACID) (PMAA) COMPLEXES

5.1 NMR Relaxation

Nuclear magnetic relaxation may be resolved into two components - relaxation
along an axis parallel to the external magnetic field and that in the plane perpendicular to
the field. The former is called spin-lattice or longitudinal relaxation, while the latter is
called spin-spin or transverse relaxation. These nuclear magnetic relaxations are

characterized by the exponential time constants T1 and T2 respectively. Relaxation

ultimately arises from fluctuating magnetic fields experienced by the nuclei as they
interact with other molecules while undergoing random thermal motion. Fluctuating
magnetic field components of the proper frequency lead to nuclear magnetic relaxation.
Because the frequencies of the fluctuating fields are dependent upon the mobility of the

nuclei, so are the relaxation times, T1 and T2. In general T2 decreases monotonically as
the mobility decreases, while T1 passes through a minimum. This behavior is due to the

fact that transverse relaxation relies upon low-frequency contributions to the spectral
density function for efficient relaxation, while longitudinal relaxation relies upon higher

frequency contributions (76,77).

51

52

5.2 NMR Sample Preparation

Monodisperse samples of poly(ethylene glycol) (PEG) (Polysciences) with
molecular weights ranging from 200 to 20,000 as well as polydisperse samples of larger
molecular weight were used as received from the supplier. The structure of PEG is
shown in Figure 13. The molecular weights and polydispersity of the PEG samples used
in this study are listed in Table 9 in the Results and Discussion section of this chapter.

NMR samples were prepared by dissolving appropriate amounts of the PEG in D20

(99.9%, Cambridge Isotope Laboratories). The concentration of the NMR samples was

0.1 +/-0.01 weight percent polymer.

5.2.1 Synthesis and NMR Sample Preparation of PEG/PMAA Complexes

Poly(methacrylic acid) (PMAA) for NMR samples was synthesized at 40°C by
reacting 20 vol% MAA monomer in water with 0.50 wt% of both sodium bisulfite and

ammonium persulfate based upon the total mass. The polymer product was dialyzed
with D20 to exchange H+ with U”, and the resulting polymer was dried in vacuo. The
resulting PMAA was used for NMR relaxational studies of PEG complexed with
PMAA. The structure of PMAA is shown in Figure 14. Samples containing both
species were prepared by dissolving appropriate amounts of each constituent in D20.
The concentration of the NMR samples was approximately 0.1 weight percent polymer.

These samples were typically prepared by first forming dilute solutions of each

polymeric species separately, and then mixing the solutions in appropriate proportions.

53

Figure 13: Structure of poly(ethylene glycol) (PEG).

54

en, en, en, en, en,
4 | ...
...cfiz\é’ cfi:\'c, ctr“? cm“? CH. ‘c\’ CH:

\
,' , 4C , ”
O’c‘oa O’c‘or-I 0 ‘OH O’c‘oa O c‘oH

Figure 14: Structure of poly(methacrylic acid) (PMAA).

55

5.3 Spin-Echo NMR Experiments

High resolution 1H NMR relaxational studies were conducted using a VXR-300
spectrometer (Varian, Palo Alto, CA) located in the Max T. Rogers NMR facility at
Michigan State University or a Gemini-300 spectrometer (Varian) at the Dow Chemical

Company (Midland, Michigan). These experiments were performed with a controlled

temperature of 25°C, using a transmitter frequency of 299.949 MHz. The 180 degree
pulse was measured between 30.0 and 31.8 microseconds. The delay between
successive pulses was at least fifteen seconds (T1 by inversion recovery was less than
two seconds). At least 16 transients were added before Fourier transformation. Spin-
spin relaxation times were measured using a Carr Purcell Meiboom Gill (CPMG) pulse
sequence (78) with a Levitt Freeman 180 degree composite refocussing pulse (79). The

sequence is shown below.
90° (1 180° 21 180° t)n acquire.

The CPMG pulse sequence effectively removed relaxational effects due to magnetic
field heterogeneity and greatly limited the effects of diffusion, while the composite pulse

was used to remove pulse imperfections (79) and off-resonance effects. The NMR tube

was not allowed to spin during the spin echo experiments. The solution state 1H NMR
spectrum of PEG in D20 is very simple with all of the ethylene protons occurring in one

peak at 3.6 5 (Figure 15). The D20 solvent peak occurred at 4.6 ppm. For each polymer

sample the intensity or integral of the ethylene peak was measured as a function of total

time for relaxation (equal to 4nt).

56

"'0 V C C
“\fllmkfll \CH'I M'\“/ .u'\CH.°H

 

 

1L 1.

TFIIIITTIIITIIITTFIIIIIIIITW

7 6 5 4 3 NM

 

Figure 15: High resolution 1H NMR Spectra of poly(ethylene glycol) (PEG) in D20 at

25°C.

57

5.4 Results and Discussion

5.4.1 Spin-Spin Relaxation Times of PEG Ethylene Protons

The spin-spin relaxation of dilute PEG solutions were determined using spin echo
pulse sequences. The spectral results are shown in Figure 16. These experiments were

performed on solutions of 0.1 wt% PEG in D20. The data for the intensity of the

ethylene peak as a function of the total time for relaxation were fit by non-linear least

squares regression to a single decaying exponential, thereby providing a value for T2. In
all cases the error in the value of T2 was less than 5%. Results for a series of

experiments with different PEG molecular weights are shown in Table 9. These results
serve as a basis of comparison for the subsequent experiments using complexed PEG.
As expected, the value of the spin-spin relaxation time decreased as the molecular
weight of the PEG decreases. As illustrated in Table 9, the incremental decrease in the
relaxation time with increasing molecular weight is most pronounced for the smallest
molecular weights. This trend probably arises from a diminishing incremental decrease

in mobility with increasing chain length for long polymer chains.

58

 

 

 

 

 

 

 

 

 

 

 

i J «ab. JOLJLJ»

0.1 03L 0 .‘t

Total Time for Relaxation (seconds)

Figure 16: Ethylene protons of PEG (1H NMR) as a function of total time for delay

using spin echo pulse sequence.

59

Table 9: NMR Spin-Spin Relaxation Times of the PEG Ethylene Protons

 

 

 

 

 

 

 

 

 

 

 

Molecular Weight Polydispersity Relaxation Time (sec)
62 1.00 3.70
202 1.05 0.99
600 0.64
1500 1.05 0.61
5000 1.05 0.55
11, 000 1.04 0.48
20,000 1.13 0.47
5,000,000 0.45

 

 

 

 

60

5.4.2 Spin Echo Experiments on Mixtures of Different Molecular Weight PEG

Mixtures of two different molecular weight PEG species were prepared by

dissolving known amounts of the two PEG species in D20. The total polymer

concentration was 0.1 wt% PEG. Spin echo NMR experiments were performed on these

solutions of two different molecular weight PEG chains in D20. The ethylene peak for

both the lower and higher molecular weight PEG species showed at the same chemical
shift (Figure 17). Therefore, the integral for the peak containing both species was

measured.

When the data for the composite integral of the ethylene peak as a function of
total time for relaxation was fit to a single decaying exponential, there was a significant

amount of error in determining a single T2 for the mixture (Figure 18). A computer

program (see Appendix) was written in order to carry out a non-linear least squares fit to
three parameters: mass fraction and the spin-spin relaxation time for each species. The

equation used in the regression analysis is shown here:
(Composite Integral) = (mass fractionl) (e't/TZ') + (mass fraction) (e'wn). (1)

In this equation, the left side is the value for the integral of the NMR peak
containing both PEG species. The right hand side contains terms for both PEG species
including mass fractions and spin-spin relaxation time. The variable “t” stands for the
total delay time for relaxation in the spin echo experiments. Figure 19 is a graphical
representation of equation 1 showing a relative contribution from each molecular weight
polymer chain to the overall NMR relaxation data. The ethylene proton peak of PEG for

samples containing 62 and 5,000,000 molecular weight PEG are shown in Figure 17.

61

Two experiments were carried out on samples containing two different molecular
weight PEG. The results are shown in Table 10. The known mass fraction for each
polymer species is given in column 2 of Table 10. A computer program was supplied
with initial guesses for individual spin-spin relaxation times (T21, T22) for each PEG

species as well as an initial guess for the mass fractions (m.f.1, mfg). Supplied with

these initial values, the non-linear least squares regression program produced values for
mass fraction shown in the third column of Table 10. The best fit was shown to be
independent of initial guess for the mass fractions. This was shown to be true even for
initial guesses with large deviations away from the known mass fraction values. The
value obtained by regression analysis for the spin-spin relaxation time of each PEG
species was within 10% of the T2 value obtained for the individual species in D20
reported in Table 9. However, the T2 value in the mixture cannot necessarily be

expected to match the T2 value obtained for the individual PEG species alone, due to the

dependence of spin-spin relaxation time on concentration and molecular weight.

 

62

."CH' sail 0H.\cfl.’0‘\ cm/cul'saz c“.\c“....

 

 

 

 

l
.11.. MAJ—Mt.

0—1 0.5 l-‘i 23 5.0

 

 

 

 

 

Total time for relaxation (seconds)

Figure 17: Ethylene protons of a mixture of 62 and 5,000,000 molecular weight PEG

(1H NMR) as a function of total time for delay using spin echo pulse sequence.

63

 

120 L
100-1

E integral

Integral
8

 

 

 

0 ‘ 1 1 I ' I ' I ' I ' i '
o 1000 2000 3000 4000 5000 6000 700
deleyoneee) ‘

Figure 18: Data of composite integral versus total time for delay for the spin echo pulse

sequence on mixtures of 62 and 5,000,000 molecular weight PEG.

- Integral = (m.f.,)(e"’72‘) + (m.f.2)(e'W22)

 

 

\

 

Normalized Amplitude

 

 

J I I
O 250 500 750 1000

 

Delay Time (msec)

Figure 19: Graphical representation of equation 1 with contribution fi'om each molecular

weight polymer in the mixture.

65
When the data were fit to these two decaying exponentials (equation 1), there was

agreement between the known and calculated mass fi'actions as shown in Table 10.
Table 10 shows the results of the spin echo measurements of two mixtures containing

202 and 62 molecular weight PEG; and 62 and 5,000,000 molecular weight PEG chains.

Table 10: Spin Echo Measurements for Mixtures of Different Molecular Weight PEG

 

 

 

 

 

mass fractions
Molecular Weights mass fractions (actual)

(regression)
62, 202 0.45. 0.55 0.42, 0.58
62, 5X10° 0.40, 0.60 0.40, 0.60

 

 

 

The two species fit the proposed two decaying exponential model. Shorter T2

values were consistently calculated for the higher molecular species in the mixture. In
both cases there was good agreement between regression and actual mass fractions. Spin
echo measurements on mixtures of three different molecular weight PEG species were
also performed. However, non-linear least squares regression analysis was unsuccessful.
The program did not give fitted mass fractions that were within an acceptable (+/-10%)
error limit when compared to known values. Therefore improved fitting techniques will
need to be employed when trying to analyze multicomponent samples. Fitting the
spectrum of relaxation time, as a function of delay time, to an inverse Laplace transform
(80) is currently being investigated in our laboratory as a method for resolving

polydisperse polymer samples.

 

66

5.4.3 Spin-Spin Relaxation Times of PEG/PMAA Complexes

Relaxational studies were also performed on PMAA and on PEG/PMAA
complexes. Solutions of 0.1 wt% PMAA and 0.01 wt% PEG were formed by dissolving
appropriate amounts of these polymers in D20, while solutions of PEG/PMAA

complexes were formed by dissolving 0.1 wt% PMAA and 0.01 wt% PEG in the same
tube. Again, values of the spin-spin relaxation time were obtained by fitting the data for
the peak integrals as functions of the time for relaxation to a single decaying exponential.
The relaxation times of both the (rt-methyl and methylene protons on PMAA were less
than 10 msec, while values for PEG are shown in Table 11. The relaxation time for
complexed PEG chains is expected to be shorter than that for uncomplexed chains
because complexation effectively decreases the mobility of the chains, thereby
enhancing the low frequency contributions to the spectral density function. In fact, if

irreversible complexes were formed, the T2 value for the complexed chains would likely

approach that of the PMAA protons (approximately 10 msec), an order of magnitude
lower than the values for uncomplexed PEG (over 500 msec). However, the complexes
are reversible and the PEG chains undergo rapid exchange relative to the NMR
relaxation timescale. Therefore the observed relaxation time is a weighted average of
the values for the complexed and uncomplexed chains (81). As shown in Table 11,

although the presence of the PMAA has a relatively small effect on the T2 value of PEG
of molecular weight 1500, it has a pronounced effect on the T2 value of PEG of

molecular weight 5000. These results suggest that the spin-spin relaxation is indeed
sensitive to the formation of complexes, and that there is a measurable chain length
effect on the complex stability between PEG molecular weights of 1500 and 5000. For

the 1500 molecular weight PEG, the ten percent decrease in T2 from 570 msec to 511

msec provides evidence for complexation even at this relatively short chain length.

67

Table 11: T2 Values for Complexed and Uncomplexed PEG

 

Molecular Weight

     

Complexed (msec)

    

Uncomplexed (msec)

  

 

5000 530 310

 

 

 

 

The experimental results may be explained in terms of the statistical
thermodynamic description of the complexation equilibrium reported by Scranton et al.
(73). These authors considered the complexation thermodynamics of dilute solutions of
free and graft oligomers with complementary polymers. The total free energy change
upon complexation was divided into two contributions which were considered
separately. An internal contribution to the free energy accounted for the conformational
degrees of freedom as well as the segmental binding interactions, while an external
contribution accounted for the configurational (translational) degrees of freedom. The
external contribution could be evaluated from combinatorial considerations, while the
internal canonical and grand canonical partition functions were formulated in terms of
sequence generating functions for trains, tails and loops. A simple random walk model
was used to evaluate statistical weights of those generating functions. Details of these

calculations may be found in reference 73.

In agreement with the experimental results reported here and in the literature, the
theoretical simulations revealed a marked dependence of the equilibrium bound fraction
on the segmental binding free energy and the PEG chain length. Moreover, simulation
results indicated that covalently attaching the complementary polymers to one another
promotes the formation of complexes primarily due to a decrease in the configurational

entropy change upon complexation. If the complementary polymers are covalently

68
attached to one another, they are effectively immobilized and are in close proximity prior
to complex formation. This result corroborates NOE experiments done in this laboratory
(6) which indicate the lack of critical chain length for graft copolymers. Other
simulation results indicate that the loops may be important in the conformation of a
complexed oligomer, and that the conformational averages for the ungrafted case
asymptotically approach those for the grafted case as the segmental binding free energy,

polymer concentration and PEG chain length increase.

CHAPTER 6

CONCLUSIONS and RECOMMENDATIONS

6.1 Complexes of PVPh with PDMA

In the first system, complexes of poly(4-viny1 phenol) (PVPh) with poly(N,N-
dirnethylacrylamide) (PDMA) were synthesized by precipitation from alcohol solutions
and were characterized as to their composition, solubility, thermal properties and
refractive index. A compositional azeotrope in which the feed solution and the complex
exhibited the same composition was observed at about 45 mole% vinyl phenol repeating
units. For overall vinyl phenol mole fractions below the azeotrope, the complex was
enriched in PVPh relative to the feed, while above the azeotrope the mole% of vinyl
phenol in the complex plateaued at a value of about 53 mol%. These results suggest that
a minimum coverage of PDMA by PVPh is required for precipitation, and that once a
PDMA chain is completely covered with PVPh, an increase in the PVPh concentration in
the solution does not lead to an increase in the PVPh content of the complex. The

refractive index of the complex varied linearly with the composition.

Solubility studies revealed that many solvents can dissolve both constituent
homopolymers, but cannot dissolve the complex. These results illustrate that
macromolecular complexation can drastically change polymer-solvent interactions since
polar functionalities are associated with one another in the complex, and are therefore
inaccessible to the solvent. Only strongly hydrogen bonding solvents could dissolve the

complex, and the polymer-polymer hydrogen bonds were apparently disrupted during
69

7O

dissolution. Complexes containing more than 30 mole percent VPh repeating units
exhibited glass transition temperatures higher than either homopolymer. This trend is
due to the fact that hydrogen bonds act as physical crosslinks which must be overcome
for the transition from the glassy state to the rubbery state to occur. The degradation
temperature decreased as the VPh content was increased, with no enhancement due to

complexation.

The future work on the PVPh/PDMA system should include incorporating both
complementary species into one molecule. This should be done by synthesizing random,
block and graft copolymers. The synthesis of graft copolymers could be carried out by
using a macromonomer. Copolymerization would allow further investigation into the
underlying mechanism of complexation between PVPh and PDMA. Copolymerization
could also form a tougher material than the homopolymer complex. The complex should
be molded for the purpose of thermomechanical analysis, dynamic mechanical analysis,
tensile testing, and izod testing. These mechanical tests would complete the picture as to
the strength, processability and range of application for the PVPh/PDMA complex.
Since it has been difficult to melt the constituent homopolymers, the first attempt in this
area might be to solution cast the complex in a volatile solvent. The use of ultra violet-
visible (UV-Vis) and fluorescence spectroscopy for detecting complexation as well as
quantitative determination of the stoichiometry of hydrogen bonding should also be
investigated. Because of the aromatic substituent groups in PVPh, the compound should

fluoresce.

The synthesis of these complexes should also be carried out at different
concentrations in order to see if the apparent complex compositions change. Using more
dilute homopolymer solutions for synthesis of the complexes than are used here (0.1 or
0.01 wt% polymer) would allow further investigation into complexation equilibrium and

entrainment phenomena. The synthesis of the PVPh/PDMA complexes should be

71
studied at different temperatures because hydrogen bonding equilibria is greatly affected

by temperature. The concentration of the supernatant, decanted ofl during isolation of

the complex precipitate, could be measured by 1H NMR. These data would provide
further confidence in the determination of the complex composition by assuming

conservation of mass.

Fourier transform infrared (FT-IR) spectroscopy could be used as a method for
quantitative determination of complexation stoichiometry. Wang et al. (42) have shown

that the absorbance due to the para substitution on the aromatic ring of PVPh at

1 can be used to establish compositions of‘the complex and

approximately 870 cm’
stoichiometry of hydrogen bonding. This can be accomplished due to the lack of any
absorbing group in this area on PDMA. FT-IR could also be used to establish the degree
of hydrogen bonding by resolving both the free and hydrogen bonded carbonyl (C=O)
and hydroxyl (-OH) moieties of the constituent homopolymers in the complex. This
would require the fitting of spectral peaks. These types of methods are outlined in detail
in the book by Coleman et al. (25). In order to obtain the resolution between
intermolecularly complexed and uncomplexed carbonyl and hydroxyl groups, spectra
will have to be collected at elevated temperatures. Moskala et al. (26) and Landry and

Teegarden (41) have shown that at elevated temperatures the absorbances due to

hydrogen bonding interactions are distinguishable and can be resolved.

The glass transition temperature for the complex could be fit to the Kwei (82),
Gordon-Taylor and Fox models for polymer blends. Between these three equations, the
Kwei equation has been shown to be most applicable to hydrogen bonding polymer
systems. Fitting to these models would require the collection of a few more data points
preferably covering the entire composition range for the complex. The simplified Kwei
equation (41,83) requires a relatively small number of data points. This equation has one

adjustable parameter (q) which represents the stabilization energy in the polymer

72
backbone due to molecular interaction. When the data here for the PVPh/PDMA

complex was fit to the simplified Kwei equation, the value obtained for q was between
150 and 200. This value is misleading due to the limited number of data points collected
and the limited composition range covered by the data. Suzuki et al. (44) and Landry
and Teegarden (41) both obtained q values near 100.

Molecular simulations using packages such as POLYGRAFR (Molecular
Simulations, Inc., Waltham, MA) could provide in depth understanding to the underlying
interpolymer hydrogen bonding mechanism. By doing molecular dynamics and
molecular mechanics simulations on the PVPh/PDMA system, probable donor and
acceptor moieties could be identified and the stoichiometry of complexation could be

theoretically calculated.

The complexation chemistry should be applied to complementary polymer
systems in which the interaction is even stronger than PVPh/PDMA such as poly(4-vinyl
phenol) (PVPh) with poly(allyl diphenyl phosphine oxide) (PADPO). These materials

should have greater toughness and more enhanced thermal properties.

 

73

6.2 NMR Analysis of PEG/PMAA Complexes

The spin-spin nuclear magnetic resonance time (T2), as measured by spin echo

experiments, decreased as the molecular weight of the poly(ethylene glycol) (PEG)
chain was increased. The individual mass fractions of two PEG samples with differing
molecular weight were obtained by regression analysis from spin echo data. However,
regression analysis for individual mass fractions on mixtures of three different molecular
weight PEG samples was unsuccessful. Complexation between poly(ethylene glycol)
and poly(methacrylic acid) in dilute aqueous solution was detected by NMR relaxation

spin echo experiments. The change in spin-spin relaxation time (T 2) was correlated with

a change in mobility of the PEG upon complexation. Complexation was detected at

lower poly(ethylene glycol) molecular weights than previously reported.

The future work on the characterization of polymers by spin echo NMR relaxation
should include analysis of systems with many molecular weights (polydisperse). To
accomplish this task, an improved data fitting technique will be needed. The analysis of
spin echo experiments yields the relative contributions of species with different spin-spin
relaxation times by fitting the peak integral as a function of delay time to multiple
decaying exponentials. This technique worked well for the system used here in which
the total number of components making up the sample was known. Resolution of
overlapping peaks based upon relaxation rate constant could be performed by taking the
inverse Laplace transform of the data for integral as a function of delay time (80). The
relaxation rate is the inverse of the spin-spin relaxation time (T2). The inverse Laplace
transform technique would be useful in determining the number of components

contained in the NMR peak, and could provide useful information concerning the

distribution of the relaxation rate constants. Use of this technique would require

74
obtaining the inverse Laplace transform of a discrete data set. In addition, acceptable

resolution in the relaxation rate domain would require a large number of experiments to

be performed (80).

The conditions of the spin echo experiments could also be varied to help broaden

the T2 range. The temperature, oxygen content, concentration, viscosity and solvent are

some of these variables. The most valuable application of this method would in the
analysis of the molecular weight distribution and other useful characteristics of
biological protein chains such as heparin (84). The use of spin echo NMR experiments
to detect complexation should also be applied to other complementary polymer systems

such as the strongly associating systems suggested in section 6.1.

10

11.

12.

13.

14.

15.

REFERENCES

. Tsuchida, E. and Abe, K., Adv. Polym. Sci., 45, 1 (1982).

. Bekturov, E. and Birnendina, L.A., Adv. Polym. Sci., 41, 100 (1980)

Osada, Y., Adv. Polym. Sci., 82, 1 (1987).

Kabanov, V.A. and Papisov, I.M., Vysokomol Soyed., A21, 243 (1979).

Klier, J. and Peppas, N.A., Proc. Adv. Biomed. Polym, 1, 107 (1989).

Klier, J .; Scranton, AB. and Peppas, N.A., Macromolecules, 23, 4944 (1990).

Vinogradov, SN. and Linnell, R.H. Hydrogen Bonding, Van Nostrand Reinhold,

New York, NY (1971).

. Painter, P.C.; Park, Y. and Coleman, M.M., Macromolecules, 21, 66 (1988).

Painter, P.C.; Park, Y. and Coleman, M.M., Macromolecules, 22, 570 (1989).

. Painter, P.C.; Park, Y. and Coleman, M.M., Macromolecules, 22, 580 (1989).

Coleman, M.M.; Lichkus, A.M. and Painter, P.C., Macromolecules, 22, 586 (1989).

Landry, C.J.T. and Teegarden, D.M., Macromolecules, 24, 4310 (1991).

Painter, P.C.; Graf, J.F. and Coleman, M.M., Macromolecules, 24, 5630 (1991).

Yang, T.P.; Pearce, EM. and Kwei, T.I(., Macromolecules, 22, 1813 (1989).

Masson, J.F. and Manley, R. S. J., Macromolecules, 24, 5914 (1991).

75

16

17

l8

19

20

21

22.

23.

24.

25.

26.

27.

76
. Pearce, E.M.; Kwei, T.K. and Min, B.Y., J. Macomol. Sci, Chem, A21(8&9), 1181

(1984).

. Pennacchia, J.R.; Pearce,E.M.; Kwei, T.K.; Bulkin, BJ. and Chen, J-P,
Macromolecules, 19, 973 (1986).

. Coleman, M.M.; Hu, J.; Park, Y. and Painter, P.C., Polymer, 29, 1659 (1988).

. Ting, S.P.; Pearce, EM. and Kwei, T.K., J. Poly. Sci, Poly. Lett. Ed., 18, 20 (1980).

. Makhija, S.; Pearce, E.M.; Kwei, T.K. and Liu, F., Polym. Eng Sci, 30(13), 798
(1990).

. Weiss, R.A.; Beretta, C.; Sasongko, S. and Garton, A., J. Appl. Polym. Sci, 41, 91

(1990).

Ohno, N. and Kumanotani, J., Polym. J., 11, 947 (1979).

Simmons, A. and Natansohn, A, Macromolecules, 25(15), 3885 (1992).

Hara, M. and Eisenberg, A., Macromolecules, 17, 1335 (1984).

Coleman, M.M.; Graf, J.F. and Painter, P.C., Specific Interactions and the Miscibility

of Polymer Blends, Technomic, Lancaster, Pennsylvania (1991).

Moskala, E.J.; Howe, S.E.; Painter, PC. and Coleman, M.M., Macromolecules, 17,

1671 (1984).

Moskala, E.J.; Vamell, DE and Coleman, M.M., Polymer, 26, 228 (1985).

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

77
Moskala, E.J.; Runt, JP. and Coleman, M.M., In Multicomponent Polymer

Materials; Paul, DR. and Sperling, L.H. Eds., Advances in Chemistry Series 211,

American Chemical Society, Washington, DC. (1986).

Meftahi, M.V. de and Frechet, J.M.J., Polymer, 29, 477 (1988).

Qin, C.; Pires, A.T.N. and Belfiore, L.A., Macromolecules, 24, 666 (1991).

Qin, C; Pires, A.T.N. and Belfiore, L.A., Polymer, 31, 177 (1990).

Zhang, X.; Takegoshi, K. and Hikichi, K., Macromolecules, 25(9), 2336 (1992).

Ting, S.P.; Bulkin, BI. and Pearce, E.M., J. Poly. Sci, Poly. Chem. Ed., 19, 1451

(1981).

Serman, C.J.; Xu, Y.; Painter, PC. and Coleman, M.M., Polymer, 32(3), 516 (1991).

Serman, C.J.; Painter, PC. and Coleman, M.M., Polymer, 32(6), 1049 (1991).

Jong, L.; Pearce, EM. and leei, T.K., Macromolecules, 23(24), 5071 (1990).

French, R.N.; Machado, J.M. and Lin-Vien, D., Polymer, 33(4), 755 (1992).

French, R.N.; Machado, J.M. and Lin-Vien, D., Polymer, 33(4), 760 (1992).

Machado, J.M. and French, R.N., Polymer, 33(2), 439 (1992).

Coleman, M.M., Xu, Y. and Painter, P.C., Makromol Chem, Makromol. Symp., 52,

75 (1991).

Landry, C.J.T., and Teegarden, D.M., Macromolecules, 24, 4310 (1991).

42.

43.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

78
Wang, L.F.; Pearce, EM. and Kwei, T.K., J. Polym. Sci, Polym. Phys. Ed., 29, 619

(1991).

Suzuki, T.; Pearce, EM. and Kwei, T.K., Polym. Mater. Sci. Eng., 65, 329 (1991).

Suzuki, Y.; Pearce, EM. and Kwei, T.K., Polymer, 33(1), 198 (1992).

Bailey, F.E.; Lundberg, RD. and Callard, R.W., J. Polym. Sci , A2, 845 (1964).

Antipina, A.D.; Baranovsky, V.Y.; Papisov, I.M. and Kabanov, V.A.,Vysokomol.

Soyed., A14, 941 (1972).

Baranovsky, V.Y.; Litmanovich, A.A.; Papisov, I.M. and Kabanov, V.A., Eur.

Polym. J., 17, 696 (1972).

Osada, Y. and Sata, M., J.Polym. Sci, Polym. Lett., 14, 129 (1976).

Ikawa, T.; Abe, K.; Honda, K. and Tsuchida, E., J. Polym. Sci, Polym. Chem, 13,

505 (1975).

Osada, Y., J. Polym. Sci, Polym. Chem, 17, 3485 (1979).

Papisov, I.M.; Baranovsky, V.Y.; Serfgieva, Y.I.; Antipina, AD. and Kabanov, V.A.,

Vysokomol. Soyed., A16, 1133 (1974).

Oyama, H.T.; Tang, WT. and Frank, C.W., Polym. Mater. Sci. Prepr., 57, 13 (1987).

Oyama, H.T.; Tang, W.T.; and Frank, C.W., Macromolecules, 20, 474 (1987).

Char, K.; Frank, C.W.; Gast, AP. and Tang, W.T., Macromolecules, 20, 1833 (1987).

 

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

79
Oyama, H.T.; Tang, WT. and Frank, C.W., Macromolecules, 20, 1839 (1987).

Chen, H.L. and Morawetz, H., Eur. Polym. J., 19, 923 (1983).

Bednar, B.; Morawetz, H. and Schaeffer, J.A., Macromolecules, 17, 1636 (1984).

Heyward, U. and Ghiggino, K.P., Macromolecules, 22, 1159 (1989).

Abe, K.; Koide, M. and Tsuchida, E., Macromolecules, 10, 1259 (1977).

Osada, Y., J. Polym. Sci, Polym. Chem, 15, 255 (1977).

Osada, Y., J. Polym. Sci, Polym. Lett., 18, 281 (1980).

Osada, Y. and Takeuche, Y., J. Polym. Sci, Polym. Lett., 19, 303 (1981).

Osada, Y.; Honda, K. and Ohta, M., J. Membr. Sci, 27, 327 (1986).

Nishi, S. and Kotaka, T., Macromolecules, 19, 978 (1986).

Scranton, A.B.; Klier, J. and Aronson, C.L., In Polyelectrolyte Gels; Harland, RS.

and Prud’homme, R.K., Eds.,; ACS Symposium Series 480, American Chemical

Society, Washington, DC. (1992).

Papisov, I.M. and Litmanovich, A.A., Vysokomol. Soyed., A19, 716 (1977).

Magee, W.S.; Gibbs, J .H. and Zimm, B.H., Biopolymers, 1, 133 (1963).

Damle, V.N., Biopolymers, 9, 353 (1970).

Poland, D., Cooperative Equilibria in Physical Biochemistry, Oxford University

Press, Oxford (1978).

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

80
Poland, D. and Sheraga, H.A., Theory of Helix-Coil Transitions in Biopolymers,

Academic Press, New York, NY (1970).

Lifson, S., J. Chem. Phys., 40, 3705 (1964).

Eichinger, B.E.; Jackson, D.M. and McKay, B.D., J. Chem. Phys., 85, 5299 (1986).
Scranton, A.B.; Klier, J. and Peppas, N.A., J. Polym. Sci, Polym. Phys., 29, 221
(1991).

Trossarelli, L. and Meirone, M., J. Polym. Sci, 57, 445 (1962).

Drago, R.S; Vogel, CC. and Needham, T.E., J. Am. Chem. Soc., 93(23), 6014 (1971).
Abragam, A., Principles of Nuclear Magnetism, Oxford University Press, New York,
NY (1961).

Slichter, C.P., Principles of Magnetic Resonance, Springer-Verlag, New York, NY
(1990).

Carr, H.Y. and Purcell, E.M., Phys. Rev., 94, 630 (1954).

Levitt, M.H. and Freeman, R. J. Magn. Reson., 43, 65 (1981).
Scranton, A.B., Ph.D. Thesis, Purdue University, May, 1990.

Sanders, J.K.M. and Hunter, B.K., Modern NMR Spectroscopy, Oxford University

Press, New York, NY (1987).

82.

83.

84.

81
Ting, S.P.; Pearce, EM. and Kwei, T.K., J. Polym. Sci, Polym. Lett. Ed., 18,

201 (1980).
Lin, A.A.; Kwei, T.K. and Reiser, A., Macromolecules, 22(10), 4112 (1989).
Schwartz, K.A., Michigan State University, East Lansing, Michigan, personal

communication, 1991.

APPENDIX

Computer Program with IMSL Subroutine for Calculating the NMR Spin-Spin Time

Constant (T2) by Non-Linear Least Squares Regression Using Equation 1 on page 60.

nnnn

WWW

WW

Carl Aronson

Computer Program with IHSL subroutine for
Non-linear Regression to solve to: best (It to two
Decaying exponentials

Integer 1dr.noba,nparm

parameter lnoba-ZZ. nparmelo. Ids-nparml

integer Idertv, Irank,nout

zeal dte.axampl. :(1dx,npa:ml,aae, theta(nparmi,
xdatalnoba),ydata(nobsl
comon Ixydata/xdata. ydata

external exampl.zn11n,umach,wrrrn

data theta/46.44,-O.2857,0.0.-0.3.0.0.-.34.-56.76.°1.0.0.0.-1.26/
call umach (2,noutl

ideriv-O

call rnlinlexampl,nparm,idetlv,theta.t.1dr.irank,dfe.sae)
writetnout.') 'theta- '. theta

uritelnout.‘) 'trnk- ', irank, ' dfe- ',dfe. ' sse- '.sse

call wrtrnt'r',nparm,nparm,r,1dr,0)
end

subroutine exampl (nparm,theta.topt. robs.trq,et,e. de,tend)
integer nparm.iopt.iobs.iend

real thetaInpatml,trq,vt,e.delll

integer nobs

parameterlnobs-Z?)

real exp, xdatalnobs), ydatalnobsi

common Ixydata/ xdata,ydata

intrrnsrc exp

if (Iobs .le. nobs) then
vt-I.Oe0
{sq-1.0e0
iend-O
e- yoataliobs) - (thetall)'explthetal?)'xdataliobs))0
thetal3l'explthetat4)'xdataltobs))OthetaISI'explxdataltobsl'
thetal6)letheta(7)'explxdata(iobsl'thetatat10theta(9)'exp(xdata
(tobst'thetaIIOIII

else

tend-1

end If

return

end

block data xy
integer nob:
parameter (nobs-22)

real xdatatnobsl,ydata(nobs)
comon lxydata/ xdata, ydata

data xdata/O.l,0.2,0.3.0.4,0.5,0.6.O.7.O.B,O.9,1.0.1.4,1.6,l.8,
2.3.2.6.2.9.3.2,3.6.3.9.4.4.5.4,6.0/

data ydata/103.2,101.999,93.79.82.12.85.7728,80.278,71.5402,
70.9123.65.2595.60.2525,50.9971.45.9541.44.6843.33.8635.30.3355o
27.119.23.74,20.7712,18.4181,14.8399.12.3019o10.9268/

end

82

VITA

Carl Lawrence Aronson was born in Bethesda, Maryland, United States of
America on January 17, 1968 to two ecstatic Jewish parents. He grew up in several
places throughout the USA due to his father’s extensive medical training. These places
included Bethesda, Maryland; Santa Barbara, California; Cumberland, Maryland; East
Lansing, Michigan; Tulsa, Oklahoma and East Grand Rapids, Michigan. In the Spring
of 1990, he graduated from East Grand Rapids, Michigan High School in the top ten of
his class. In the Fall of 1990, he entered Hope College, Holland Michigan with the help
of Presidential and Dow Chemical Company scholarships. During four summers at
Hope, he engaged in chemistry research with financial support from the Dow Chemical
Company, Midland, Michigan. After four years in Holland, he graduated Cum Laude
from Hope College in June of 1990 with a Bachelor of Science in Chemistry and a

Bachelor of Arts in Music (Trumpet Performance) as well as a Mathematics minor.

During the Summer of 1989, Carl did research under Dr. R. Mark Worden in the
Department of Chemical Engineering at Michigan State University, East Lansing,
Michigan. Due mainly to this research experience, he decided to attend graduate school
at Michigan State University in Chemical Engineering. He accepted an assistantship
from Michigan State University and entered graduate school in Chemical Engineering
eight days after graduating from college. In the Summer of 1991, he received a four year
graduate fellowship from the Michigan Polymer Consortium. During his graduate
research, Carl studied under the tutelage of Professor Alec Byron Scranton. Carl is the

first student to earn an advanced degree under Dr. Alec B. Scranton.

MICHIGAN smr: UNIV. Lreeenrss
Mill ”Hill "Will Illlllllllllll lllllllllllllllll
31293007929098