Skip to main content

Full text of "Applications of a ruthenium catalyst in acyclic diene metathesis (ADMET) chemistry : depolymerization and polyolefin modelling studies"

See other formats


APPLICATIONS OF A RUTHENIUM CATALYST IN ACYCLIC DIENE 
METATHESIS (ADMET) CHEMISTRY: DEPOLYMERIZATION AND 
POLYOLEFIN MODELLING STUDIES 



By 

MARK D. WATSON 



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

UNIVERSITY OF FLORIDA 



1999 



ACKNOWLEDGEMENTS 



I have been told many times by Dr. Ken Wagener, my graduate advisor, that 
graduate school can be the best time of one’s life. As a result of his guidance, the 
stimulating environment provided by the University of Florida’s Department of 
Chemistry and the people within it; and of course the arrival of my wife, Kim, I am 
inclined to agree. 

The polymer group is made up of a unique collage of people excited about their 
work and life outside the lab. I am truly grateful to members, past and present, of the 
polymer group for providing both a stimulating working environment discussing 
chemistry on the chalkboards, written on paper towels, etc. and a great social life from 
shooting pool to canoe trips. My time here was made rich by the acquaintances of 
Fernando Gomez, Debby Tindall, Jason Smith, Jason Portmess (the group cheerleader, 
his presence is sorely missed), Tammy Davidson, Dominique Valenti, Todd Younkin. 
and Karen Dai. Special thanks go to Shane Wolfe my long time friend and roommate for 
nearly the duration, BOO-YOW! BOTBOTH! I am very fortunate to have shared a lab 
with Krystyna Brzezinska as she shared friendship, unique perspective, and an 
incomparable professional example in the laboratory. 

Thanks go to members of the faculty including Dr. Merle Battiste (for many late 
afternoon discussions of chemistry and otherwise), Dr. William Dolbier, Dr. Lisa 
McElwee-White, and Dr. Eric Enholm for sharing with one struggling to become a 



ii 



successful chemist. Lorraine Williams remains a marvel to me. Her delightful persona, 
which accompanies an endless capacity for helping in any situation, is the glue that holds 
the polymer group together. In similar fashion, Donna Balkcom made my life here at UF 
much smoother doing all the things which she does behind the scenes for all graduate 
students. Thanks go to members of the faculty and staff who offered excellent technical 
support in characterization including Dave Powell, Lidia Matveeva, Ion Ghiviriga, and 
Kathryn Williams. 

Dr. Ken Wagener truly fulfilled the designations of advisor and mentor not only 
in professional but personal growth. He not only provided the where-with-all for 
conducting research but also exuded the excitement of someone who sincerely pursues 
the task of mentor. 

Gratitude is also extended to the National Science Foundation for financial 
support which made this work possible. 

I am most grateful for having met my new and permanent roommate, my wife 
Kim. If I had only known what I know now nine years ago when our paths first briefly 
crossed. 



iii 



TABLE OF CONTENTS 



page 

ACKNOWLEDGMENTS ii 

ABSTRACT xii 

CHAPTERS 

1 INTRODUCTION 1 

The Olefin Metathesis Reaction 2 

Development of the Olefin Metathesis Reaction 6 

Well-Defined Metathesis Catalysts 8 

Ruthenium Metathesis Catalysts 1 1 

Discrete Ruthenium Carbene Complexes 13 

Acyclic Diene Metathesis (ADMET) Polymerization 16 

Summary of Results Presented in this Dissertation 20 

2 EXPERIMENTAL 24 

Instrumentation and Analysis 24 

Materials and Techniques 26 

Synthesis and Characterization 27 

ADMET Depolymerizations 27 

Metathesis and in situ Hydrogenation 28 

Synthesis ofFunctionalized Diene Monomers for Model Ethylene 

Copolymers 32 

Synthesis of extended chain co-alkenyl bromides and carboxylic acids 32 

Synthesis of symmetrical alcohol-functionalized dienes 35 

Synthesis of symmetrical dienes with pendant acetate groups 38 

Synthesis of symmetrical dienes with pendant carboxylic and 

alkoxycarbonyl groups 40 

Synthesis of symmetrical dienes with pendant chloride and phenyl 

groups 43 

Synthesis of ADMET Model Ethylene/Polar Monomer Copolymers 45 

Synthesis of ADMET model ethylene/vinyl acetate copolymers 46 



IV 



Synthesis of ADMET model ethylene/acrylate copolymers 47 

Synthesis of ADMET model ethylene/vinyl chloride and styrene 

Copolymers 48 

3 ADMET DEPOLYMERIZATION OF 1 ,4-POLYBUTADIENE UTILIZING 

A WELL-DEFINED RUTHENIUM METATHESIS CATALYST 50 

Ethenolysis of 1,4-Polybutadiene 53 

Ethenolysis of 1,4-Polybutadiene Catalysed by a Well-Defined Ruthenium 

Complex 54 

Bulk Depolymerization of Polybutadiene 62 

Bulk Depolymerization of Polybutadiene to Produce Telechelics 67 

Conclusions 69 

4 TANDEM HOMOGENEOUS METATHESIS/HETEROGENEOUS 

HYDROGENATION 71 

Immobilization of a Well-Defined Ruthenium Complex on the Surface of 

Silica 72 

Catalytic Activity of _L4 Adsorbed on Silica 74 

Hydrogenation with the Residue of Catalyst L4 Adsorbed on Silica 78 

Preparation of Saturated Polymers by Tandem Homogeneous 

ADMET/Heterogeneous Hydrogenation 83 

Preparation of End-Functionalized Polyethylene by 

ADMET/Hydrogenation 83 

Synthesis of a Polyester with Long Aliphatic Segments 85 

Conclusions 89 

5 THE SYNTHESIS OF MODEL ETHYLENE/POLAR MONOMER 

COPOLYMERS VIA ADMET AND HYDROGENATION 90 

Ethylene/Polar Monomer Copolymers 93 

Model Polymers 95 

Preparation of Model Polymers via Metathesis 97 

Model Polymers via Metathesis Polymerization and Hydrogenation 98 

Design and Synthesis of Symmetrical Diene Monomers for Ethylene/Polar 

Monomer Model Polymers 102 

Synthesis of Acetoxy-Functional Monomers 105 

Chloro- and Phenyl-Substituted Dienes 106 

Monomers with Pendant Alkoxy-Carbonyl Moities 107 



Preparation and Characterization of ADMET Ethylene/Polar Monomer 

Copolymers 110 

ADMET Ethylene/Vinyl Acetate Model Copolymers 1 1 1 

Thermal Analysis of EVA Model Copolymers 1 15 

ADMET Ethylene/Alkyl Acrylate Copolymers 122 

ADMET Ethylene/Styrene Copolymer 126 

ADMET Ethylen/Vinyl Chloride Copolymer 128 

Conclusions 130 

REFERENCES 132 

BIOGRAPHICAL SKETCH 139 



VI 



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



APPLICATIONS OF A RUTHENIUM CATALYST IN ACYCLIC DIENE 
METATHESIS (ADMET) CHEMISTRY: DEPOLYMERIZATION AND 
POLYOLEFIN MODELLING STUDIES 



By 

Mark D. Watson 



May 1999 



Chairman: Kenneth B. Wagener 
Major Department: Chemistry 

Investigation of the utility of Grubbs’ ruthenium carbene, Ru(CHPh)(PCy 3 ) 2 Cl 2 , 
as a catalyst for various operations of Acyclic Diene Metathesis (ADMET) chemistry is 
presented. Applications include the metathesis depolymerization of cis- 1,4- 
polybutadiene, ADMET polymerization of functionalized dienes, and tandem metathesis/ 
hydrogenation reactions in which the ruthenium complex serves to catalyze both 
reactions. 

In Chapter 3, the unprecedented stability of the ruthenium complex in the 
presence of ethylene is exploited for effective depolymerization of 1,4-polybutadiene by 
cross-metathesis with ethylene. Vastly improved yields of hexadiene over those 
previously reported utilizing well-defined early transition metal complexes were 



vii 



realized. The first accounts of the remarkable ability of the ruthenium complex to effect 
the solvent-free depolymerization of 1,4-polybutadiene and its application in the 
synthesis of end-functionalized polymers are also disclosed. 

In Chapter 4, the dual utility of the catalyst for both olefin metathesis and 
hydrogenation is described. Adsorption of the catalyst residue on the surface of silica gel 
converts it to a heterogeneous system, which is effective for catalytic olefin 
hydrogenation at room temperature and moderate pressures. The two reactions may be 
conducted in tandem to produce saturated polymers, including telechelic polyethylene. 
Further application of the tandem process for C-C bond construction in small molecule 
synthesis is described. 

In Chapter 5, the metathesis/hydrogenation process disclosed in Chapter 4 is 
exploited for the production of model periodic ethylene copolymers. A homologous 
series of symmetrical dienes with varied central pendant functionality were synthesized 
to this end. ADMET polymerization of these monomers and subsequent hydrogenation 
yielded periodic models for copolymers of ethylene with vinyl acetate, vinyl chloride, 
acrylates, and styrene. Thermal analysis of the resulting polymers revealed the predicted 
relation of increasing melting point with increasing ethylene run length. Within a series 
of polymers with identical ethylene run lengths, increasing steric bulk of the regular 
pendant functionality resulted in greater melting point depression. 



viii 



CHAPTER 1 
INTRODUCTION 



The now continuous stream of advances in polymer chemistry attending new 
developments in catalytic organometallic chemistry 1 can generally be classified into two 
categories: (1) original chemical transforms and (2) the betterment of known chemical 
transforms. The latter class may be manifestations of improved catalyst efficiency, 
chemoselectivity, broader scope to encompass substrates with diverse functionalities, and 
resistance to ubiquitous trace poisons allowing more practical implementation. 

The subject of this dissertation, which falls into the second category above, 
involves the exploitation of unique properties of a well-defined ruthenium metathesis 
catalyst to improve upon and expand the scope of various applications of Acyclic Diene 
Metathesis (ADMET) chemistry. One aspect of the studies focussed on ADMET 
depolymerization where the resilience of the catalyst was exploited for improved 
conversions in reactions involving ethylene as a substrate. Also the first, to the best of 
our knowledge, “solid state” depolymerizations of ultra-high molecular weight 
polybutadiene catalyzed by a discrete metathesis catalyst were performed. Exciting 
observations were made regarding the dual-purpose use of the catalyst for post- 
metathesis hydrogenation chemistry, made possible by the nature of the late-transition 
metal center. The timing of these observations coincided with the growing implication of 
metathesis polymerization in tandem with hydrogenation as a powerful method for 



1 



2 



macromolecular engineering. This prompted the development of a facile process for the 
production of periodic model ethylene copolymers involving homogeneous metathesis in 
tandem with heterogeneous hydrogenation. 

To facilitate cohesive presentation, the somewhat disparate subjects of 
depolymerization, hydrogenation, and model polymer engineering are apportioned into 
chapters 3, 4, and 5, respectively. The remainder of this chapter will serve to briefly 
introduce the metathesis reaction and advances in catalyst development pertinent to the 
the establishment of the ADMET reaction as a viable synthetic scheme. The ADMET 
reaction, which is the unifying concept common to all the studies reported herein, 
deserves separate discussion for adequate understanding of the subject matter. 

The Olefin Metathesis Reaction 

The olefin metathesis reaction is a powerful organic transform, which formally 
divides olefinic compounds at the carbon-carbon double bonds into carbene fragments 
and then redistributes them, forming new olefins. 2 The process is catalyzed by a metal 
carbene complex which may be preformed or formed in situ, and may be homogeneous, 
heterogeneous, or supported on an insoluble substrate. The key steps in the mechanism 
are shown in Figure 1-1. Coordination of a substrate olefin to the metal center leads to 
the formation of a metallocyclobutane. Cleavage of this ring may occur either along the 
vertical or horizontal leading to, respectively, starting materials or a new olefin and new 
metal carbene. The olefin product may be cis, trans or a mixture as determined by the 
steric demands of the substrate and the choice of catalyst. In the special case where R = 
R”, termed degenerate metathesis, no new products are formed in the forward reaction. 



3 



The mechanism finds analogy in the Wittig reaction, shown for comparison in 
Figure 1-1, where a carbene fragment is delivered to a carbon doubly bonded to oxygen 
forming a new C=C bond. A major difference between the two mechanisms lies in the 
reactivity of the other doubly bonded product liberated upon cleavage of the cyclic 
intermediate. In contrast to the Wittig reaction, the new metal carbene formed in the 
metathesis reaction can repeat the reaction with any other reactive C=C bond. 



Wittig Reaction(Stoichiometric, irreversible) 



.R 



UP= 



O: 



l 3 p 



K 






O 



'R‘ 



^R' 



Unreactive 




Olefin Metathesis (Catalytic, reversible) 




R" 




L n M-r 



R"' 



.R 



’R‘ 



Reactive 




Figure 1-1. Key steps of the olefin metathesis reaction compared to the Wittig reaction. 



2 i • 

The olefin metathesis reaction may be divided into categories depending on the 
structure of the starting materials and products as depicted in Figure 1-2. Regardless of 
the particular pathway, it is now widely accepted that each of these processes has as 
common intermediates the metal carbene and the metallocyclobutane. In the simplest 
case of a single symmetrical olefin, the substrate will be isomerized to an equilibrium 
mixture of cis/trans stereoisomers via repetitive degenerate metathesis. For an 
unsymmetical olefin substrate or a mixture of different starting olefins of similar 
reactivity, the result is a statistical redistribution of carbene fragments upon reaching 



4 



equilibrium. Following Le Chatelier’s Principle, removal or addition of any of the 
reactants or products may shift the position of the equilibrium. 

Acyclic 

R'^^R 

Ring Opening Metathesis Polymerization (ROMP) 




Ring Opening Cross-Metathesis 




Acyclic Diene Metathesis Polymerization (ADMET) 

^"R^ - ^R^ + C 2 R 4 

Figure 1-2. Classifications of olefin metathesis reactions. 

By ring opening metathesis polymerization (ROMP), cyclic olefins may be 
converted to high polymer via a chain process provided the energetics are favorable. The 
process is thermodynamically favored for 3-, 4-, 8-, and larger membered rings but is 
unfavorable for cyclcohexene due to stability of the six-membered ring. Cyclic olefins 
may also be ring opened to monomeric compounds by cross-metathesis of cyclic and 



5 



acyclic olefins. This “arrested ROMP” reaction occurs when cross-metathesis of the 
newly formed metal carbene from either the cyclic or acyclic olefin is favored over self- 
metathesis. 

Conversely, compounds with more than one double bond appropriately spatially 
disposed may be cyclized by ring closing metathesis (RCM). This particular transform 
has been extensively exploited for the preparation of fine chemicals and extended to 
complicated natural product syntheses. Dienes may also be condensed intermolecularly 
to polyenes and high polymers by acyclic diene metathesis polymerization (ADMET), to 
be discussed in detail later. Competition between RCM and ADMET often exists, and its 
outcome may be governed to varying degrees by altering reaction conditions. The most 
significant variable is concentration, where dilute conditions favor intramolecular (RCM) 
reaction and bulk conditions favor intermolecular (ADMET) reaction. Chemical 
characteristics that hold the double bonds in close proximity, such as bulky groups 
between the olefins (Thorpe-Ingold Effect) and intramolecular hydrogen bonding may 
augment cyclization of dienes at the expense of linear products. 1 

Recently, Grubbs reported 4 an elegant study in which ADMET, RCM, or ROMP 
could be driven in high conversion from the same olefinic substructure depending on 
reaction conditions as shown in Figure 1-3. Under bulk conditions, the oligoether diene 
was condensed to high polymer by ADMET. Introduction of an appropiate metal cation 
to a solution of either the polymer or the monomeric diene leads to hapto-coordination of 
the ether oxygens to the cation, bringing sequential olefins into proximity favoring RCM. 
The resulting crown ether could be obtained either from the diene or from high polymer 



6 



in high yields. Removal of the cation in the presence of active metathesis catalyst leads 
to ROMP of the crown ether to high polymer. 



ADMET 




Figure 1-3. Interplay of ADMET, RCM and ROMP utilizing a template. Metathesis 
conversion >95% in all three cases by NMR. 

Development of the Olefin Metathesis Reaction 

The history of the reaction parallels other catalytic processes such as Ziegler- 
Natta polymerization where obvious technical utility spurred almost immediate 
commercial exploitation preceding a fundamental understanding. 2 The reaction found 
application in the interconversion of hydrocarbon streams, diversifying olefin- 
manufacturing processes. These processes typically used empirically developed 
heterogeneous catalysts where a transition metal precatalyst was loaded onto an inorganic 
support and pretreated to generate the active catalyst, the structure of which was 
unknown. Two such processes 2 ' 5 that were developed by Phillips Petroleum Co. are 
depicted in Figure 1-4. The Phillips triolefin process, implemented in 1966, utilized 
metathesis to convert propene to ethylene and but-2-ene. In 1969, the Neohexene 
process was initiated, where neohexene is generated from ethylene and 2,4,4- 
trimethylpent-2-ene. 



7 



In spite of these early commercial applications of the process, it was not until the 
beginning of the 1970’s that fundamental research resulted in proposal of the now- 
accepted mechanism 6 (cf Figure 1-1) involving the key intermediacy of metal carbenes 
and metallocyclobutanes. The reaction became the subject of a number of research 
programs, both academic and applied, where the reasons for fundamental investigation 
were manifold. 7 The initial commercial catalyst systems required excessive temperatures 
for practical activity, limiting the substrates to those with functionality that could 
withstand the harsh conditions. Application was also limited by low tolerance of the 
catalysts to Lewis basic functionality. 8 In addition, lack of insight into the mechanism 
prohibited prudent catalyst engineering. 

Phillips Triolefin Process 

^ 

Neohexene Process 




Figure 1-4. Early commercial applications of the olefin metathesis reaction. 



Specific refinements were necessary to make the metathesis reaction a more 
general tool: (a) The catalyst must promote only one reaction (metathesis) with olefins; 
(b) the catalyst must tolerate and remain inert to other functionality (c) the catalyst should 
be active under milder conditions. Refinement in these aspects continues today through 
variation in the choice of the metal center, cocatalysts, and ligand environment. 



8 



Early catalysts were two part systems most commonly based on Mo, W, and Re 
activated by a main-group metal alkyl cocatalyst 9 such as A1 and Zn alkyls. The 
cocatalysts may serve more than one functions such as: A) provide alkyl ligands via 
transmetallation which can rearrange to the alkylidene by elimination processes, B) act as 
a ligand, modifying electron density of the transition metal through, for example, 
coordination to oxo and halo ligands. Due to their high Lewis acidity, some cocatalysts 
initiated side reactions via electrophilic reaction with C=C bonds. This shortcoming was 
overcome with the realization of cocatalysts with diminished Lewis acidity such as those 
based on Pb IOa,b and Sn IOc f . Byproducts of a given catalyst system might also induce 
deleterious chemistry as in the case of WCl 6 /Me 4 Sn which generates HC1. 2 

Greater functional group tolerance was realized with the development of catalysts 
based on late transition metals. 8 The late transition metals are softer Lewis acids and as a 
result, metathesis catalysts derived from these are more selective towards olefin groups in 
the presence of polar functionality. Heterogeneous systems based on Re have been 
studied extensively" and have proven to be more tolerant of functionality than early 
transition metal species. Though the process has yet to achieve commercial success, 
heterogeneous Re systems have been proven effective for the metathesis of unsaturated 
esters. 



Well-Defined Metathesis Catalysts 

Research efforts aimed at delineating the mechanism of the olefin metathesis 
reaction shifted towards the development of active preformed metal carbenes. Although 
stable isolable metal carbenes were reported as early as 1964 by Fischer, these species 



9 



1 9 

proved to be inactive for metathesis. The first isolable, metathetically active carbene 
complex was reported in 1974 13 and shown to effect the ROMP of cyclooctene shortly 
thereafter 14 . This complex, C 05 W=CPh 2 , required activation for metathesis through loss 
of a CO ligand induced by heating or UV irradiation but still represented a significant 
advance as the necessity for cocatalyst was eliminated. 

Various researchers subsequently secured their place in the history of this drive to 
obtain well-defined metathesis catalysts, but most were to be overshadowed by the names 
Schrock and Grubbs. The former prepared the first class of highly active species based 
on W and Mo systems which were soon complemented by the Ru systems of Grubbs, to 
be discussed later. 





R 


M = W 


\ 

M = Mo 


/-Pr-f^-Pr 


t- but 


1.1a 


1.2a 


_ _ N 



«/ Y 


CF 3 (CH 3 ) 2 C 


1.1b 


1.2b 


CH 3 (CF 3 ) 2 C 


1.1c 


1.2c 


V 









Neopentylidenes 



i-P r 




/- Pr 



CH3(CF 3 ) 2 COm..^ 0 

Ch^CF^Cd^ ^ Ph 

1.2d 



Neophylidene 



Figure 1-5. Well-defined, highly active W(VI) and Mo(VI) alkylidines. 



Through systematic variation of ligand environment, Schrock eventually 
developed a series of four-coordinate W(VI) and later Mo(VI) alkylidenes with tunable 
reactivity, shown in Figure 1-5. The bulky imido and alkoxy ligands serve in one fashion 
to diminish bimolecular catalyst decomposition while still allowing the approach and 
coordination of olefinic substrates. Variation of the steric and electronic nature of the 
alkoxide ligands allows the reactivity of the metal alkylidine to be engineered to differing 



10 



degrees, with highest reactivity observed for R = CH 3 (CF 3 ) 2 C, or hexafluoro-t-butoxy. 
For example, catalyst 1.1a is effective for ROMP but not acyclic chemistry while 1.1c 
not only catalyzes ROMP but is highly active for acyclic metathesis (TON = 1000 min' 1 
for 2-pentene). While representing a landmark advance in catalyst development, the W 
analogues still suffered from low tolerance to functionality. For example, although 
catalyst 1.1c effected the metathesis of methyl oleate, catalyst lifetime was limited by 
deactivation via Wittig-type reaction with the ester carbonyl group. 

The less electrophilic molybdenum analogues proved to be more tolerant of 
functionality and became the vehicle by which olefin metathesis was transformed into a 
trustworthy synthetic tool. The Mo catalysts’ improved chemoselectivity has allowed the 
ROMP, RCM and acyclic metathesis of substrates containing a broad range of 
functionality including halo, ether, amine, amide, ester, thioether, and metal containing 
groups. 15 The neophylidene complex 1.2d , although no more effective than the 
neopentylidene analogue, is more easily prepared and therefore preferred. 

Utilization of these catalyst systems for ROMP chemistry offers several 
advantages in addition to greater mechanistic insight. They allowed the preparation of a 
broad range of polymers in living fashion with predictable and narrow PDI's as well as 
block copolymers with well-defined block lengths. 1:1 Also mild reaction conditions 
together with well-delineated reactivity minimized defects in the polymer microstructure. 
The Mo catalysts are still sufficiently oxophilic to react with aldehydes and ketones in a 
Wittig fashion, and this particular reaction has been capitalized upon for stoichiometric 
cyclizations of unsaturated carbonyl compounds 16 and for end-capping of living ROMP 
reactions as shown in Figure 1-6. If functionalized benzaldehydes are used as end- 



capping reagents in ROMP reactions, polymers with narrow polydispersity and useful 

1 7 

end-groups can be obtained. 

[Mo ]=0 
[Mo ]=0 

Figure 1-6. Reaction of Mo alkylidenes with carbonyl-containing substrates: Ring- 
closing olefination 16 and end-capping of living ROMP 17 . 

As mentioned previously, these catalyst systems would soon be overshadowed by 
well-defined ruthenium catalyst systems which are superior in many aspects. Schrock's 
W and Mo alkylidines, like their ill-defined predecessors based on early transition metals, 
still displayed moderate functional group tolerance. They are easily deactivated by 
compounds containing active hydrogens (acids, alcohols, H 2 0, etc) and oxygen. Their 
widespread usage is also limited by thermal instability during storage and expense of 
preparation. 




Ruthenium Metathesis Catalysts 

The promise of ruthenium systems as robust, functional group tolerant catalysts 
was foreshadowed during the first stages of development of the metathesis reaction, 
where R11CI3 was shown to be effective for the ROMP of cyclobutenes and norbornene in 
aqueous and alcoholic environments. 2 Polynorbornene has been produced commercially 
for over 15 years in a process utilizing RuCls in alcohol solvent and the reaction is even 



12 



promoted by the presence of oxygen. However, utility is limited to ROMP of high strain 
ring systems. 

In 1992, reports began to emerge of the in situ activation of ruthenium complexes 
for ROMP by addition of diazo compounds. 18 ' 20 Activation occurred presumably via the 
formation of ruthenium carbenes, the presence of which was supported by NMR studies. 
Initial studies 19 with diruthenium (II) tetracarboxylates activated by ethyl diazoacetate 
were limited to high strain olefins and suffered from competing cyclopropanation of the 
olefin substrates. This observation finds precedent in extensive studies of transition 
metal catalyzed cyclopropanation, which is most effective with diazocarbonyl 
compounds. 21 Ethyl diazoacetate may be used to activate Ru(H 2 0) 6 (OTs) 2 for the ROMP 
of cyclooctene, which also indicates that the difficulty in polymerizing the low strain 
monomers with Ru(II) complexes lies not in propagation but in initiation. 20 




Figure 1-7. Functionalized cyclooctenes polymerized via ROMP catalyzed by 
[Ru(p-cymene) 2 Cl 2 ] 2 + PCy 3 + trimethylsilyldiazomethane. 

A variety of ruthenium complexes have been studied, with relative success 
reported 19 for the type [Ru(arene) 2 Cl 2 ] 2 activated by trimethylsilyldiazomethane (TMSD) 
both with and without added trialkyl phosphines. These multicomponent systems have 
been applied to the ROMP of various functionalized norbornenes, oxanorbornenes, and 
cyclooctenes. Each of the functionalized cycloctenes shown in Figure 1-7 were 



13 



polymerized to high molecular weight (M n = 5-12 X 10 4 g/mol, PDI = 1.4 - 2.2) in 
good to moderate yields. 



While these multicomponent ruthenium systems enabled the metathesis of 
substrates with a variety of functionality not previously possible with other transition 
metals, concurrent developments in the synthesis of related preformed ruthenium 
carbenes would come to virtually dominate the attention of the metathesis community. 8 ’ 22 
The first preformed ruthenium species which were highly active for acyclic metathesis 
were reported" 3 in 1993 and an improved version 24 , IA, followed shortly thereafter 
(Figure 1-8). While other versions were prepared with differing phosphine and carbene 
ligands, JL4 is preferred as its synthesis is extremely fascile, utilizing diazo compounds 
similar to the process described earlier for in situ activation of ruthenium complexes. In 
addition, L3 has been shown to be a slower initiator due to the greater stability of the 
more highly conjugated vinylidene relative to the benzylidene. 



Figure 1-8. Discrete ruthenium carbene olefin metathesis catalysts. 

While it is certain that the continually growing knowledge base will lead to ever- 



Discrete Ruthenium Carbene Complexes 





PCy 3 



PCy 3 



L4 



increasingly effective catalysts, the explosion of applications made possible by this series 
of Ru carbenes makes them seem the current zenith. Their technical and synthetic value 



14 



has sparked a flurry of research efforts aimed at more facile catalyst preparations, which 
now approach the degree of difficulty of sophomore organic chemistry laboratory 
exercises. 

Some of the outstanding features of the catalysts are as follows. They can be 
employed in both aqueous' 3 " 6 and alcoholic solutions 23 . The unprecedented stability of 
the catalysts to adventitious moisture and other typical poisons enables their practical use 
without highly specialized equipment, and they may be stored for long periods of time in 
air without decomposition. The practical organic chemist has only to sparge a simple 
reaction vessel, the substrate and any solvents with nitrogen before reaction. The catalyst 
1-3 has been used for RCM of dienes containing aggressive functionality such as OH, 
C0 2 H, aldehyde and amide groups 27 (Figure 1-9). These transforms are impossible with 
early transition metal catalysts due to intolerance of the catalysts to these functionalities. 




Figure 1-9. RCM of dienes with aggressive functionality. 



It should be noted however that the success of the reactions in Figure 1-9 does not 
show that the complex is impervious to these functionalities, but rather the catalyst is 
sufficiently tolerant to allow good yields of product when the kinetics of the desired 
reaction (e.g. RCM) are sufficiently rapid to compete with any deleterious side reactions. 
The success of a metathesis reaction in the presence of these functionalities therefore 



15 



depends on the substrate under study. For instance, while hydroxy-functionalized 
substrates may be employed successfully in ROMP and RCM schemes, dimerization of 
acyclic co-alkenol alcohols gives poor yields and some isomerization of the alcohol 
substrates to aldehydes due to slower kinetics of acyclic chemistry. 28 

While catalysts L3 and 1A are much more tolerant to oxygenated groups, they are 
less effective than early transition metal complexes for the metathesis of sulfur 29 and 
phosphorous 20 containing substrates. This can be explained in terms of hard-soft theory 
where the softer S and P atoms more effectively complex the soft ruthenium center than 
the harder earlier transition metal centers. Catalyst L4 has, however, been proven 
effective for the first ever RCM reaction of phosphonate compounds as shown in Figure 
1 - 10 . 




3 mole % ±A 



MeCI2, reflux, 30 min 





3 mole % ±A 



MeCI2, reflux, 30 min 




99% 



75% 



Figure 1-10. RCM of dienes with phosphonate functionality catalyzed by L4. 



Complexes JL3 and L4 have promoted the affirmation of the metathesis reaction 
by the synthetic community, where growing confidence is indicated by the willingness to 
incorporate the reaction as a transform in advanced stages of multi-step synthetic 
sequences. 2 The most common application is ring closure, noteworthy examples 22 ' 34 of 
which are shown in Figure 1-11. 



16 




Figure 1-11. Applications of RCM catalyzed by L4 in multistep synthesis. 

Acyclic Diene Metathesis (ADMET) Polymerization 

ADMET polymerization 35,36 is a step-condensation polymerization where dienes 
are connected via metathesis in a stepwise fashion to produce polyenes as depicted in 
Figure 1-12. Monomer is consumed in the early stages of the reaction to produce dimer, 
trimer, tetramer, etc with the evolution of the condensate, a small monoene. Each species 
in the system can react with any other species resulting in a distribution of homologous 
products. As each step is reversible, the condensate may also react with homologues 
splitting the chains to smaller species and in a closed system, this reverse reaction 
restrains the equilibrium product mixture to relatively low molecular weight species. In 
the typical case involving a,o>dienes, the ethylene liberated with each connection is 
easily removed under reduced pressure to drive the equilibrium to the right yielding a 
product distribution with progressively higher average molecular weight. The 
equilibrium nature of the reaction permits the reaction of all species in the system, which 
randomizes the product distribution to the most probable distribution described by the 
polydispersity index, PDI = M w /m„ = 2.0. 



17 






Dimer 

+ C 2 H 4 



Dimer 

Trimer 

Tetramer 

Oligomers 

C 2 H 4 




^ r k 



m 



+ C 2 H 4 



Figure 1-12. The ADMET step/condensation polymerization of an a, co-diene. 



For typical step/condensation polymerizations, the relationship between the 
number average degree of polymerization (DP„), the conversion (p), and the 
stoichiometric imbalance (r) is described by Equations 1-1. In this relation, r is the molar 
ratio of groups "A" to groups "B" which condense to form the polymer linkages. In the 
limit of 100% conversion (p = 1.000), Equation 1-la reduces to 1-1 b. It can be seen from 
Equation 1-lb that any decrease in “r” from unity results in a decrease in the maximum 
obtainable average molecular weight. 



DP. = 



1 + r 
1 + r - 2 rp 



Eqn (1-1 a) 



DP„ = 1 + r Eqn ( 1 - lb) 



- r 



Attainment of high molecular weight requires near perfect stoichiometric balance 
of the reactive groups 37 , a concept that must be modified to apply to ADMET. The 
ADMET polymerizations to date are somewhat unique amongst condensation 
polymerizations in that the functional groups, which condense to form the linkages in the 
polymer backbone, are not chemically different. They do not fit into the schemes of A-A 
+ B-B or A-B type polymerizations where different functional groups “A” and “B” react, 
common examples of which are polyesterifications schematically represented in Figure 1- 



13. In the second case in which a hydroxy-functional carboxylic acid condenses to a 
polyester, r = 1.000 due to the fact that both the “A” and “B” groups are in a single 
monomer. An exception to these polycondensation schemes similar to ADMET is the 
hydrolytic step-polymerization of dichlorosilanes 38 in which Si-Cl groups are converted 
to silanol groups which then condense with each other to form siloxanes and the 
condensate, water. Of course in this particular type of polymerization, the SiOH groups 
may also react with any residual SiCl groups to form the same polymer. 



O O 

X X 

HO R OH 
AA monomer 



HO-R'-OH 
BB monomer 



HO 



O O 

^ R X 



O-R'-O+H 
n 



HpO 



O 

ho A r' oh 

AB monomer 



HotV 0 ^ 



HoO 



R 

Cl— Si-Cl + H 2 0 
R 

AA monomer 



-HCI 



R 

HO-Si-OH 

R 

A TV monomer 



— HO-j-Si-O-jH + H 2 0 



R 



Figure 1-13. Comparison of classical condensation polymerizations. 



Therefore, perfect stoichiometric balance in the case of ADMET simply means 
that vinyl groups terminate all linear molecules in the system at both ends throughout the 

course of the reaction. In this light, the relationship between DP n and conversion is 
more appropriately described by Equation l-2a, often referred to as the Carothers 
equation, where / avg is the average functionality of all linear species. For perfectly 
difunctional species ,/ avg = 2.0, and the more commonly presented Equation 1 -2b results. 



19 



o 1 

DP = Eqn (l-2a) DP n = - Eqn(l-2b) 

" 2-p/„„ 1-P 

Any loss of vinyl groups due to competing chemistry decreases / avg and lowers the 
maximum achievable molecular weight. This critical requirement tor successful 
ADMET polymerization deferred its development until the arrival of the well-defined 
Schrock W(VI) and Mo(VI) carbenes (cf Figure 1-5). Initial attempts to polymerize 
dienes as early as 1970 were destined to failure as the classical catalyst systems of that 
time period participated in side reactions which decreased f av g.~ The reaction was 
revisited in 1987 by Wagener and coworkers with the realization that the key to success 
was the utilization of single-site catalysts. 41 The a,o>-dienes 1,5-hexadiene and 1,9- 
decadiene were condensed to high molecular weight linear polymers utilizing W catalyst 
1.1c. 35 ’ 41 The reaction has since been demonstrated with other Schrock alkylidines, most 
commonly catalyst 1.2d , with homogeneous W(VI) classical catalysts 4 " activated by alkyl 
tin cocatalysts, and most recently with ruthenium catalyst L4. 43 

Successful demonstration of the ADMET reaction was followed by systematic 
studies designed to delineate monomer structure reactivity relationships, the bulk of 
which has centered on Mo catalyst L2d. It became clear that the slower kinetics of the 
ADMET reaction relative to faster processes such as ROMP and RCM placed greater 
restraints on the type and placement of functionality in substrates. For dienes with Lewis 
basic functionalities appropriately placed within the structure, reaction may be 
prohibitively slow or even completely halted. This effect has been attributed to 
intramolecular coordination of the Lewis basic functionality, schematically represented in 
Figure 1-14, and finds precedent in similar ROMP systems. 44 The degree to which the 



20 



ADMET reaction is hindered may depend on the stability and persistence of this chelate. 
This effect, termed the negative neighboring group effect, has been noted tor a variety of 
functionality but can generally be overcome through monomer design by separating the 
Lewis basic group "X" from the olefin groups by a suitable number of methylene spacers. 




Figure 1-14. The proposed negative neighboring group effect. 



With the knowledge of the keys for successful ADMET chemistry in hand, 
polymers with a broad range of functionality have been prepared including ether, ketone, 
ester, amine, thioether, conjugated olefins, silane, siloxane, and metal-containing 
groups . 36 Just as for other catalyzed polymerizations, the scope of the ADMET reaction 
continues to evolve as a result of new developments in catalyst design. Most recently, 
the arrival of the extremely robust ruthenium catalyst L4 has allowed the direct 
preparation of hydroxy-functional polymers. With the realization of classical systems 
activated by main-group metals with diminished Lewis acidity, ADMET polymers can 
now be prepared without the necessity for preformed metal carbenes . 42,44 A particularly 
novel twist to this concept was reported where tin-containing dienes were prepared which 
function both as monomer and as cocatalysts . 44 

Summary of Results Presented in this Dissertation 

The ADMET reaction and the polymers produced thereof share many 
characteristics with other common step/condensation polymerization systems. High 



molecular weight polymers can only be obtained when the equilibrium between terminal 



21 



and internal olefins can be driven to high conversion, a direct consequence of the relation 



between DP n and conversion described by Equation 1 -2b. A further consequence is that 
the reverse reaction may also be driven in high conversion, where high molecular weight 
polyalkenamers may be depolymerized to low molecular weight species. The 

relationship between / avg and DP„ (Equation 1 -2a) may also be capitalized upon to 
control molecular weight through the incorporation of monoenes, or chain limiters, in the 
forward ADMET reaction. End-functionalized oligomers and polymers may be prepared 
by either the forward or reverse reaction if functionalized chain limiters are employed 
(Figure 1-15). 

Chapter 3 describes results obtained concerning the depolymerization of high 
molecular weight polybutadiene to prepare low molecular weight species and end- 
functionalized polymers utilizing complex E4. The unprecedented stability ot IA in the 
presence of ethylene implicated it as a good candidate to catalyze the exact microscopic 
reverse of the ADMET reaction, whereby polyalkenamers are depolymerized with 
ethylene as chain limiter. This assumption was born out as much higher conversions in 
this reaction were achieved than previously possible with well-defined W and Mo 
alkylidenes. In addition, the unique ability of E4 to initiate depolymerization of 
polybutadiene in the solid state was observed and exploited to prepare end-functionalized 
polybutadiene. 




FT 











[M]= 




+ C2H4 



Figure 1-15. ADMET depolymerization and polymerization with chain limiters. 



22 



With judicious choice of catalyst system with reactivity appropriate to a given 
type of diene monomer, linear polymers with highly pure microstructure are obtained. 
The only variation in structural homogeneity is the olefinic cis/trans ratio, which is 
typically >70% trans. This backbone unsaturation provides sites for post-polymerization 
reaction by which novel polymers may be prepared that are difficult or impossible to 
obtain by direct means. One such reaction, which has generated both commercial' and 
high academic interest is hydrogenation whereby the carbon-carbon double bonds are 
converted to single bonds. It was discovered in our laboratories that addition of silica gel 
to a completed metathesis reaction catalyzed by ruthenium complex E4 followed by 
exposure to hydrogen gas constitutes an efficient method for catalytic heterogeneous 
hydrogenation of the metathesis products. The development of this process and its 
application to the indirect synthesis of saturated polymers, including telechelic 
polyethylene, are described in Chapter 4. 

The highly pure microstructure of ADMET polymers appoints this polymerization 
route as a means to prepare well-defined models for polymers with less pure constitution. 
ADMET polymerization coupled with quantitative olefin hydrogenation has been 
shown 46 ' 51 to be particularly well suited for the preparation of models for polyethylene 
and ethylene copolymers as shown in Figure 1-16. Any functionality (R), within the 
restraints imposed by catalyst compatibility, may be incorporated at regular intervals 
along a polymer which is in essence a perfectly linear terpolymer of ethylene, butadiene 
and a monomer containing the functionality. The number of methylene spacers (“n” in 
Figure 1-16) incorporated in the monomer synthesis precisely controls the frequency of 
the desired functionality in the polymer backbone. Saturation of the backbone double 



23 



bonds produces model polymers for ethylene copolymers. In Chapter 5, the tandem 
ADMET/ hydrogenation process described above is utilized to prepare a range of model 
ethylene copolymers including copolymers with vinyl acetate, acrylates, styrene, and 
vinyl chloride. The synthesis of the necessary dienes and the effect of the chemical 
nature of the incorporated substituents on thermal properties of the polymers are also 
disclosed. 




ADMET 





n 



n 



n n J x 



2n+2 



Figure 1-16. Indirect synthesis of model polymers via ADMET/hydrogenation. 



CHAPTER 2 
EXPERIMENTAL 

Instrumentation and Analysis 

'H NMR 300 MHz and l3 C NMR 75 MHz were recorded on a Gemini Series 
NMR Superconducting Spectrometer System or a Varian VXR-300. NMR spectra were 
recorded in chloroform-J (CDCR), benzene-ek (C6D 6 ), toluene-dg, and bromobenzene-ds. 
Resonances are referenced to residual protio solvent and reported in § units downfield 
from TMS at 0.0 ppm. Heteronuclear decoupled quantitative l3 C NMR spectra were 
acquired over 8-12 hours with a pulse delay of 20s. Gas chromatographic measurement 
of purity and product distributions were conducted with a Hewlett-Packard HP5880A gas 
chromatograph with cross-linked methyl silicone gum capillary column and flame 
ionization detector. Low and high-resolution mass spectrometry data were recorded on a 
Finnigan MAT 95Q or Finnigan MAT GCQ Gas Chromatograph/Mass Spectrometer 
utilizing either electron ionization or chemical ionization. Elemental analyses were 
conducted by Atlantic Microlab, Inc., Norcross, GA, where each recorded value is the 
average of duplicate analyses. 

Differential Scanning Calorimetric (DSC) and Thermogravimetric Analysis 
(TGA) data were obtained with a Perkin Elmer 7 Series Thermal Analysis System. DSC 
samples (5-10mg) were analyzed with either liquid nitrogen or ice as coolant and under a 
Helium flow rate of 25 ml/min. All samples were pre-dried at 40-50°C under reduced 



24 



25 



pressure (<0. 1 mm Hg) for 8-12 hours. The instrument was calibrated in the appropriate 
temperature ranges with cyclohexane and/or ultra-high purity water (subambient 
measurements), or Indium. Samples were heated to at least 40°C above the melting point 
to erase thermal history and then scanned at 5°C/min, cooling cycle first. TGA samples 
(5-10 mg) were heated at 20°C/min under N 2 or air with a flow rate of 30 ml/min, until 
complete combustion. 

Gel Permeation Chromatography (GPC) was performed using a Waters 
Associates Liquid Chromatograph U6K equipped with tandem ABI Spectraflow 757 UV 
absorbance detector and a Perkin-Elmer LC-25 RI detector. The GPC solvent delivery 
system was configured for elution with either HPLC grade CDCI3 or THF through an 
Ultrastyragel linear mixed bed column (CHC1 3 ) or through 5 x 10 3 A and 5 x 10 4 A 
Phenogel columns in series (THF). Solvents were sonicated immediately prior to use and 
delivered at a rate of 1.0 ml/min. Retention times were calibrated against narrow (PDI < 
1.07) polystyrene standards (Scientific Polymer Products, Inc) with M p = 1900, 7700, 
12000, 30000, 59000, 79000, and 139400 g/mol. Polymer samples were prepared in 
HPLC grade THF or CHCL3 (-1% w/w) and slowly passed through a 50p syringe filter 
prior to injection. Molecular weight measurements by vapor pressure osmometry (VPO) 
were conducted utilizing a Jupiter Instruments Model 233 with toluene as solvent at 
50°C. The instrument constant was determined by calibration (R 2 > 0.98) with 
recrystallized sucrose octaacetate and verified against narrow PDI polystyrene GPC 
standards (error in measured M n < 5%). 



26 



Materials and Techniques 

The olefin metathesis catalysts Ru(=CHPh)(PR 3 ) 2 CI 2 , R = cyclohexyl or phenyl, 
and Mo[=CHC(CH 3 ) 2 Ph](N- 2 , 6 -C 6 H 3 -/-Pr 2 )[OCCH 3 (CF 3) 2 ] 2 were prepared according to 
published methods. These catalysts were graphically depicted in the appropriate sections 

of Chapter One. Polybutadiene (MW = 2 x 10 ^, 98% cis- 1,4-repeat unit, Aldrich # 
18,374-4) was purified by precipitation from methylene chloride into methanol, then 
degassed and azeotropically dried by dissolution in dry toluene followed by removal of 
toluene under reduced pressure. It was then stored in an amber bottle inside an argon- 
purged dry box. Polymer grade ethylene (99.9% min.) was obtained from Matheson and 
purified by passage through activated molecular sieves followed by two Alltech 8115 
High Pressure Oxy-Traps in series. Silica gel-60 (Selecto Scientific cat # 162824, 
particle size 32-63p) used in hydrogenation reactions was sonicated twice in reagent 
grade CHCI 3 , dried under reduced pressure at 80°C for 24 hours, and stored in an argon- 
purged dry box until use. 1-Octene, trans-7-tetradecene, 1 ,9-decadiene, HMPA, 
diisopropyl amine, bromobenzene, iodomethane, ethyl iodide, and CDCI3 were dried over 
CaH 2 and vacuum transferred. Olefin substrates were further degassed by three 
consecutive freeze/pump/thaw cycles. Pyridine and acetyl chloride were stored over 
CaH 2 and manipulated under argon. Carbon tetrachloride and methyl formate were 
distilled under argon from P 2 Os, with the latter subsequently distilled from CaH 2 . 
Acetone (reagent grade) was stirred over anhydrous CaS0 4 (lOOg/L) overnight, decanted 
to fresh CaS0 4 (50g/L), and distilled immediately prior to use. Acetic anhydride was 
distilled from anhydrous Na 2 C0 3 . Triphenylphosphine (Aldrich, 99%) was dried in the 
molten state (85°C) with stirring under reduced pressure (5 mm Hg) for 8-12 hours then 



27 



stored in an argon purged dry box until use. Diethyl ether and tetrahydrofuran were 
distilled under argon from Na/K alloy with the respective metal benzophenone ketyls and 
stored over 5 A molecular sieves. Toluene was washed with cold cone. H2SO4, distilled 
from Na/K alloy, and stored over 5 A molecular sieves. All other materials were used as 
received. 



Synthesis and Characterization 
ADMET Depolymerizations 

Ethenolysis of cis-1 ,4-Polybutadiene . Depolymerizations were conducted in 
either glass pressure tubes or glass-lined stainless steel bombs appropriate to the applied 
pressure. Inside an argon purged dry box polybutadiene 2bl (500mg, 9.26mmol repeat 
unit, 400 eq) cut in small chunks and catalyst L4 (19 mg, 1 eq) were combined inside the 
appropriate pressure vessel together with 3 ml CDC1 3 . The vessel was sealed, removed to 
a gas manifold, and opened to a constant dynamic pressure of C 2 H 4 with vigorous stirring 
at room temperature for 48 hours. The reaction vessel was then chilled to avoid loss of 
volatile products, and the ethylene pressure was slowly vented to the atmosphere. The 
homogeneous solutions were then immediately exposed to air to decompose any 
remaining active catalyst. See text for data analysis. 

Synthesis of Ester-terminated Polybutadiene(3.5 and 3.6) , The chain limiter (3.4 ) 
was prepared immediately before use by dimerization of ethyl 10-undecenoate (3.3), 
Inside an argon-purged dry box, 3 3 (0.788g, 3.71 mmol, 400 eq) and 1.4 (7.6 mg, 
9.25pmol, 1 eq) were combined in a high vacuum flask. The flask was removed to a high 
vacuum manifold and the contents stirred under reduced pressure at room temperature for 



28 



3 hours after which time the reaction was heated to 40°C for an additional 24 hours. 
Inside the dry box, portions of this dimer/catalyst mixture were combined with 
polybutadiene 3J_ (500 mg, 9.24 mmol repeat unit) cut in small chunks in glass vials, in 
two different proportions to generate end-functionalized polymers with predicted DP = 
20(3.5) and 40(3.6), Stoichiometry: Polymer .^5 (183 mg dimer 3A 46.2pmol), 

Polymer 3.6(91 .7 mg dimer 3^4, 23.1pmol). Magnetic stir bars were inserted and the 
vials were sealed and removed from the dry box. The mixtures were occasionally 

t 

agitated until viscosity was sufficiently reduced that they could be stirred with a magnetic 
stir plate(12 hours). After stirring for 24 hours, reaction was quenched by placing under 
reduced pressure then opening to air for 2 hours with stirring. Each polymer was 
dissolved in toluene and stirred with alumina and filtered to remove catalyst residue, and 
the toluene was evaporated under reduced pressure to yield clear colorless viscous oils. 
Yield: X5, 592.8 mg, 86.7%; 3,6, 467.3 mg, 78.9%. *H- NMR 5 (ppm) 3^ 5.39(m, 
42H); 4. 10(q, 4H); 2.26(t, 4H); 2.03(br, 94H); 1.59(t, 6H); 1.25(br)+1.23(t)=30H. I3 C- 
NMR 5 173.82, 130.27, 130.07, 130.04, 129.95, 129.92, 129.53, 129.37, 60.07, 34.32, 
32.65, 32.53, 29.55, 29.26, 29.18, 29.09, 29.03, 27.35, 24.92, 14.21. 'H- NMR 8 (ppm) 
M 5.39(m, 80H); 4.10(q, 4H); 2.27(t, 4H); 2.03(br, 180H); 1.59(t, 6H); 

1.26(br)+1.23(t)=30H. 

Metathesis and in situ Hydrogenation 

Model Study of Hydrogenation using frans-7-Tetradecene . Inside an argon- 
purged dry box, trans- 7-Tetradecene (1.0 g, 5.09mmol) and 3 (5.2 mg, 0.13 mol%) were 
combined in a pressure tube containing a magnetic stir bar and stirred 1 hour to allow 



29 



complete dissolution of catalyst and alkylidine exchange, during which time the purple 
heterogeneous mixture became a clear maroon solution. 520 mg silica was added, the 
tube was immediately sealed, removed to a hydrogen manifold and exposed to a constant 
120 psig H 2 at room temperature. Adsorption of the catalyst residue onto the silica 
surface was indicated by rapid decolorization of the substrate/catalyst solution. After 30 
minutes, the reaction was quenched by releasing pressure to the atmosphere and filtration 
to remove catalyst residue and silica, yielding a clear, colorless liquid. Extent of 
reduction was estimated by ‘H NMR(CDC1 3 ) comparing the resonances at 5.4 ppm 
(multiplet, internal olefin) to that at 0.88 ppm (triplet, CH 3 group). Experiments at each 
pressure were conducted in triplicate and the average conversions calculated. See text for 
tabulated results. 

Synthesis of 1,18-Bisacetoxyoctadecane (4.2) . 9-Decenyl acetate ( 4.1 ) (l.Og, 
5.04mmol) and catalyst JL4 (5.2 mg, 52pmol) were combined in a high vacuum flask 
containing a magnetic stir bar. The vessel was sealed, removed to a schlenk line, and 
opened to a slight positive pressure of argon. It was heated to 60°C with vigorous stirring 
for 4 hours, after which reaction was continued for an additional 2 hours at ~0.2 mm Hg. 
The reaction mixture was taken up in 3 ml toluene, combined with 600 mg silica in a 
pressure tube, and the whole was exposed to 120 psig H 2 at 90°C for 5 hours. Filtration 
and evaporation of solvent yielded the product as a white powder (0.90 g, 96% yield). 
LRMS: 372 (m+2). Elemental anal, calcd for C 22 H 42 0 4 : C, 71.31; H, 11.42. Found: C, 
7 1 .37; H. 1 1 .50. 'H- NMR 8 (ppm) 4.02(t, 4H); 2.02(s, 6H); 1 ,59(m, 4H); 1 ,23(br, 28H). 
13 C-NMR 5 171.24, 64.65, 29.65, 29.56, 29.52, 29.24, 28.59, 21.02. 



30 



Synthesis of Telechelic Polyethylene (4.6). Inside an argon-purged dry box, 1 ,9- 
decadiene (3.0g, 21.7mmol, 1 eq), 9-decenyl acetate ( 4.1) (0.861g, 4.34 mmol, 0.2 eq) 
and catalyst L4 (49mg, 0.06mmol) were combined for a total of 800: 1 olefin-to-catalyst 
ratio. The reaction vessel was sealed, removed to a schlenk line, opened to slight positive 
pressure of argon and heated to 45°C. Over a period of 6 hours, pressure was slowly 
decreased (intermittant vacuum) then left open to dynamic vacuum of 10' 2 mmHg, while 
slowly increasing temperature to 65°C. After 24 hours, the reaction vessel was sealed 
and returned to the dry box. 1.5g crude polymer was initimately mixed with 2.5 g silica 
in a pressure tube, the tube was sealed and removed to a hydrogen manifold. The 
contents were exposed to 120 psig H 2 at 90°C for 30 minutes. 15 mL toluene was then 
added via syringe under H 2 purge. The vessel was then exposed to 120 psig H 2 at 90°C 
with vigorous stirring for 5 hours. Reaction was quenched by release of pressure and 
filtered while still hot through a heated course glass frit to remove catalyst residue and 
silica. Toluene was removed under reduced pressure to yield 1.45 g white solid. 
Elemental anal, calcd for [Cio 2 H 202 0 4 ]: C, 82.07; H, 13.64. Found: C, 82.03; H, 13.70. 
'H- NMR 6 (ppm) (90°C, toluene-d 8 ) 3.98(t, 4H); 1.73(s, 6H); 1.50(m, 4H); 1.30(br, 
220H). M n = 1.5 x 10 3 (PDI = 1.9, high temp. GPC vs PE), 1.7 x 10 3 (‘H NMR end- 
group analysis). 

Synthesis of saturated polyester (4.9). Monomer 47 was dried by stirring at 50°C 
for 8-12 hours under reduced pressure (< 0.1 mm Hg) in an appropriate storage flask 
equipped with a high-vacuum valve. The atmosphere in the flask was deoxygenated by 
three cycles of evacuation under reduced pressure and replacement with argon, and the 
monomer was stored under argon. Inside an argon-purged dry box 47 (3.00 g, 9.31 



31 



mmol, 600 eq) and catalyst L4 ( 12.7 mg, 15.5 jimol, 1 eq) were combined in a small 
glass tube with a 24/40 ground glass joint and a magnetic stir bar. The flask was fitted 
with a Kontes® high vacumm valve, sealed, and attached to a high vacuum scklenk line 
via ground glass joint sealed with silicone grease. Evolution of ethylene was observed 
within 5 minutes as evidenced by formation of small bubbles which rapidly increased as 
the stirrer was set in motion and a 45°C oil bath was raised. The pressure was slowly 
decreased by applying intermittent vacuum eventually reaching a dynamic vacuum of 10' 
2 - 10' 3 mm Hg while the temperature was slowly increased to 70°C. After 48 hours, the 
vessel was sealed and returned to the dry box. Crude yield: 2.70g, 97.8%. A l.OOg 
chunk of the crude polymer/catalyst residue mixture was combined with 500 mg silica 
gel in a glass pressure tube containing a magnetic stir bar. 10 mL dry degassed toluene 
was added and the tube was rapidly sealed with a brass valve fitted via a threaded 
Teflon™ flange, removed to a gas manifold where it was charged with H 2 (120 psig, 
dynamic pressure) and heated to 90°C with vigorous stirring. Note: Hydrogen pressure is 
applied prior to dissolution of the polymer chunk. After 24 hours, pressure was released 
to the atmosphere and the heterogeneous mixture was filtered while still hot through a 
course glass frit. The reddish-brown silica was washed with hot toluene and the 
washings were combined with the clear, colorless filtrate. The polymer precipitated on 
cooling to below 60°C. The toluene was evaporated under reduced pressure to yield 
as a white solid. Yield: 0.96g, 95% based on lg 4JI. Elem. Anal. Calcd. for [Q9H36O2]: 
C, 76.97; H, 12.24. Found: C, 77.06; H, 12.22. . 'H- NMR 5 (ppm) (90°C toluene-d 8 ) 
4.02(t, 2H); 2. 1 7(t, 2H); 1.56(m,4H); 1 ,28(br, 28H). n C-NMR 5 177.01, 139.12, 114.06, 



32 



51.21, 45.67, 33.76, 32.47, 29.49, 29.35, 29.05, 28.86, 27.42. M n = 1.02 x 10 4 (PDI = 
2.3, high temp. GPC vs PE). 

Synthesis of Functionalized Diene Monomers for Model Ethylene Copolymers 

Synthesis of extended chain oalkenyl bromides and carboxylic acids 

Synthesis of 9-decenyl bromide(5.1), 9-Decenol (lOg, 64 mmol, 1 eq) and CBr 4 
(25.47 g, 76.8 mmol, 1.2 eq) were dissolved in 100 ml anhydrous ether in an argon-filled 
schlenk flask. A small amount of freshly ground CaH 2 was added and stirred for one 
hour. Solid PPh^ (20.14 g, 76.8 mmol, 1.2 eq) was added under argon in ~4 g increments 
over 20 minutes. After 20 hours the reaction mixture was diluted with 50 mL pentane, 
filtered, and concentrated under reduced pressure. 1 1.2 g (80% yield) pure 5J_ (100% 
GC) was obtained after distillation from CaPL (bp = 96-98°C/10 mmHg). 'H- NMR 8 
(ppm) 5.78(m, 1H); 4.94(m, 2H); 3.38(t, 2H); 2.02(q, 2H); 1.83(m, 2H); 

1.37(m)+1.28(br) = 10H. 13 C-NMR 5 139.06, 1 14.14, 33.95, 33.73, 32.78, 29.24, 28.97, 
28.83,28.68,28.12. 

Synthesis of 10-undecenyl bromide(5.2). The alkenyl bromide was prepared in 
two steps from the respective alcohol via the tosylate. 10-Undecenol (82. 5g, 0.484 mol, 
1 eq) was dissolved in 500 mL CHCI3 in a 1 L schlenk flask. The solution was placed 
under argon and chilled to 0°C. Pyridine (78 mL, 2 eq) was decanted into the flask from 
CaPL. Solid TsCl (138.4 g, 0.726 mol, 1.5 eq) was added under argon in increments over 
- 1.5 hours. After 8 hours the reaction mixture was washed with water, sat’d NaHCCL, 
3N HC1, and brine. The organic layer was dried with Na 2 S 04 and concentrated to 154.7 



33 



g pale yellow oil which was converted to the bromide without further purification. The 
crude tosylate was dissolved in 900 mL dry acetone under argon. LiBr anh (126 g, 3 eq) 
was added in one portion, the flask was fitted with a water-cooled condenser and the 
mixture was refluxed for 12 hours. The mixture was filtered, diluted with 300 mL 
pentane, filtered a second time and concentrated. The crude product was taken up in 300 
mL pentane and washed with water, sat’d NaHCCL, and brine. The organic layer was 
dried with Na 2 S0 4 , concentrated, and the bromide (97% GC) was vacuum transferred 
from CaH 2 . Yield: 100.8 g, 89 % for two steps. 'H- NMR 8 (ppm) 5.80(m, 1H); 4.95(m, 
2H); 3.40(t, 2H); 2.04(q, 2H); 1.85(m, 2H); 1 ,38(m)+l ,29(br) = 12H. 13 C-NMR 8 
139.12, 1 14.12, 33.96, 33.76, 32.81, 29.33, 29.04, 28.88, 28.71, 28.13. 

Synthesis of 1 1-dodecenoic acid(5.3). 10-Undecenyl bromide ( 5 . 2 ) (10 g, 42.9 
mmol, 1.0 eq) was added slowly via syringe to freshly scoured Mg turnings (1.053 g, 
43.3 mmol, 1.01 eq) in 50 mL dry THF with rapid stirring under argon. After stirring 2 
hours at 45°C, the resulting Grignard solution was rapidly poured into a large beaker 
containing 19g (—10 eq) finely crushed C0 2 (s). The mixture was stirred with 50 mL 3N 
HC1 and 80 mL pentane, after which the aqueous layer was discarded. The organic layer 
was washed with 3N HC1 and water, dried over Na 2 S0 4 , and concentrated to 8. 1 6g pale 
yellow oil. The oil was further purified by distillation under reduced pressure (bp = 1 10- 
1 15°C/0<P<0. 1 mmHg) to yield 6.7 1 g colorless oil (79% yield). 'H-NMR 8 (ppm) 
1 1.76(br, 1H); 5.80(m, 1H); 4.93(m, 2H); 2.32(t, 2H); 2.01(q, 2H); 1.60(m, 2H); 1.26(br, 
12H). 13 C-NMR 8 180.62, 139.13, 114.09, 34.11, 33.78, 29.67, 29.38, 29.35, 29.20, 



29.07,29.01,28.89, 24.63. 



34 



Synthesis of 1 3-tetradecenoic acid(5.5). 10-Undecenyl bromide (5;2) (22.25 g, 
95.4 mmol, 1.0 eq) was added slowly via syringe to freshly scoured Mg turnings (2.34 g, 
96.2 mmol, 1.01 eq) in 100 mL dry THF with rapid stirring under argon. After stirring 2 
hours at 45°C, the resulting Grignard solution was cooled to RT and transferred via 
cannula to LiCl (anh) (0.161 g, 1.6 mmol) in 100 ml dry THF at 0°C. (3-Propiolactone 
(Acros, unopened bottle) (5 mL, 79.5 mmol, 0.83 eq) was added dropwise via syringe 
maintaining the temperature below 5°C. After one hour, reaction was quenched with 50 
mL 3N HC1 and then concentrated. The crude product was taken up in 200 ml pentane 
and washed with 3N HC1, water, and brine. The organic layer was dried over Na 2 S0 4 
and concentrated to 20.03g (over theory) pale yellow oil, which solidified on standing. 
The product can be further purified by distillation under reduced pressure (bp s 
1 26°C/0<P<0. 1 mmHg) or by flash chromatography (20%EtOAc/Pentane) rendering 85 
and 91% isolated yields, respectively of a white waxy solid. 'H- NMR 5 (ppm) 1 1.22(br, 
1H); 5.79(m, 1H); 4.92(m, 2H); 2.32(t, 2H); 2.02(q, 2H); 1.6 l(m, 2H); 1.24(br, 16H). 
n C-NMR 8 180.64, 139.19, 114.07, 34.12, 33.80, 29.55, 29.47, 29.41, 29.22, 29.13, 
29.03, 28.93, 24.64. 

Synthesis of 11-dodecenyl bromide(5.4). Following the LAH reduction of 11- 
dodecenoic acid ( 5.3 ) to the respective alcohol, fL4 was obtained in two steps via the 
tosylate as described previously for preparation of 53 . Solid LAH pellets (7.34 g, 1 .5 eq) 
were added in increments over 1 hour to 53 (25. 6g, 0.129 mol) in 300 mL anhydrous 
ether under argon and stirred for 8 hours. The solution was chilled below 0°C, quenched 
with 300 mL 3N HC1, then combined with 100 ml ether. The organic layer was washed 



35 



with 3N HC1, water, IN NaOH, and brine, then dried over Na 2 S0 4 and concentrated to 
22. 8 g clear colorless liquid (>99% GC). The resulting 11-dodecenol was converted to 
the tosylate as described previously. The crude tosylate was recrystallized from 
MeOH/H 2 0, taken up in HCCI 3 and dried with Na 2 S0 4 , and finally concentrated to 
38. 1 9g white solid. The toslylate was converted to 54 by the action LiBr in refluxing 
acetone as described previously, and purified by fractional distillation. Yield: 22. 5g, 71% 
overall for 3 steps. ‘H- NMR 8 (ppm) 5.78(m, 1H); 4.93(m, 2H); 3.36(t, 2H); 2.01(q, 
2H); 1.83(m, 2H); 1.35(m)+1.28(br) = 14H. I 3 C-NMR 8 139.06, 114.15, 33.96, 33.73, 
32.79, 29.30, 29.02, 28.80, 28.67, 28.10. 

Synthesis of 13-tetradecenvl bromide(5.6). Prepared from 1 3-tetradecenoic acid 
( 5 . 5 ) in three steps as previously described for the preparation of 54 , excluding 
recrystallization of the intermediate tosylate. The bromide was purified by flash 
chromatography (pentane). Yield: 56 % overall for 3 steps. 1 H- NMR 8 (ppm) 5.78(m, 
1H); 4.94(m, 2H); 3.38(t, 2H); 2.02(q, 2H); 1.83(m, 2H); 1.36(m)+1.25(br) = 18H. 13 C- 
NMR 8 139.16, 1 14.06, 33.93, 33.81, 32.82, 29.53, 29.42, 29.13, 28.92, 28.77, 28.16. 

Synthesis of symmetrical alcohol-functionalized dienes 

Synthesis of 1 l-oxo-l,20-heneicosadiene ( 5 . 7 ). The ketone diene 5/7 was 
prepared in two steps via Claisen condensation of ethyl 10-undecenoate to form the (3- 
keto ester diene , followed by dealkoxycarbonylation without isolation of the first step. 
Claisen condensation: Ethyl 10-undecenoate (60mL, 52.44 g, 247 mmol, 1 eq) was 
added dropwise to a stirring suspension of KH [34.97g (35%wt/wt in mineral oil), 296.4 
mmol, 1.2 eq] in 300 ml dry THF at room temperature. After 12 hours, the reaction was 



36 



cooled to OC, quenched by addition of 30 ml acetic acid, and the solution was poured into 
I L cold D.I. water. The layers were separated and the aqueous phase was extracted with 
ether (2 x 100 ml). The combined organic solutions were washed with D.I. water (3 x 
1L), dried over MgS04, and concentrated under reduced pressure to a yellow oil. Crude 
yield can not be determined as it contains mineral oil from the KH dispersion. 
Dealkoxycarbonylation: The crude oil was combined with 93.4 g LiCl, 450 ml DMSO, 
and 20 ml D.I. water and the mixture was sparged with argon for 30 minutes. The 
mixture was then vigorously stirred at 155°C under argon for 12 hours. Upon cooling to 
room temperature, the mixture separated and the upper phase solidified. The solid was 
taken up in 500 ml ether and washed with D.I. water(3 x 300 ml). The lesulting crude 
product was reduced with LAH without further purification. An analytical sample was 
obtained by recrystallization from methanol. %. 'H- NMR 8 (ppm) 5.78(m, 2H); 4.92(m, 
4H); 2.35(tr, 4H); 2.0 l(q, 4H); 1.53(m, 4H); 1.32(br, 24). 13 C-NMR 5 21 1.59, 139.13, 

1 14.1 1, 42.78, 33.75, 29.33, 29.27, 29.05, 28.88, 23.87. 

Synthesis of 1 1 -hydroxy- 1.20-heneicosadiene ( 5 . 8 ). Prepared by LAH reduction 
of crude ketone diene 577 . The crude ketone was dissolved in 300 ml dry THF under 
argon in a 500 ml schlenck flask equipped with an addition funnel. LAH (62 mL 1M in 
THF, 0.5 eq) was added dropwise via the addtion funnel at room temperature with 
stirring. After 12 hours, the solution was cooled to 0°C and quenched by slow addition 
of 100 ml 3M HC1. Transferred to separatory funnel with 300 ml ether and discarded the 
aqueous phase. The ether solution was washed with sat'd NaHCCL (300 ml), D.I. water 
(3 x 300 ml), dried over MgS0 4 , and concentrated to an oily off-white solid. The pure 
diene 5.8 was obtained as a white solid after flash chromatography (10% 



37 



EtO Ac/hexanes). Yield 27.0 g, 70.7% overall for three steps. 'H- NMR 5 (ppm) 5.78(m, 
2H); 4.94(m, 4H); 3.55(b, 1H); 2.01(q, 4H); 1.37(m)+1.25(br) = 28H. 13 C-NMR 8 
139.16, 1 14.07, 71.94, 37.45, 33.77, 29.66, 29.54, 29.42, 29.10, 28.89, 25.63. 

Synthesis of l,22-tricosadiene-12-ol ( 5 . 9 ). 10-Undecenyl Bromide ( 5 . 2 ) (5 g, 
21.4 mmol, 1.0 eq) was added slowly via syringe to freshly scoured Mg turnings (0.525 
g, 21.6 mmol, 1.01 eq) in 50 mL dry THF with rapid stirring under argon. After stirring 
lhour at 45°C, the resulting Grignard solution was chilled to 0°C and methyl formate 
(0.63 mL, 0.48 eq) was added dropwise, while maintaining the temperature below 5°C. 
The contents were warmed to RT, stirred 30 minutes and then concentrated. The product 
mixture was taken up in pentane and 3N HC1. The organic layer was washed with 3N 
HC1 and water, dried over Na 2 S0 4 , and concentrated to 3.0 g white solid. The alcohol 
was purified by flash chromatography (10%EtO Ac/Hexanes). Yield: 2.76g, 76.6%. 'H- 
NMR 8 (ppm) 5.78(m, 2H); 4.92(m, 4H); 3.55(b, 1H); 2.01(q, 4H); 1.37(m)+1.25(br) = 
32H. I3 C-NMR 8 139.16, 1 14.06, 71.94, 37.47, 33.77, 29.69, 29.59, 29.53, 29.46, 29.1 1, 
28.91,25.63. 

Synthesis of l,24-pentacosadiene-13-ol ( 5 . 10 ). Diene 5.10 was prepared by 
reaction of 11-dodecenyl magnesium bromide (from 54) with methyl formate utilizing 
similar protocol described previously for preparation of diene 5 . 9 . Yield: 81.4%. NMR 
spectra identical within experimental error with that of 5 . 9 . 

Synthesis of l,28-nonacosadiene-15-ol ( 5 , 11 ). Diene 5.11 was prepared by 
reaction of 13-tetradecenyl magnesium bromide (from 5 . 6 ) with methyl formate utilizing 



38 



similar protocol described previously for preparation of diene 5 ^ 9 . Yield: 39 %. NMR 
spectra identical within experimental error with that of 5^9. 

Synthesis of symmetrical dienes with pendant acetate groups 

General procedure for acetvlating alcohol-functional dienes. The alcohol diene 
was weighed into a schlenk flask and dried by either stirring in the molten state at 45- 
70°C under reduced pressure (< 5 mmHg) for several hours or by dissolution in dry 
toluene followed by evaporation under reduced pressure. Excess AcCl (4-10 eq) was 
added dropwise to a solution of the alcohol in either dry toluene or anhydrous ether 
containing pyridine (slight excess over that of AcCl). Both AcCl and pyridine were 
transferred via syringe after storage over CaH 2 under argon. After aqueous workup, the 
crude acetates were purified by flash chromatography (2-3%EtOAc/pentane or hexanes). 
Products obtained in this manner were free of detectable impurities (GC. NMR) with the 
exception of 15-acetoxy-l,28-nonacosadiene which required further purification by 
HPLC (normal phase, 2%EtOAc/hexanes) to separate impurities with nearly identical R f 
values, possibly caused by side reaction (see below). 

Synthesis of 1 l-acetoxv-l,20-heneicosadiene ( 5 . 12 ). Prepared as described above 
from x 8 . Yield: 82%, clear, colorless oil (100% GC). LRMS: 351 (m+). Elem. Anal. 
Calcd. for C 23 H 42 O 2 : C, 78.80; H, 12.08. Found: C, 78.82; H, 12.21. 'H- NMR 8 (ppm) 
5.77(m, 2H); 4.94(m)+4.83(p) = 5H; 2.02(q)+l .99(s) = 7H; 1.46(m, 4H); 

1.34(m)+1.23(br) = 24H. 13 C-NMR 5 170.71, 139.02, 114.04,74.27,34.05,33.73,29.41, 



29.33, 29.03, 28.85, 25.24, 21.15. 



39 



Synthesis of 12-acetoxy-l,22-tricosadiene (5.13). Prepared as described above 
from 5^9. Yield: 86%, clear, colorless oil (100% GC). LRMS: 378 (m+). Elem. Anal. 
Calcd. for C^RkA: C, 79.30; H, 12.25. Found: C, 79.25; H, 12.29. 'H- NMR 8 (ppm) 
5.77(m, 2H); 4.94(m)+4.83(p) = 5H; 2.02(q)+1.99(s) = 7H; 1.46(m, 4H); 

1.34(m)+1.23(br) = 28H. 13 C-NMR 8 170.82, 139.13, 114.06,74.36,34.10,33.76,29.47, 

29.42, 29.09, 28.91, 25.28, 21.23. 

Synthesis of 13-acetoxv-l,24-pentacosadiene (5.14). Prepared as described above 
from 5Jj). Yield: 93%, clear, colorless oil (100% GC). LRMS: 406 (m+). Elem. Anal. 
Calcd. for C 27 H 50 O 2 : C, 79.74; H, 12.39. Found: C, 79.81; H, 12.45. 'H- NMR 5 (ppm) 
5.77(m, 2H); 4.94(m)+4.83(p) = 5H; 2.02(q)+1.99(s) = 7H; 1.46(m, 4H); 

1.34(m)+1.23(br) = 32H. 13 C-NMR 8 170.86, 139.10, 114.07,74.36,34.15,33.70,29.49, 

29.43, 29.03, 28.85, 25.24, 21.17. 

Synthesis of 15-acetoxy-l,28-nonacosadiene (5.15). Prepared as described above 
from 5.11 . Violent reaction occurred upon addition of AcCl (evolution of gas, exotherm, 
and formation of black particulates) perhaps due to adventitious moisture. Yield: 62 %, 
white solid (100% GC). LRMS: 463 (m+1). Elem. Anal. Calcd. for C 3 iH 58 0 2 : C, 80.45; 
H, 12.63. Found: C, 80.39; H, 12.75. 'H- NMR 8 (ppm) 5.78(m, 2H); 4.93(m)+4.80(p) = 
5H; 2.02(q)+2.00(s) = 7H; 1.47(m, 4H); 1.33(m)+1.23(br) = 40H. I3 C-NMR 8 170.89, 
139.21, 114.06, 74.41, 34.10, 33.81, 29.62, 29.60, 29.56, 29.53, 29.50, 29.14, 28.93, 



25.30, 21.28. 



40 



Synthesis of symmetrical dienes with pendant carboxylic acid and alkoxycarbonyl 
groups 

Synthesis of 2-(9-decenvl)-l 1-dodecenoic acid(5.18). Diene 5.18 was prepared 
by C-alkylation of the dianion (enolate) of 1 1-dodecenoic acid ( 5.3 ) with 9-decenyl 
bromide (5.1) . The dianion of 5J was generated by reaction with 2 eq LDA prepared in 
situ. All manipulations were conducted using standard schlenk techniques under argon 
atmosphere. Butyl lithium (74.4 mL 2.5M in hexanes, 0.186 mol, 2.05 eq) was added 
slowly via syringe to diisopropyl amine (26.2 mL, 0.186 mol, 2.05 eq) in 100 mL THF 
while maintaining the temperature at -15°C>T>-30°C. The solution was stirred 5 min at 
-20°C, then 15 minutes at RT. 5J (18 g, 90.8 mmol, 1.0 eq) was added slowly via 
syringe while maintaining the temperature at -15°C>T>-30°C during which a milky 
white suspension formed. The solution was cooled to 0°C and HMPA (15.8 mL, 90.8 
mmol, 1.0 eq) was added via syringe followed by rapid addition of 5J[ (19.9 g, 90.8 
mmol, 1.0 eq) via syringe at which time the solution temperature elevated to 15-20°C. 
After stirring 12 hours the reaction was chilled to 0°C, quenched with 500 mL ice cold 
3N HC1, and concentrated under reduced pressure. The heterogeneous aqueous mixture 
was extracted with pentane and the organic layer washed with 3N HC1, water and brine, 
then dried over Na 2 S0 4 and concentrated to ~38g pale yellow oil. The oil was further 
purified by flash chromatography (15%EtOAc/pentane) to render pure 5.18 in 71% 
isolated yield as a white solid. ‘H- NMR 5 (ppm) 5.7(m, 2H); 4.94(m, 4H); 2.33(m, 1H); 
2.0 l(q, 4H); 1.63(m, 2H); 1.48(m)+1.27(br) = 26H. I3 C-NMR 5 183.26, 139.15, 1 14.1 1, 
45.57, 33.79, 32.12, 29.51, 29.39, 29.09, 28.89, 27.34. 



41 



Synthesis of 2-( 1 l-dodecenyD-13-tetradecenoic acid(5.19). Diene 5.19 was 
prepared by C-alkylation of the dianion (enolate) of 1 3-tetradecenoic acid ( 5.5) with 1 1- 
dodecenyl bromide (5.4), following the same procedures for preparation and purification 
described previously for diene 5.18 . The carboxylic acid was isolated as a white powder 
in 82% yield. ‘H- NMR 5 (ppm) 5.7(m, 2H); 4.94(m, 4H); 2.33(m, 1H); 2.0 l(q, 4H); 
1.63(m, 2H); 1.48(m)+1.27(br) = 34H. 13 C-NMR 5 183.34, 139.10, 114.105, 45.63, 
33.82, 32. 1 1 , 29.49, 29.38, 29. 1 1 , 28.80, 27.32. 

Synthesis of 2-(8-nonenvl)-12-tridecenoic acid(5.25). Diene 5^25 was prepared 
by C-alkylation of the dianion (enolate) of 10-undecenoic acid (Acros) with 10- 
undecenyl bromide (5.2), following the same procedures for preparation and purification 
described previously for diene 5.18 . The carboxylic acid was isolated as a clear, 
colorless oil in 85% yield. 

Synthesis of methyl 2-(l l-dodecenyl)-13-tetradecenoate (5.22). Diene 5.22 was 
prepared by base-catalyzed O-alkylation of carboxylic acid diene 5.19 with Mel. 
Anhydrous K 2 C0 3 (1.32 g, 9.6 mmol, 2.5 eq) and 5,19 (1.5 g, 3.8 mmol, 1 eq) were 
combined with 50 mL dry acetone in a scklenk flask fitted with a water-cooled condenser 
under argon. Mel (1.9 mL, 30.5 mmol, 8 eq) was added via syringe. The mixture was 
heated with a 60-70°C oil bath for 24 hours with vigorous stirring. To minimize loss of 
Mel due to evaporation, the condenser temperature was maintained at 15°C and the argon 
backpressure was slightly increased over atmospheric. Two additional aliquots of Mel (2 
mL) were added at intervals via syringe. The acetone was evaporated under reduced 
pressure and the concentrate was taken up in 50 mL pentane and water. The organic 



42 



layer was washed with sat’d NaHC0 3 , 3N HC1, and brine, dried over Na 2 S0 4 and 
concentrated to 1.57g pale yellow oil. The oil was further purified by flash 
chromatography (5%EtO Ac/hexanes) to 1.45g (93.5%) 5,22 as a clear colorless oil 
(100% GC). LRMS(EI): 406 (m+). Elem. Anal. Calcd. for C27H50O2: C, 79.74; H, 
12.39. Found: C, 79.79; H, 12.54. 'H- NMR 8 (ppm) 5.80(m, 2H); 4.94(m, 4H); 3.70(s, 
3H) 2.35(m, 1H); 2.04(q, 4H); 1.61(m, 2H); 1.39(m)+1.23(br) = 34H. 13 C-NMR 5 
177.06, 139.18, 1 14.06, 51.24, 45.70, 33.80, 32.49, 29.54, 29.46, 29.13, 28.91, 27.46. 

Synthesis of methyl 2-(9-Decenvl)-ll-dodecenoate ( 5 . 20 ), Diene 5.20 was 
prepared by base-catalyzed O-alkylation of carboxylic acid diene 5.18 with Mel, utilizing 
similar procedures for synthesis and isolation described previously for 5.22 (100% GC). 
Yield; 88%. LRMS(EI): 351 (m+). Elem. Anal. Calcd. for C23H42O2: C, 78.80; H, 12.08. 
Found: C, 78.90; H, 12.26. 'H- NMR 8 (ppm) 5.77(m, 2H); 4.92(m, 4H); 3.63(s, 3H) 
2.29(m, 1H); 2.00(q, 4H); 1.55(m, 2H); 1.33(m)+1.23(br) = 26H. I3 C-NMR 8 177.01, 

139.12, 1 14.06, 51.21, 45.67, 33.76, 32.47, 29.49, 29.35, 29.05, 28.86, 27.42. 

Synthesis of ethyl 2-(9-Decenvl)-l 1 -dodecenoate ( 5 . 21 ), Diene 5^21 was 

% 

prepared by base-catalyzed O-alkylation of carboxylic acid diene 5.18 with EtI, utilizing 
similar procedures for synthesis and isolation described previously for 5.22 (100% GC). 
Yield: 83%. LRMS(EI): 365 (m+). Elem. Anal. Calcd. for C24H44O2: C, 79.06; H, 12.16. 
Found: C, 79.10; H, 12.26. 'H- NMR 8 (ppm) 5.77(m, 2H); 4.92(m, 4H); 4.13(q, 2H) 
2.34(m, 1H); 2.04(q, 4H); 1.55(m, 2H); 1 .33(m)+l ,23(br) = 29H. I3 C-NMR 8 176.51, 

139.13, 1 14.06, 59.85, 45.72, 33.76, 32.47, 29.50, 29.36, 29.06, 28.89, 27.39, 14.32. 



43 



Synthesis of 2-(methyl)-l 1-dodecenoic acid(5.25). 5.25 was prepared by C- 

alkylation of the dianion (enolate) of propionic acid with 9-decenyl bromide ( 5 , 1 ), 
following the same procedures for preparation and purification described previously for 
diene 5 , 18 . The carboxylic acid was isolated as a clear colorless oil in 70.8% yield. 'H- 
NMR 5 (ppm) 1 1.34(br, 1H); 5.78(m, 1H); 4.94(m, 2H); 2.43(m, 1H); 2.01(q, 2H); 
1.62(m, 1H); 1.32(m)+1.27(br)+1.15(d,) = 16H. I3 C-NMR 5 183.69, 139.16, 114.09, 
39.39, 33.78, 33.49, 29.45, 29.39, 29.07, 28.89, 27.10, 16.80. 

Synthesis of 2-(9-decenyl)-2-methyl-l 1-dodecenoic acid(5,23). 5.23 was 

prepared by C-alkylation of the dianion (enolate) of 2-(methyl)-l 1-dodecenoic acid 
( 5.25 ) with 9-decenyl bromide ( 5 . 1 ) , following the same procedures for preparation and 
purification described previously for diene 5 . 18 . The carboxylic acid was isolated as a 
clear colorless oil in 76 % yield. *H- NMR 8 (ppm) 5.78(m, 2H); 4.93(m, 4H); 2.01 (q, 
4H); 1.59(m, 2H); 1 ,35(m)+l ,25(br)+l . 1 l(s) = 33H. I3 C-NMR 8 184.73, 139.16, 114.11, 
45.75, 39.06, 33.80, 30.12, 29.43, 29.1 1, 28.91, 24.43, 20.99. 

Synthesis of methyl 2-(9-decenyl)-2-methyl-l 1-dodecenoic acid ( 5 . 24 ). Diene 
5.24 was prepared by base-catalyzed O-alkylation of carboxylic acid diene 5.23 with 
Mel, utilizing similar procedures for synthesis and isolation described previously for 5.22 
(100% GC). Yield: 96%. LRMS(EI): 364 (in+). 'H- NMR 8 (ppm) 5.77(m, 2H); 

4.92(m, 4H); 3.61(s, 3H); 2.00(q, 4H); 1.55(m, 2H); 1.34(m)+1.22(br) = 26H; 1.07(s, 
3H). I3 C-NMR 8 178.09, 139.12, 114.07,51.38,45.92,39.44,33.77,30.07,29.40,29.08, 



28.88, 24.51, 21.13. 



44 



Synthesis of symmetrical dienes with pendant chloride and phenyl groups 

Synthesis of 1 l-chloro-l,20-heneicosadiene ( 5 . 16 ). 5.8 (2.0 g, 65 mmol, 1 eq) in 
10 mL CC1 4 was added via cannula under argon to Ph^P (5.5 g, 21 mmol, 3.2 eq) in 20 
mL CC1 4 . The vessel was heated in an 80°C oil bath for 16 hours, during which time a 
white precipitate formed and the solution developed a yellow tinge. The mixture was 
filtered and the solids washed with pentane. Addition of the pentane washings to the 
filtrate caused additional precipitation, which was again filtered, and the resulting liquid 
was concentrated under reduced pressure. Repeated dissolution in minimal pentane, 
chilling, and filtration followed by concentration yielded 2. 1 7g slightly cloudy liquid. 
Pure (100% GC) diene 5.16 was obtained after flash chromatography (pentane) as a clear, 
colorless liquid. Note: Ph 3 P and 5^8 were dried in the molten state under reduced 
pressure prior to use and CC1 4 was distilled from P 2 0 5 (A previous attempt with less 
rigorous exclusion of moisture gave much lower yields). Yield: 1.8g, 85%. LRMS(EI): 
326 (m+). Elem. Anal. Calcd. for C 21 H 39 CI: C, 77.14; H, 12.02. Found: C, 77.07; H, 
12.13. . 'H- NMR 5 (ppm) 5.79(m, 2H); 4.94(m, 4H); 3.86(m, 1H); 2.02(q, 4H); 1.69(m, 
4H); 1.50(m)+1.36(m)+1.27(br) = 24H. 13 C-NMR 5 139.13, 114.12,64.29,38.50,33.80, 
29.45, 29.40, 29. 16, 29.09, 28.90, 26.48. 

Synthesis of 1 1 -phenyl- 1,20-heneicosadiene ( 5 . 17 ), Diene 5.17 was prepared via 
nucleophilic attack of phenyl lithium on ketone diene 577 , followed by in situ 
deoxygenation of the resulting tertiary benzylic alkoxide. Bromobenzene (2.0 mL, 19.6 
mmol, 2 eq) in 20 ml ether was added slowly to lithium foil (1.087g, 157 mmol, 16eq) in 
40 mL ether under argon. After refluxing for 1 hour, ketone diene 577 (3g, 9.79 mmol, 1 



45 



eq) in 20 mL ether was added slowly and the mixture was stirred for 1 hour. The mixture 
was chilled to -50°C and ~50 mL NH 3 was condensed into the flask from a tank at which 
time the reaction turned blue. After 10 minutes, the reaction was quenched via addition 
of 1 1 g NH 4 Cl(anh) in portions through a solids addition funnel. The ammonia was 
evaporated, the mixture was filtered and the ether solution was washed with brine and 
D.I. water. Pure (100% GC) diene 5.17 was obtained after flash chromatography 
(pentane) as a clear, colorless liquid. 2.2 g, 61 % yield. LRMS(EI): 368 (m+). Elem. 
Anal. Calcd. for C 27 H 4 4 : C, 87.97; H, 12.03. Found: C, 88.00; H, 12.06. . 'H- NMR 5 
(ppm) 7.25(m, 2H); 7.20(m, 3H); 5.79(m, 2H); 4.95(m, 4H); 2.45(m, 1H); 2.01 (q, 4H); 
1.54(m, 4H); 1.34(m)+1.21(br) = 24H. 13 C-NMR 5 146.37, 139.18, 128.10, 127.63, 
125.66, 1 14.06, 46.06, 36.97, 33.79, 29.73, 29.47, 29.10, 28.91, 27.60. 

Synthesis of ADMET Model Ethylene/Polar Monomer Copolymers 

General Procedure for Preparation of Model Polymers . The pure diene monomers 
may be adequately dried and degassed by replacing the atmosphere three times with 
argon via three successive pump/purge cycles and then stirring under reduced pressure at 
50°C for several hours. The ADMET polymerizations and subsequent hydrogenations 
were all conducted utilizing the procedure described for polymers fL6 and 4/9. Inside an 
argon-purged dry box, monomer and catalyst (400: 1 ) were combined in a small tube with 
a single 24/40 neck and containing a magnetic stir bar. The tube was sealed with a 
Kontes high vacuum valve, removed to a vacuum line and the pressure was reduced in 
steps ultimately reaching 10 " 2 - 10 ‘ 3 mm Hg while increasing the temperature to 65°C. 
Alternately, the reaction may be conducted in a pressure tube fitted with a vacuum valve 



46 



via teflon bushing thereby eliminating the need to transfer the ADMET polymer to 
another vessel before hydrogenation. A homemade magnetic stirrer constructed from a 
powerful magnet and a stir motor was utilized to maintain agitation as long as possible as 
the viscosity increased during the polymerization. After 48 hours, the vessel was sealed 
and returned to the dry box. In a glass pressure tube, the polymer/catalyst mixture was 
combined with silica gel-60 (100 X the weight of catalyst) and sufficient toluene to aid 
dispersion (generally 10-15 ml for 500 mg polymer). Hydrogen pressure was applied as 
quickly as possible, within 2-3 minutes, to deactivate the catalyst residue toward 
metathesis prior to dissolution of the polymer. The mixture was then heated to 90°C with 
vigorous stirring while maintaining a constant pressure of 120 psig H 2 for 24 hours. The 
mixture was filtered to remove the silica/catalyst residue composite and the toluene 
evaporated to yield colorless viscous or water-white solid polymers. Number average 
molecular weights for these polymers as measured by GPC(vs polystyrene) were 
consistently in the range 2 - 5 x 10 4 g/mol (PDI = 2.0). 

Synthesis of ADMET model ethylene/vinyl acetate copolymers 

ADMET/Hydrogenation of 1 l-acetoxy-l,20-heneicosadiene (HP5.12). Elem. 
Anal. Calcd. for [C 2 iH 4 o0 2 ]: C, 77.72; H, 12.42. Found: C, 77.63; H, 12.50. 'H- NMR 8 
(ppm) 4.83(p, 1H); 2.02(s, 3H); 1.50(br, 4H); 1.23(br, 32H). 13 C-NMR 8 170.86, 74.45, 
34.15, 29.71, 29.58, 25.33, 21.28. 

ADMET/Hydrogenation of 12-acetoxy-l ,22-tricosadiene (HP5.13), Elem. Anal. 
Calcd. for [C^H^Cb]: C, 78.35; H, 12.58. Found: C, 78.16; H, 12.63. ‘H- NMR 8 (ppm) 



47 



4.82(p, 1H); 2.03(s, 3H); 1.49(br)+1.25(br) = 40H . 13 C-NMR 5 170.88, 74.42, 34.13, 
29.65, 29.53, 29.29, 29.17, 25.31, 21.26. 

ADMET/Hydrogenation of 1 3-acetoxy- 1 ,24-pentacosadiene (HP5.14), Elem. 
Anal. Calcd. for [C 2 5H 48 0 2 ]: C, 78.88; H. 12.71. Found: C, 78.83; H, 12.76. 'H- NMR 5 
(ppm) 4.83(p, 1H); 2.0 1 (s, 3H); 1.47(br)+l ,22(br) = 44H. 13 C-NMR 5 170.93, 74.44, 
34.1 1, 29.70, 29.58, 29.54, 25.31, 21.23. 

ADMET/Hydrogenation of 15-acetoxv-l,28-nonacosadiene (HP5.15). Elem. 
Anal. Calcd. for [C 29 H 56 0 2 ]: C, 79.75; H, 12.92. Found: C, 79.61; H, 12.94. 'H- NMR 5 
(ppm) 4.83(p, 1H); 2.01(s, 3H); 1.47(br, 4H); 1.22(br, 48H). 13 C-NMR 5 170.80, 74.38, 
34.10, 29.62, 29.564, 29.50, 29.09, 25.30, 21.29. 

Synthesis of ADMET model ethylene/acrylate model copolymers 

ADMET/Hydrogenation of methyl 2-(l l-dodecenyl)-13-tetradecenoate (HP5.22). 
Elem. Anal. Calcd. for [C 25 H 48 0 2 ]: C, 78.88; H, 12.71. Found: C, 78.92; H, 12.78. *H- 
NMR 8 (ppm) 3.7 l(s, 3H) 2.33(m, 1H); 1.61(m)+1.39(m)+1.23(br) = 44H. I3 C-NMR 5 
1 77.06, 5 1 .24, 45.70, 32.49, 29.52, 29.37, 29. 1 0, 28.9 1 , 27.43. 

ADMET/Hydrogenation of methyl 2-(9-Decenyl)-l 1 -dodecenoate (HP5.20). 
Elem. Anal. Calcd. for [C 2 ,H 40 O 2 ]: C, 77.72; H, 12.42. Found: C, 77.44; H, 12.43. 'H- 
NMR 5 (ppm) 3.63(s, 3H) 2.29(m, 1H); 1.54(m)+1.40(m)+1.21(br) = 36H. 13 C-NMR 5 
177.06, 51.22, 45.71, 32.50, 29.67, 29.57, 29.47, 27.47. 



48 



ADMET/Hydrogenation of ethyl 2-(9-DecenyI)-l 1 -dodecenoate (HP5.21). Elem. 
Anal. Calcd. for [C 22 H 42 O 2 ]: C, 78.05; H, 12.50. Found: C, 77.81; H, 12.49. 'H- NMR 8 
(ppm) 4.05(q, 2H); 2.38(m, 1H); 1.61(m,)+1.39(m)+1.22(br) = 39H. I 3 C-NMR 8 176.43, 
59.89, 45.67, 32.47, 29.50, 29.36, 29.06, 28.92, 27.43. 14.32. 

Synthesis of ADMET model ethylene/vinyl chloride and styrene copolymers 

ADMET/Hydrogenation of 1 l-chloro-l,20-heneicosadiene (HP5.16), Although 
the ADMET polymerization proceeded in similar fashion to all others, analysis of the 
polymer after attempted hydrogenation showed very little conversion/virtually unchanged 
olefin signals in *H and 13 C NMR. The resulting polymer was fully saturated using Pd/C. 
Unsaturated polymer P5.16 was dried and degassed by placing under vacuum at 80°C for 
several hours. Polymer P5.16 (483 mg) was combined with Pd/C (10% wt/wt, 500 mg) 
and 15 ml dry, degassed toluene under argon in a pressure tube. The mixture was 
exposed to a constant 80 psig H 2 at room temperature with vigorous stirring. After 8 
hours, the polymer was seen to precipitate as evidenced by gelatinous particles and the 
Pd/C was no longer freely dispersing. The mixture was heated to 40°C at which time 
stirring resumed and the gelatinous particles disappeared. After 24 hours, the mixture 
was filtered through a medium glass frit (while heated) and then passed through a 0 . 2 |im 
syringe filter to insure complete removal of fine Pd/C particles. The water-white solid 
polymer was obtained in quantitative yield after precipitation into methanol. Elem. Anal. 



Calcd. for [C, 9 H 37 C1]: C, 75.83; H, 12.39; Cl, 11.78. Found: C, 75.74; H, 12.51; Cl, 
11.68. 'H- NMR 8 (ppm) 3.86(m, 1H); 2.02(q, 4H); 1.68(m, 4H); 



49 



1.47(m)+1.38(m)+1.23(br) = 32H. 13 C-NMR 5 64.39, 38.53, 33.80, 29.70, 29.59, 29.53, 
29.21,26.51. 

ADMET/Hydrogenation of 1 1 -phenyl- 1 ,20-heneicosadiene (HP5.17) , Elem. 
Anal. Calcd. for [C 25 H 42 ]: C, 87.64; H, 12.36. Found: C, 87.39; H, 12.46. . *H- NMR 8 
(ppm) 7.25(m, 2H); 7.20(m, 3H); 2.43(m, 1H); 1.60(m, 4H); 1.21(br, 32H). ,3 C-NMR 8 
146.43, 128.10, 127.64, 125.64, 46.07, 37.00, 29.79, 29.71, 29.67, 29.57, 27.63. 



CHAPTER 3 

ADMET DEPOLYMERIZATION OF 1 ,4 -POLYBUTADIENE 
UTILIZING A WELL-DEFINED RUTHENIUM CATALYST 



Extensive study has been made of the use of olefin metathesis to produce small 
molecules from unsaturated elastomers where the bulk of the work has focused on 
reaction of 1,4-polybutadiene (1,4-PB) with substituted olefins as chain transfer agents 
(CTA). 52 As discussed in Chapter One, initial studies in this field, just as in all areas of 
metathesis, utilized ill-defined catalyst systems consisting of some species with 
sufficient lewis aciditiy to invoke side reactions such as vinyl addition via cationic 
pathways. Thus these reactions were termed metathesis degradation rather than 
depolymerization. With the advent of well-defined catalysts, and more recently classical 
systems with cocatalysts of diminished Lewis acidity, clean metathesis depolymerizaiton 
could be achieved. 

The reaction is depicted in Figure 3-1, where cross-metathesis of the CTA with 
the unsaturated linkages of the polymer backbone results in chain-scission and end- 
fuctionalization. As in other normal cross-metathesis reactions, the product distribution 
is governed by statistics. In the event of complete depolymerization, the only products 
would be diene and excess CTA. In most studies, this has only been achieved with a 
large excess of CTA. For example, when using the catalyst WCl 6 /Et 4 Sn/Et20, 200 
equivalents of hex-3-ene were required to achieve above 90% conversion of 1,4-PB to 
the monomer, deca-3,7-decadiene (Figure 3-1, R = Et, n = l). 2 



50 



51 




Figure 3-1. Inter- and intramolecular depolymerization of 1 ,4-polybutadiene. 

Unsaturated elastomers may also be depolymerized via intramolecular metathesis 
to produce macrocycles or oligomeric cyclics. This ring-chain equilibrium, familiar in 
step/equilibrium polymerizations is affected by temperature, repeat structure, and of 
course concentration. At sufficient dilution, high molecular weight 1,4-PB may be 
completely depolymerized to small cyclics (Figure 3-1, m = 2-8). In depolymerizations 
employing CTA’s, these cyclics are present in neglible amounts at equilibrium provided 
CTA/PB repeat unit > 10. Thus careful control of reaction conditions can determine the 
position of the equilibrium which is established between the monomeric species, linear 
and cyclic oligomers, and higher molecular weight species. 

Recently, Wagener and coworkers reported the first example of a complete, non- 
statistical depolymerization of 1,4-PB to monomeric diene using a silane-functional 
CTA (Figure 3-1, R = CIUSiMeaCl, n = 1) catalyzed by 1.2b , 20 Approximately 3 
equivalents of the CTA were sufficient to achieve this unprecendented observation. 
Crowe and coworkers later reported 53 that allyl silanes function as selective olefin 
metathesis substrates. Cross metathesis is much more favored between allyl silanes and 
alkyl olefins than self-metathesis of either class of olefin due to electronic and steric 



52 



factors. It may be assumed that this phenomenon is responsible for the above-mentioned 
non-statistical depolymerization. 

Metathesis depolymerization with appropriately functionalized CTA’s results in 
elastomers with reactive end-groups. Hydroxy-telechelic polybutadienes (HTPB) are 
highly desired due in no small part to their technical utility in polyurethane formulations. 
Due to the intolerance of most metathesis catalysts for hydroxyl functionality, HTPB s 
have been prepared by indirect routes where the hydroxyl group is masked during the 
metathesis reaction. Borane functional CTA’s were used in combination with 
WCl 6 /Me 4 Sn in one such example (Figure 3-2), where the borane functional telechelics 
could then be converted under mild conditions to HTPB’s. 52r 



Figure 3-2. Indirect synthesis of HTBP via metathesis depolymerization. 

Reports of the broad utility of the ruthenium metathesis catalysts (Figure 3-3) 
described in part in Chapter One prompted their investigation in metathesis 
depolymerization chemistry. Of particular significance are their tolerance to functional 
groups and stability in the presence of ethylene. In this chapter the depolymerization of 
cis- 1,4-polybutadiene with ethylene and with ester-functionalized CTA’s catalyzed by 






1.4 is described. Observations of the remarkable ability of catalyst IA to effect 



53 



metathesis depolymerization of ultra-high molecular weight polybutadiene in the bulk 
state are also disclosed. 



CI/„ 

Cl 



PCy 3 

Ru=^ 

PCy 3 

1.3 



.Ph 

Ph 



Cl/ 



Cl 



PCy 3 

Ru=^ 

I 

PCy 3 

1.4 



Ph 



Figure 3-3. Well-defined ruthenium metathesis catalysts. 



Ethenolysis of 1,4-Poly butadiene 

The cross-metathesis between internal olefins and ethylene was first dubbed as 
ethenolysis in 1967. 54 This specific reaction has been the subject of a number of studies 
and has been exploited commercially. 55 Applications include the production of o> 
unsaturated carboxylic esters from esters of naturally occurring fatty acids and a,o 
dienes from cyclic olefins, shown in Figure 3-4. 





Figure 3-4. 




C 2 H 4 



Commercial applications of ethenolysis. 



The depolymerization of 1,4-polybutadiene, as well as other unsaturated 
elastomers, via cross metathesis with ethylene catalyzed by the well-defined metathesis 
catalyst, [(CF3)2CH 3 CO]2(NAr)W=CHC(CH3) 2 Ph has been reported. 56 This reaction 



54 



is essentially the reverse of the ADMET polymerization of 1,5-hexadiene, and if the only 
reaction occurring is ethenolysis, then the principal products would be a, co-vinyl 
terminated polybutadiene oligomers (Figure 3-1, R = H). Should conversion be 
complete, only the monomer 1,5-hexadiene would be produced. Conversions were far 
from complete, but instead this cross metathesis reaction produced clean oligomers with 
M n = 1,000 along with trace amounts of 1,5-hexadiene. It was assumed that the low 
conversions of polymer to pure ADMET monomer were due to inactivation of the 
catalyst through a previously described competing reaction of the catalyst with ethylene. 

Ethenolysis of 1,4-Polybutadiene Catalyzed by a 
Well-Defined Ruthenium Complex 

A polymeric substrate with highly pure microstructure was chosen for these 
studies (cis- 1,4-polybutadiene: 98% minimum cis 1,4-microstructure) (3J_) to simplify 
analysis of product mixtures. A systematic study was performed to determine the effect 
of ethylene pressure on conversion with all other variables held constant, including 
[ 3.11 / 11.41 , polymer concentration, solvent and temperature. Solutions of polymer and 
catalyst in CDCI 3 ([3.11 = 2.5M, [3.1 1/[ 1 .41 = 400) were exposed to a constant pressure 
of C 2 H 4 with vigorous stirring for 48 hours at room temperature, at which time reactions 
were quenched by slowly bubbling air through the solution. To minimize loss of volatile 
products, the reaction mixtures were chilled to T < -50°C and the C 2 H 4 was slowly 
released prior to quenching. The catalyst was removed by briefly stirring with silica gel 
followed by filtration. 



I 



55 




Figure 3-5. Ethenolysis of 1,4-polybutadiene. 



Analysis of the product mixtures by GC showed each to be consistently 
composed of a complex mixture of products, schematically represented in Figure 3-5. 
GC/MS analysis indicated that major products of these reactions were low molecular 
weight a, co-vinyl terminated butadiene oligomers with mass spectral peaks consistently 
found at FW = 82 + n(54), where n has the values between 0 and 5. Oligomers with 
n>5 were not detected. Of course the product with n=0 is the monomer 1 ,5-hexadiene. 
Trace peaks were identified with mass = m(54) which may be due to cyclic butadiene 
oligomers. 

Due to the complex nature of the product mixtures, individual products were not 
isolated. A simple alternative is to simply to calculate a relative moment of the product 
distribution from each experiment to determine the trend in conversion. Therefore, 
product mixtures were analyzed as a whole by gas chromatography/mass spectrometry 
(GC/MS) in concert with 'H and 13 C NMR to determine the trend in conversion vs 
pressure. 

The concept of average molecular weight ' 7 is familiar to the polymer chemist as 
most polymerizations produce not one discrete species with a single molecular weight 
but instead a homologous series of products characterized by an average molecular 
weight. One such average, the number average molecular weight ( M „ ), is directly 
related to conversion for step/condensation polymerizations. Since ethenolysis is the 



56 



microscopic reverse of the forward ADMET reaction, a step polymerization, it seems 

that the most straightforward measure of conversion is to determine M „ . M n values 
were calculated from 'H NMR end-group analysis and from GC plots and the two values 
were in good agreement. 




7 6 5 4 3 2 1 PPm 

Figure 3-6. 'H NMR of 1,4-Polybutadiene (3.1). Insets are partial spectra of product 
mixtures from ethenolysis at A: 30 and B: 400 psig C 2 H 4 . 

NMR spectra of all product mixtures were consistent with a,co vinyl-terminated 

butadiene oligomers. Figure 3-6 demonstrates the extent of ethylene depolymerization 

chemistry. It shows the 'H NMR spectrum of the reactant, cis- 1,4-polybutadiene with 

peaks near 5.4 and 2.1 ppm corresponding to the olefinic and allylic protons 

respectively. The insets "A" and "B" are partial spectra obtained from the product 

mixtures resulting from ethenolysis at 30psig (lowest conversion) and 400psig (highest 



57 



conversion) C 2 H 4 , where the signals centered near 5 and 5.8 ppm represent the terminal 
and penultimate protons, respectively, of terminal vinyl groups. For mixtures consisting 
entirely of linear a,co vinyl-terminated species, number average molecular weights can 
be calculated using equation 3-1: 



M n = 108 



82 



Eqn 3-1 



where Ij 0 and Ie 0 are the areas under the peaks for the internal olefinic protons and 
terminal vinylic protons, respectively. This method of end-group analysis has been 
applied successfully to moderately high-molecular weight polymers with well-defined 
vinyl end groups prepared by the forward ADMET reaction. Since the concentration of 
other products (cyclics) was low in all cases as determined by GC, this may be 
considered a reasonably reliable method to follow conversion. 

The M n also was estimated from the peak areas of GC plots of each reaction 
mixture. A representative GC plot for reaction at 400 psig is shown in Figure 3-7 with 
peaks labeled by their masses determined from GC/MS. No species larger than the 
tetramer were detected in this case. The trace peaks immediately preceding oligomers 
have masses, m(54), corresponding to cyclic butadiene oligomers. The trace peaks 
immediately following major peaks have identical masses as the preceding major peak 
and are assumed to be linear cis/trans isomers. The relative mass percents of each 
species (GC) were converted to mole percents then normalized neglecting contribution 
from solvent and trace peaks due to cyclics. From these values, M ,, can be calculated 



using the well-known equation 3-2: 



58 



M „ 



IM.JV, 

IK 



Eqn 3-2 



where M x and N x equal the molar mass and number of moles, respectively of species x. 




Figure 3-7. Representative GC trace with mass labels (GC/MS). Reaction mixture 
from ethenolysis at 400 psig C 2 H 4 . 



The results calculated from ‘H NMR and GC are compiled in Table 3-1 along 
with the relative mole percents 1,5-hexadiene produced as calculated from GC 
measurements. These results show a general trend in conversion with ethylene pressure 
with highest conversion of polybutadiene to 1,5-hexadiene being observed at 400 psig 
C 2 H 4 . This may be expected as an increase in ethylene pressure results in an increase in 
ethylene concentration relative to polybutadiene. That is, CTA/PB repeat unit is 



59 



increased, making ethenolysis progressively more favored over self metathesis of 
polybutadiene. 

Surprisingly, beyond 400 psig conversions decrease which could be a 
consequence of catalyst mortality. Grubbs has reported 57 that various versions of the 
complexes L4 (L n Ru=-R, R = alkyl) slowly decompose in solution. He proposed 
bimolecular decomposition pathways as evidenced by the formation of structures 
"R-=-R" in solutions consisting of only catalyst and solvent. The methylidene, 
L n Ru=CH2 , which is most definitely the catalytic species in ethenolysis, was reported 
to decompose more quickly in solution; within a matter of hours. 

At higher concentrations of ethylene, it may be that non-productive self- 
metathesis of ethylene begins to offer substantial competition to ethenolysis of the 
polymer. Further, the probable bimolecular decomposition of the methylidene could be 
depleting the system of active catalyst, and these events could lead to the lower 
conversions observed at higher ethylene pressures. The overall result is an asymptotic 
upper limit to increasing conversion with increasing ethylene pressure. Similar limiting 
conversions have been observed in our laboratories in the ethenolysis of low molecular 
weight 1 ,2-disubstituted olefins catalyzed by 1A. 

The identity of trace products with mass = m(54) could not be assigned 
unequivocally to cyclic cyclic oligomers. Although a trace peak assigned the mass of 
108 could be 1,5-cyclooctadiene, the cyclic dimer, an authentic sample of 1,5- 
cyclooctadiene gave a different retention time from GC. 



60 



c 2 H 4 

(psig) 


Mn (g/mol) 
a b 


Mole % (a) 

1 ,5-Hexadiene 


30 


135 


133 


43 


100 


106 


114 


68 


250 


99 


104 


74 


400 


93 


96 


82 


600 


1 18 


128 


52 


800 


122 


134 


47 



Table 3-1. Results of ethenolysis at varied C 2 H 4 pressure (a: GC, b: 'H NMR). 

Abendroth 58 and Thorn-Csannyr 4 investigated the metathesis degradation of 
polybutadiene containing substantial amounts of 1,2 linkages. Figure 3-8 shows the 
cleavage of a 1,4- 1,2- 1,4 triad from the polymer backbone with hex-3-ene as CTA 
(R=Et) to form a triene. Subsequent ring closing metathesis yields either of the cyclic 
products shown. Polymer 3J_, like all 1,4-polybutadiene produced by chain 
polymerization of butadiene contains 1,2-linkages. The presence of these irregularities 
is indicated by a small unresolved multiplet in the 'H NMR of 3J_ (cf Figure 3-6) at 
approximately 5 ppm, signifying the terminal CFF of pendant vinyl groups. RCM and 
ethenolysis of these groups, in either order, as depicted in Figure 3-8 (R = H) renders 
allyl cyclopentene and/or vinyl cyclohexene, both with FW = 108 making them likely 
candidates for the unknown product. Reexamination of the insets in the 'H NMR 
spectra in Figure 3-6 reveals a new small unresolved multiplet at 5.64 ppm which 



61 



corresponds well with the olefinic protons of cyclohexene and cyclopentene (both at 5.7 
ppm). This small multiplet was observed in the 'H NMR spectra of all depolymerized 
3.1 in this study, and will be discussed further later in this chapter. 




Figure 3-8. Products from extrusion of 1 ,4- 1 ,2- 1 ,4 triad of PB . 

To determine the extent to which 3J_ is intramolecularly depolymerized in the 
absence of ethylene, a solution of 3J_ and catalyst in CDCI 3 was prepared in the same 
proportions and concentrations as in the ethenolysis experiments. The PB chunks 
visibly dissolved much faster (within 30 minutes) than in the absence of catalyst, 
indicating rapid reaction. The solution was stirred 48 hours then quenched by exposure 
to air. Silica was used to remove catalyst residue as previously described and the solvent 
was removed under reduced pressure to yield a viscous liquid. H and C NMR were 
consistent with 1,4-PB but with trans content increased to 81% (quant l3 C NMR). The 
number average molecular weight determined by VPO was 2 x 10 1 g/mol. This value is 
not a true measure of the molecular weight produced as any small cyclics would have 
been removed under vacuum during isolation. 



62 



Bulk Depolymerization of Polybutadiene 

A remarkable observation was made regarding the solid-state reaction of catalyst 
1.4 with polymer 3J_ when the two are placed in intimate contact; in the absence of any 
other reagents or solvent, the solid polymer begins to liquefy over a period of a few 
hours! To determine the outcome 3J_ was cut into small chunks and its surface was 
coated with catalyst (1 3.11 / |T.41 =400), and a magnetic stir bar inserted. During the first 
few hours of reaction, the mixture was occasionally agitated with a large magnet, until 
viscosity was sufficiently diminished that slow stirring could be accomplished with a 
magnetic stir plate. After 24 hours, the reaction was quenched by exposure to air, and 

■ I n 

the polymer was dissolved in benzene and isolated as described above. H and C 
NMR were consistent with 1,4-PB but with 72% trans microstructure(quant l3 C). 
Characterization of the polymer by GPC indicated a decrease in M„ to 9.7 x 10 3 g/mol 
(PDI=1.76, monomodal)! Again, any small cyclics, which may have been produced, 
would have been removed during isolation. 

Before discussing the possible route to achieve this depolymerization, the 
uniqueness of its initiation must first be noted. Evidence suggests during reaction of 
catalyst L4 with olefinic substrates, phosphine must dissociate to free the active site. For 
metathesis to occur, the catalyst and the polymer must come in contact on a microscopic 
scale, i.e. the catalyst must dissolve in the polymer. At room temperature, the polymer is 
above its crystalline melting temperature, T m of -6"C, and well above its glass transition 
temperature, T, of -102"C. 37 Under these conditions adequate translational, rotational, and 
vibrational energies exist to allow molecular motion. This molecular motion together 



63 



with the solubility of the catalyst in a range of organic media perhaps is sufficient to 
allow the polymer to imbibe the catalyst and form a "solid solution" at the interface. The 
well known molybdenum metathesis catalyst, 1.2b did not display this reactivity under 
the same conditions, perhaps due to lower solubility in the polymer. 

Assuming the M n value obtained from GPC for the bulk depolymerized PB is 
correct within an order of magnitude, the molecular weight of the polymer was reduced 
by two orders of magnitude (2-3xl0 6 — » 9.7x10' g/mol). In the absence of added 
sources of end-groups, this depolymerization could be considered to have occurred by 
one or a combination of three metathetical routes: a:) conversion to lower molecular 
weight macrocycles via intramolecular metathesis, b:) incorporation of the catalyst 
residues as end groups, and c:) conversion of 1 ,2-linkages to end groups. 




x X = CHPh, RuL n 
Figure 3-9. Incorporation of residues from F4 as end-groups in depolymerization. 

The catalyst itself can be converted to end-groups, as depicted in Figure 3-9. 
Any productive metathesis reaction of the catalyst with the backbone of the polymer 
causes chain scission. The products are a styril -terminated PB chain and a ruthenium 
alkylidene terminated chain. Facile chain transfer of the ruthenium chain end with high 
molecular weight chains begins the equilibration toward lower molecular weight species. 
The total number of linear chains should be equal to the total number of catalyst 




64 



molecules and therefore the average degree of polymerization equals the repeat unit to 
catalyst ratio. For this experiment this gives DP=400 which corresponds to M „ s 2.2 x 
10 4 . Given the inherent error in GPC measurements using dissimilar polymers as 
standards, this calculated M „ agrees within experimental error with that obtained from 
GPC (9.7 x 10 3 g/mol). 

If conversion of the catalyst to end-groups is the major contributor to this 
depolymerization, then the outcome should be sensitive to the catalyst loading. The 
experiment was repeated but with the 13.1 1/11.41=4000, an increase of an order of 
magnitude. For formation of linear chains terminated by catalyst residues alone, the M n 
should be over 2x1 0 5 . Upon quenching by exposure to air, the reaction mixture was 
characterized without purification to retain any volatile products. The molecular weight 
distribution in this case was found to be bimodal with the bulk of the sample having 
essentially the same retention time and profile as the oligomerized PB prepared at higher 
catalyst loading (M, = 9.0 x 10 3 , PDI = 1.72). An additional small peak was detected 
with retention time indicative of very low M „ oligomers, which were probably removed 
from the first reaction mixture during purification. Because the M n of the 

depolymerized product was insensitive to the change in catalyst loading, it may be 
assumed that incorporation of catalyst as end-groups is not the major determinant of the 
decrease in molecular weight. 

As discussed previously, the spacial placement of double bonds in 1,4- 1,2- 1,4 
triads favors formation of cyclopentene and cyclohexene derivatives via ring closing 
metathesis (RCM). It is proposed here that these imperfections are the major determinant 



65 



of the product's molecular weight via the route illustrated in Figure 3-10. RCM of the 
backbone olefins (bonds a and b) of a 1,4- 1,2- 1,4 triad extrudes vinyl cyclohexene with 
relatively infinitesimal decrease in chain size. This molecule however can then act as 
CTA via cross-metathesis of the vinyl group with the polymer backbone, cleaving the 
chain. Likewise, RCM of the vinyl group of the triad with an adjacent backbone olefin 
(bonds a and c) results in direct chain scission. 




Figure 3-10. Conversion of 1,4- 1,2- 1,4-triads to end-groups. 



Equilibration of the termini toward the most probable distribution should produce 
polybutadienes terminated by vinyl, cyclohexenyl and cyclopentenyl groups with average 
molecular weight proportional to the mole percent 1,2 linkages. For example, the 
production of entirely linear chains via the proposed route from polybutadiene initially 
containing 0.5% 1,2 linkages would result in DP = 200, Mn =1.1 x 10 4 g/mol. The 



66 



presence of the proposed endgroups is supported by the occurrence of the same small 
unresolved multiplet(5.64 ppm) in the ‘H NMR spectra of all depolymerized samples. 
Corroborating evidence could be obtained via bulk depolymerization of a series of 1,4- 
polybutadienes with varying levels of 1,2-linkages. 

There is no doubt the increase in entropy associated with conversion of UHMW 
PB to lower molecular weight macrocycles contributes to this depolymerization. 
Synthetic ultra high molecular weight polymers may currently only be produced 
kinetically by capitalizing on the release of energy via events such as alleviation of ring 
strain or reduction of bond order. This -AH must be sufficiently large to overcome the 
obvious major -AS inherent in the process of connecting a large number of small 

* 5*7 

molecules by covalent bonds.' It is essential that no mechanisms exist for reversing the 
polymerization under the conditions employed. An example of such a mechanism is the 
thermal depolymerization of polystyrene at or above the ceiling temperature. 

ROMP of cycloctadiene results in an equilibrium mixture consisting of high 
molecular weight polybutadiene, cyclic oligomers, and monomer. The same equilibrium 
mixture may be reached whether starting with pure high polymer, pure monomer, or a 
mixture of all species. Intramolecular metathesis depolymerization of ultra high 
molecular weight polybutadiene by conversion to macrocycles provides a thermoneutral 
route to increase entropy. Once metathesis is initiated in the system reported here, the 
system should move towards the appropriate equilibrium for this catalyst/polymer 
combination in the bulk state. Contributions from these processes cannot be excluded. 



67 



Bulk Depolymerization of PB to Produce Telechelics 

Marmo reported the preparation of moderate molecular weight end-functonalized 
polybutadienes by metathesis depolymerization utilizing molybdenum catalyst 1.2b . 60 
Due to the inability of the catalyst to effect cross-metathesis between the CTA and the 
polymer in the bulk state, a two step process was employed. The polybutadiene was 
depolymerized by intramolecular metathesis in toluene prior to addition of the CTA, 
after which the toluene was evaporated to favor the formation of higher molecular 
weight oligomers at the expense of the cylics generated during solution 
depolymerization. In this manner, end-functionalized polybutadienes with M„ = 1.1 x 
10 3 and 2.6 x 10 3 g/mol were generated. 

The unique reactivity of the ruthenium catalyst with polybutadiene in the bulk 
state may be exploited to accomplish depolymerizations without the need for solvent. In 
order to demonstrate the possibilities, polybutadiene was depolymerized by ethenolysis 
in the bulk state. The polymer was cut into small chunks, intimately mixed with catalyst 
(f 3.il /r i.41 = 400) and agitated until it could be magnetically stirred. Exposure to 100 
psig C 2 H 4 (48 hours reaction time) resulted in a rapid decrease in viscosity. Analysis by 
GC and NMR were consistent with the results obtained from ethenolysis in solution. 

The M n calculated from GC was 147 g/mol, indicating lower conversion than obtained 
in solution at the same pressure. This is justifiable in terms of relative concentrations, 
where dilution of PB while maintaining constant ethylene pressure increases CTA/PB 
repeat unit, favoring higher conversion. 



68 



It is highly desired to produce telechelic products consisting entirely of linear 
chains with each and every chain end containing useful functionality, which may be 
incorporated via the CTA. This corresponds to a number average functionality ( F „ ) 
equal 2.0. While competing processes prevent the realization of this ideal case, 
optimization of reaction conditions can greatly reduce the adverse effect on F n - 2 ’ 41 
Polymer chain back-biting producing cyclic species devoid of the desired end-groups 
may be diminished by using the most concentrated conditions possible. Ill-defined end 
groups resulting from termination due to catalyst decomposition may be avoided by 
choosing a robust catalyst, such as L4. The catalyst loading should be kept as low as 
possible as each catalyst molecule introduces two unwanted end-groups (cf Figure 3-9). 

Bulk depolymerization of PB in the presence of CTA should give essentially the 
same results as the ROMP of COD with CTA. It must however be noted that if the 
ubiquitous 1,2-linkages are indeed converted to end groups the F n will be adversely 
affected. Nonetheless, two ester-terminated PB’s (ETPB) of differing molecular weights 
were prepared (Figure 3-11). Ethyl 10-undecenoate ( 3.3 ) and catalyst ([3.31/11.41 = 400) 
were combined and placed under vacuum for 24 hours to prepare the dimer (3.4). This 
dimer/catalyst mixture was then added to chunks of PB ([31]/[L4] > 4000, r3.il/f3.31 = 
20 and 40) and the heterogeneous mixtures were agitated occasionally until stirrable. 
After 24 hours, the reactions were quenched by exposure to air and the products were 
dissolved in toluene, stirred with silica and filtered to remove catalyst residue, and the 
solvent removed under reduced pressure. The molecular weights of the products were 
characterized by GPC and 'll NMR, the results of which are summarized in Table 3-2. 
The values obtained by GPC are higher than the predicted values as expected due to 



69 



calibration of the instrument with polystyrene standards. However, the ratio of the 
measured molecular weights for the two products is approximately the same as the ratio 
of the predicted values. 



Sample 


PB/CTA 


(lO 3 )m„ (g/mol) 






theory 


'HNMR 


GPC [PDI] 


X5 


20 


1.478 


1.5 


4.1 [1.76] 


3d> 


40 


2.560 


2.6 


6.3 [1.70] 



Table 3-2. Characterization of ester-telechelic PB’s. 



The *H NMR spectrum of 3^ is shown in Figure 3-1 1 along with the reaction 
scheme. The inset is a magnification of the diminutive signal attributed to the proposed 
cyclohexenyl and/or cyclopentenyl end-groups created from 1,4- 1,2- 1,4 triads (cf Figure 
3-8). 

Conclusions 

Catalyst L4 has been shown to be highly effective in various types of metathesis 
depolymerization schemes. Its unprecedented stability in the presence of ethylene 
qualifies it as a suitable homogeneous catalyst for ethenolysis. The remarkable reactivity 
of the catalyst with solid polybutadiene allows for the first solvent free depolymerizations 
of ultra high molecular weight polybutadiene and this method can successively be 
incorporated into schemes for preparing end-functionalized oligomers. 



70 





Figure 3-11. 1 H NMR of ester-terminated polybutadiene 3.5 . 



CHAPTER 4 

TANDEM HOMOGENEOUS METATHESIS/ HETEROGENEOUS 

HYDROGENATION 



Much research has been dedicated to the heterogenization of homogeneous 
catalysts. 62 Theoretically, the advantages of homogeneous and heterogeneous catalysis 
can be combined. Advantages of heterogeneous catalysts include long-lived activity, 
facile mechanical separation from substrates and products, and recyclability. These 
attributes have made them highly effective in batch processes common in industry. 
Homogeneous catalysts often offer greater sight selectivity, greater activity, and operate 
under milder conditions. Binding a homogeneous catalyst to insoluble substrates may 
combine the positive aspects of both types of systems. Immobilization produces hybrid 
heterogeneous catalysts, which are analogous to homogeneous complexes on the 
microscopic scale. However, industrial research on this topic has decreased dramatically 
as the desired benefits are seldom realized. Applications are limited by difficulty in 
preparation, characterization, slower rates, failure to achieve site isolation, and catalyst 
leaching. Regardless, as the knowledge base in this area increases, this class of catalysts 
may bear useful systems. 

In this chapter, some observations of the reactivity of catalyst L4 supported on 
silica are reported. Preliminary attempts to prepare a well-defined hybrid of L4 tethered 
to silica resulted in the formation ill-defined ruthenium species which are efficient 
catalysts for olefin migration. This serendipitous observation lead to the realization of a 
new system for one-pot homogeneous metathesis/heterogeneous 
hydrogenation. It has been proven effective as a synthetic transform for constructing C- 
C bonds and in the preparation of saturated aliphatic polymers. 



71 



72 



Immobilization of a Well-defined Ruthenium Complex on the Surface of Silica 

Catalyst 1A is effective in the solution and bulk depolymerization of 
polybutadiene(PB) to produce functional telechelics. Its robust nature allows good 
yields of hexadiene through the ethenolysis of PB. However, in the presence of excess 
C 2 H 4 , L4 decomposes in solution most likely via bimolecular pathways. If the limiting 
conversions for ethenolysis, described in Chapter 3, are a direct result then 
immobilization of the catalyst on a bulky substrate might increase conversions by 
extending catalyst lifetime. The immobilization of catalysts similar to L3 by attachment 
to crosslinked polystyrene beads via phosphine tethers has been reported 6 ’ as shown in 
Figure 4-1. This hybrid catalyst proved active for the metathesis of 2-pentene and could 
be recycled. However, turnover numbers were substantially lower than for the 
homogeneous analogue, which may be attributed to slow diffusion of the substrate and 
product through the cavities of the tortuous beads. 

The adverse effect on rate, due to slower diffusion, might be diminished by 
attaching the catalyst to the exterior of particles rather inside pores. The method used by 
Grubbs was modified to tether the catalyst to silica gel particles rather than cross-linked 
polymer beads (Figure 4-2). Simple chromatographic silica gel was oxidized with 
piranha wash to increase the surface hydroxyl content and alkyl-bromide tethers were 
attached by a well known method 64 for preparing hydrophobic silica. A fraction of the 
bromide groups were converted to phosphine groups via Sn 2 attack by LiPCy 2 . Reaction 
with Cl 2 (Ph 3 P) 2 RuCHPh should give the immobilized, well-defined ruthenium complex 
by displacement of one aryl phosphine ligand by a tethered trialkyl phosphine. The 
insoluble product was washed repeatedly with toluene to remove any homogeneous 
species. Grubbs has reported that mixed phosphine versions of ruthenium metathesis 
catalysts like that proposed here are less active. 6 ’ The accepted mechanism for olefin 
metathesis using L4 involves dissociation of one phosphine ligand to provide an open 



73 



active site. Should the trialkyl phosphine dissociate in this system, catalyst loading 
would be lowered (catalyst leaching). It has been shown that trialkyl phosphines have a 
stronger affinity for the ruthenium center than triaryl phosphines. It was hoped that this 
tendency could be exploited to decrease catalyst leaching. 




Figure 4-1. Immobilization of L3 on cross-linked polystyrene. 



To qualitatively assess the metathesis activity of this system, 1-octene was added 
and the heterogeneous mixture was stirred for 24 hours after which an aliquot was taken 
for GC analysis. A peak was seen with the same retention time as an authentic sample 
of fran.s-7-tetradecene, indicating that metathesis had occurred. In addition several other 
peaks were seen including one with essentially the same retention time as unreacted 1- 
octene. However, 'H NMR showed the complete disappearance of resonances from 
external olefins. Given the low apparent yield of 7-tetradecene and the appearance of 
various side products, the complete conversion of external olefin to internal olefins must 
have occurred by competing chemistry other than metathesis. 



74 




OH 

— OH 

+ 



OH 




Br 




LiPCy 2 



PPh 3 





Cl/,, 


.Ph 






°, Cl- R 


u — ^ 




CN 


-<A 


CI 2 (PPh 3 ) 2 RuCHPh 


CN 


O 

CO 


/'^> PCy * 
O ' J 3 


O 

CO 




Figure 4-2. Immobilization of ruthenium complex on silica via phosphine tethers. 



Catalytic Activity of L4 Adsorbed on Silica 

Attempts to obtain well-defined, single-site catalysts by tethering sometimes 
meet with failure due to incomplete coverage of the silica surface with the tethers. 65 
Highly reactive species may then result from deposition of metal complexes and metal 
particles on the silica surface itself. It was proposed that the side products obtained 
above might be the result of such competing chemistry. 

1-Octene and catalyst JL4 (0.25mole%) were combined with chromatographic 
silica gel inside an argon-purged dry box and stirred vigorously 24 hours. The silica gel 
became reddish-brown and the liquid became colorless signifying that the complex was 
adsorbed onto the silica surface. 'H and l5 C NMR spectra of the liquid were consistent 
with linear alkenes with double bonds in internal positions. Analysis of the product 
mixture by GC/MS showed the presence of a distribution of products with masses 



75 



corresponding to a homologous series of straight chain aliphatic alkenes from C 6 to Ci 6 . 
Removal of the product mixture and addition of fresh 1-octene to the heterogeneous 
catalyst residue gave similar results. 

It is proposed that the process for forming this product mixture consists of at 
least two steps, one of which is olefin migration. Contributing processes may be olefin 
dimerization. 66 ' 69 Figure 4-3 shows dimerization via olefin insertion followed by 13- 
hydride elimination to give dimers, steps that are common to Ziegler-Natta 
polymerization. With certain catalyst systems, the rate of (3-hydride elimination may be 
fast enough to favor formation of dimers as the major or even exclusive products. Paths 
A and B differ in the initial regioisomer formed from two different modes of insertion. 
Olefin insertion into the resulting metal carbon bond may also occur with regioselective 
competition. The product distribution may be further broadened due to competing 
regioselective elimination. Only linear products V and VI are consistent with the 
products obtained using L4 adsorbed on silica. The presence of odd carbon species and 
the absence of branching in the product mixture for L4 on silica suggest an alternate 
mechanism(s). Another possibility is a dimerization mechanism, which proceeds 
through metallocyclopentanes, but this also gives branched alkenes. 

A more likely combination of events is olefin migration in tandem with 
metathesis as depicted in Figure 4-4. As an example, 1-octene may be isomerized to 2, 
3, and 4-octenes through olefin migration. Metathesis of this mixture would result in 
olefin products ranging from C 2 to C !4 with the olefins placed from the 1 to the 7 
position. Further, the products larger than C[ 4 could be formed by olefin migration to 
the 8 position in any given alkene followed by metathesis. The combination of these 
two mechanisms results in alkenes with both odd and even number of carbons. 



76 




MLn 




R 




IV 



B 




1 ' R 

VI 

Figure 4-3. Example mechanism for olefin dimerization. 

The precatalyst ChCPhrPERuCHPh used in the attempted preparation of the 
tethered catalyst above was evaluated under the same conditions, i.e., adsorbed directly 
on silica. The unmodified catalyst is not active for the metathesis of acyclic olefins. In 
concord with the proposed mechanism here, 1-octene was isomerized to a mixture of 
octenes but no other products were detected. The obligatory metal hydride likely 
originates from reaction of the ruthenium complexes with silanol groups of the silica. In 
a related process, co-unsaturated alcohols are isomerized to aldehydes due to proposed 
Ru-H species formed in situ from reaction of L3 with hydroxyl functionality. 28 Given 
the inability of 2 to metathesis acyclic olefins, higher olefins in the tethered system must 
have been produced by the desired well-defined hybrid version of 1A. Metathesis with 





77 



1.4 directly adsorbed on the silica is perhaps due to intact metal carbene, which has been 
simply physisorbed. 




Olefin 



Metathesis 



U 



+ 




+ 



C 3 - Cg a-olefins 






n m 

C 4 - C ]4 Alkenes 



n, m, y = 0 - 5 



n 5 



Olefin 

► 

Migration 






Olefin 



Metathesis 






6 6 



Other alkenes 



C 15 Alkene 

Figure 4-4. Proposed olefin metathesis/migration to produce C 4 - C| 6 internal alkenes. 



The result of this process may be compared to the overall result of the latter 
stages of the Shell Higher Olefin Process (SHOP), Figure 4-5. 2 In the first step, ethylene 
is oligomerized to a-olefins. These alk-l-enes are isomerized to an equilibrium mixture 
of internal alkenes. These are then passed over a metathesis catalyst to produce a 



78 



statistical distribution of linear alkenes. The C 11 -Q 4 fraction is separated and the 
remaining fractions are recycled through the process. The Ch-Ch alkenes may be 
converted to detergent alcohols via hydroformylation. The olefin metathesis and 
isomerization steps might be conducted in one vessel using 1A adsorbed on silica. If 
this system should be proven robust, this could also be a batch process. 



Alk-l-enes 



1 



Isomerization 



1 



Metathesis 



1 



> C14 Alkenes 
— 



Fractionation 



< Cl 1 Alkenes ■ 



I 1 



Cl 3- 14 Alkenes 



Cl 1-12 Alkenes ■ 



Figure 4-5. Block diagram of later stages of Shell Higher Olefins Process . 2 



Hydrogenation with the Residue of Catalyst L4 Adsorbed on Silica 

Catalytic olefin migration and hydrogenation mechanisms have as common 
intermediates metal hydrides. Ruthenium hydride complexes have been shown to be 
highly effective hydrogenation catalysts . 70 ' 71 If this system indeed forms Ru-H species, 
then it should prove an effective catalyst for olefin hydrogenation. In two separate 
pressure vessels, tran^-7-tetradecene and catalyst L4 (0.25 mole %) were combined. To 
one vessel was added 1 g chromatographic silica gel. The two mixtures were stirred for 
1 hour and then exposed to 120 psig H 2 for 14 hours with vigorous stirring. *H NMR 



79 



showed the complete disappearance of olefin in the reaction containing silica. The 
product mixture prepared without silica was less than 30% reduced. 

Concurrent with these observations was a report utilizing the residue of a 
ruthenium metathesis catalyst for hydrogenation. McLain 72 et al. reported a one-pot 
procedure for producing an ethylene/methyl acrylate copolymer by the ROMP of an 
ester-functionalized cyclooctene catalyzed by L3, and then hydrogenating by simply 
applying hydrogen pressure to the completed ROMP reaction system as shown in Figure 
4-6. It was assumed that the catalyst residue was converted to HRuCl(PCy 3 ) 2 . H 2 
pressures of at least 400 psi were required to achieve >99% reduction. Lower 
conversions in the absence of large excess H 2 were contributed to decomposition of the 
suspected highly unsaturated hydrogenating species. One possible route is the cleavage 
of solubilizing phosphine ligands via the ubiquitous P-C oxidative addtion often seen in 
transition metal phosphine adducts. ' The result may be species that are not active in 
hydrogenation and/or precipitation of ruthenium species from solution. Precipitation 
may result in aggregates with diminished surface area and therefore fewer available 
reactive sites. Addition of silica may produce ruthenium complexes that are not 
susceptible to this decomposition pathway. Adsorption of the catalyst residue from a 
homogeneous solution not only provides for high surface area (more available reactive 
sites) and diminishes formation of aggregates but also perhaps initiates the reduction by 
forming the obligatory Ru-H species. 




Figure 4-6. One-pot ROMP/hydrogenation to produce model ethylene/acrylate 

copolymer. Ethylene run length varies (n = 6,7,8) due to unsymmetrical 
nature of monomer. 



80 



A brief model study demonstrates the effect of varied H 2 pressure on the 
effectiveness of the heterogeneized catalyst L4 for olefin hydrogenation, trans-1- 
Tetradecene and 1A were combined as above and stirred one hour to allow complete 
dissolution of catalyst. A change from purple heterogenous mixture to reddish-brown 
homogeneous solution indicated dissolution and carbene exchange. Silica was added 
and the system was exposed to a constant H 2 pressure at room temperature with vigorous 
stirring for thirty minutes. Conversions were estimated from 'H NMR, comparing the 
resonances due to the olefinic protons (~5.4 ppm) to those of the terminal methyl 
groups(~0.8 ppm). As can be seen from Figure 4-7, reaction can be driven to 
completion within 30 minutes at low to moderate pressures. 



5 5 


Ru, silica \ / \ ^ 


M ' V/ 

H 2 12 


Pressure H 2 (psig) 


Conversion 


100 


41 


150 


90 


200 


>99 


250 


quantitative 



Figure 4-7. Hydrogenation of rrans-7-tetradecene. Conversion estimated from ! H NMR. 

This hydrogenation may be incorporated into a two-step, one pot procedure for 
forming C-C bonds at pressures that are convenient in a laboratory environment without 
the need for specialized equipment. In the first step, carbon skeletons are constructed via 
metathesis. In the second step, silica is added to the reaction mixture and the system is 
exposed to hydrogen to reduce the olefinic product(s), forming the desired C-C single 
bond. For this procedure to be successful, it is necessary that hydrogen is applied 



81 



immediately after addition of silica to preclude homologation via olefin migration and 
further metathesis catalyzed by the ruthenium residue as it is adsorbed on silica. 

To confirm the utility of this process, the olefinic ester 41 was dimerized 
quantitatively by U4 (Figure 4-8). Upon completion of the reaction, silica was added and 
the system was exposed to 120 psig FU. After mechanical separation of the catalyst 
residue/silica composite, the long chain a,co-diacetate 42 was obtained in >95% isolated 
yield. 



OAc 



1.4 



AcO 



-C 2 H 4 



v / \ ^ -u / \/OAc Add silica AcCU / \^OAc 
8 8 n 2 18 



Ho 



Figure 4-8. Tandem homogeneous metathesis/heterogeneous hydrogenation to produce a 
long-chain diester. 



It has been found desirable by coworkers in our laboratory to prepare bipolar 
surfactants of the type 45 (Figure 4-9) as they can be polymerized using Langmuir trough 
techniques to produce a 2-dimensional cross-linked network polyaniline. The 
methodology reported herein was exploited in constructing the long aliphatic segment of 
4.5 . The diene 43 might be expected to give a mixture of products due to reaction of the 
styrenic double bond (bond b) in addition to the a-olefin (bond a). As predicted based on 
recent reports of selectivity in the metathesis of substrates with varied electronic and 
steric modifications, bond b is unreactive under the conditions employed for metathesis 
catalyzed by L4. Crowe 5 ’ investigated the reactivity of olefins with varied steric and 
electronic modifications including allyl silanes, acrylonitrile, and styrenes, where cross- 
metathesis of these substrates with alkyl olefins proved to be selective. [3-Alkyl styrenes 



82 



were also shown to be of lower reactivity. Perfluoroalkyl and some fluoroalkyl 
substituted olefins have also been shown to be inert towards catalyst 44 due to lowered 
electron density of the double bonds. Bond b of substrate 43 also proved to be 
unreactive at room temperature in the presence of L4. Not only is it sterically crowded, 
but the electron density of bond b is diminished by the nitro-substituted benzene ring. As 
a result, triene 44 can be obtained in 95% isolated yield after flash chromatography. 



80 % cis 



44 

C 2 H 4 



o 2 n 

4.3 



\0 







nh 2 22 nh 2 




4.4 



Figure 4-9. Attempted synthesis of bipolar aniline surfactant. 



In situ hydrogenation of 44 revealed some aspects of relative selectivity of the 
system reported here. Hydrogenations were conducted at room temperature (24 hrs) and 
conversions were estimated from 'H NMR. At 120 psig H 2 , bond a is quantitatively 
hydrogenated. However, bond b(cis) is only converted by approximately 90% while 
b(trans) is virtually unreacted. At 800 psig, conversions are near quantitative and 
approximately 80% for b(cis) and b(trans), respectively. This is in keeping with the 
selectivity typically observed in the hydrogenation of sterically hindered cis/trans isomers. 
In both cases, trace peaks were detected ('H NMR) which might be assigned to the aniline 
moiety resulting from reduction of the nitro group. Reports of hydrogenation with 



83 



various homogeneous catalysts of general formula Cl n (R 3 P) m RuL n indicate pressures in 
excess of 80 atm are required to obtain good to excellent conversion of aromatic nitro 
compounds to aniline derivatives in practical reaction times. 74 Optimization of reaction 
conditions might allow higher conversion in hydrogenation of nitro-functionalized 
substrates providing an indirect route to primary amines without the need for 
protection/deprotection schemes. 

Preparation of Saturated Polymers by Tandem Homogeneous 
ADMET/ Heterogeneous Hydrogenation 

Olefin metathesis polymerization has proven to be quite broad in scope for the 
preparation of well-defined polymers and hydrogenation of these provides additional 
polymers with unique microstuctures. 46-51 Hydrogenation also produces new polymers 
with differing physical and chemical properties, including greater oxidative stability. 
These syntheses typically involve first the synthesis and separation of unsaturated 
polymers from the metathesis catalyst followed by a second hydrogenation step. 
Hydrogenated polymers from norbenene and derivatives of norbornene have been 
produced commercially as Zeonex™ since 1991 . 

Preparation of End-Functionalized Polyethylene by ADMET/Hydrogenation 

A useful application of metathesis/hydrogenation is the production of 
commercially valuable telechelic polyethylenes. This has been demonstrated in the 
production of hydroxy-telechelic polyethylene via diimide reduction of intermediate 
telechelic polyalkenamers. 61 The polyalkenamers were produced through ROMP of 



84 



cyclooctadiene in the presence of 1,2-disubstituted olefins, as described in Chapter Three. 
Similar polyalkenamers may be produced by ADMET polymerization in the presence of 
functionalized CTA’s. 

The new tandem homogeneous metathesis/heterogenous hydrogenation technique 
utilizing catalyst L4 proved effective for preparing end-functionalized polyethylene (4.6) 
(Figure 4-10). 1 ,9-Decadiene was polymerized using E4 under typical ADMET 

conditions in the presence of 0.2 equivalents 9-decenyl acetate (4.1) as a chain transfer 
agent. Subsequent heterogeneous hydrogenation requires the addition solvent to aid 
dispersion, but dilution in the presence of active metathesis catalyst would favor 
formation of lower molecular weight cyclics at the expense of linear species. Therefore, 
the polymer was intimately mixed with silica and the system was exposed to 120 psig H 2 
for 30 minutes prior to addition of toluene. Extensive hydrogenation had already occurred 
at this point to yield a product with sufficient polyethylene character to render it poorly 
soluble in toluene. Hydrogen pressure (120 psig) was reapplied for five hours while 
heating in a 90°C oil bath to prevent precipitation of the product. Simple hot filtration 
and removal of the solvent under reduced pressure yielded a water-white solid in near 
quantitative yield. 

Molecular weight measurements agree well with the value predicted Mn = 1 .5 x 
10 3 g/mol based on the ratio of chain limiter to diene in the ADMET reaction. End-group 
analysis ( 1 H NMR) and high temperature GPC (vs PE) gave 1.5 x 10 3 and 1.7 x 10 3 
g/mol, respectively. The polydispersity ( 1 .9) approaches that expected for equilibrium 
condensation polymerization. Examination of the ‘H NMR spectrum (Figure 4-10) in 
the region from 4.8 - 6.0 ppm reveals good conversion in the hydrogenation step, as 



85 



indicated by complete disappearance of signals rising from olefinic protons. Acetate 
content of the product as calculated from * H NMR did not decrease from that expected 
from initial stoichiometry indicating that these groups remain intact throughout the olefin 
hydrogenation. This reaction was repeated, with aliquots periodically removed to 
qualitatively assess the time required for complete reduction. By ‘H NMR, signals due to 
olefinic protons completely disappear within 4 hours. 

Synthesis of a Polyester with Long Aliphatic Segments 

Pennelle 7 ^ and coworkers reported the synthesis of a polyester with regular long 
alkyl segments within the polymer backbone by classical condensation of a long chain 
aliphatic diol with a diacid (Figure 4-11). Their interests lie in the control of lamellar 
crystal structure by introducing regular defects in a polymer with high polyethylene 
character. 

Similar polymers may be constructed by the ADMET/hydrogenation method 
reported here as illustrated in Figure 4-12. The ester diene ( 4 . 7 ) was condensed by 
ADMET and the polymer was hydrogenated as described previously in this chapter for 
hydrogenation of end-functionalized polyoctenamer. The higher molecular weight and 
crystallinity of the unsaturated polyester (fL8) slows its dissolution in toluene, allowing 
the addition of solvent just prior to exposure to hydrogenation without fear of rapid 
metathesis depolymerization. The saturated polyester ( 4 . 9 ) was obtained as a white solid 
in near quantitative yield. The structure shown is not entirely correct as the ester groups 
may by incorporated in head-head, head-tail, and tail-tail distribution, a consequence of 
the unsymmetrical nature of the ester moiety. 



86 






OAc 



1A 

- C 2 H 4 AcO v 



4.1 






.OAc 




d 












" i " — r- — i " " i " " i " i " ' i ' 

6.0 5.8 5.6 5.4 5.2 5.0 



1000X 



b 

a. 



toluene 




6 5 4 3 2 1 PPm 

Figure 4-10. 'H NMR of telechelic polyethylene from ADMET/hydrogenation. 



87 



O O 




E = 0 2 C(CH 2 ) 3 C0 2 



Figure 4-11. Preparation of a polyester with long ethylene run lengths by classic 
condensation chemistry. 

The 'll NMR spectra for polyesters and fF9 are shown in Figure 4-13. Due to 
its highly crystalline nature (AHf = 122 J/g, T,,, = 97°C, DSC) and long ethylene run 
lengths, polymer 4J9 displays very limited solubility in typical organic solvents at room 
temperature, but is freely soluble above 60°C. Thus NMR measurements for this polymer 
were acquired at 90°C in toluene-dg which apparently also increases resolution of the 
overlapping signals corresponding to the protons (3 to the ester group ("a" and "d"). Note 
the signals ("g", upper and "e", lower) indicating conversion of vinyl ends groups of 
polymer 4.8 to methyl end groups upon hydrogenation. High temperature GPC (vs 
polyethylene standards) gave M» = 1.0 xlO 4 g/mol (PDI = 1.7) for polymer 4d) and these 
values agree within experimental error with those obtained for 4J5 under the same 
conditions. 




4/7 4J5 4j} 

Figure 4-12. Preparation of saturated aliphatic polyester by ADMET/hydrogenation. 



88 




6 5 4 3 2 1 PPm 

a O c 




b d e 



Toluene 




Figure 4-13. 'H NMR spectra of 4JS (upper, CDCI 3 ) and 43) (lower, toluene-dg, 90°C). 



89 



Conclusions 

The method reported herein is a facile one-pot route for forming carbon-carbon 
bonds in the synthesis of polymers and small molecules. No additional catalysts or 
reagents are needed for hydrogenation other than inexpensive silica gel and hydrogen 
gas. Addition of silica as a support for the ruthenium complex not only facilitates 
hydrogenation at low to moderate pressures but also simplifies purification of products. 
No purification step is required prior to hydrogenation and the catalyst may be removed 
by simple mechanical filtration. It shows promise both commercially in the preparation 
of telechelic polyethylenes and in the preparation of model polymers. 

The ester moieties of the polymers in Figures 4-11 and 4-12 can be easily 
incorporated into polymer backbones with long ethylene run lengths by classical ester 
condensation chemistry. However, other functional groups that may provide deeper 
insight into the engineering of crystal structures may prove difficult or even impossible 
to incorporate into a high molecular weight polymer. The method of 
ADMET/hydrogenation is very well suited to producing model polymers of this type, as 
symmetrical dienes with a wide variety of functional groups at their center may be 
constructed by typical organic transforms. The method for incorporating the functional 
group into the polymer is the same mild procedure regardless of constitution. 
Applications of this process to the preparation of such model “functionalized 
polyethylenes” will be described in the following chapter. 



CHAPTER 5 

THE SYNTHESIS OF MODEL ETHYLENE/POLAR MONOMER 
COPOLYMERS VIA ADMET AND HYDROGENATION 



Since macromolecular compounds are, almost without exception, polymolecular 
mixtures, one does not deal with uniform materials, as is the case for low 
molecular compounds. In the most favorable case, such a material is 
polymeruniform, that is, a mixture of macromolecules of the same structure, but 
different chain lengths. 

Hermann Staudinger 76 

Flawless control in the construction of discrete molecules of high molar mass has 
been exhibited only in nature, where the macromolecular engineer and the legislator of 
laws governing the implication of chemical transforms are one and the same. The 
indispensability of this level of control is demonstrated by the horrendous outcome of 
rare mistakes in these biological polymerizations. DNA, one of the many forms of 
natural polymers, stores coded information in its perfect sequences of nucleic acid repeat 
units and these codes are read and used to replicate the perfect polymers as well as to 
produce others. 77 Numerous cancers, birth defects, and diseases are now linked to what 
are errant polymerizations induced by mutated DNA or misreading of its codes. 

Man’s clear mediocrity in this arena is demonstrated by current practices to 
combat the polymerizations gone awry. These include “shotgun” approaches such as 
chemotherapy and radiation treatments, with their well-known grim physiological side 
effects due to the indiscriminate statistical termination of both properly and improperly 
functioning polymerizations. All modern synthetic schemes are governed to some extent 



90 



91 



by statistical elements, limiting the control which chemists may exert over the outcome of 
their chemical products, including polymers. Man’s subjugation to statistics has even 
prompted some to ascribe the development of nature’s perfect polymerizing machines to 
a statistical process termed evolution. 

With regard to chemical homogeneity, the aesthetic simple structures assigned to 
synthetic polymers are rarely accurate depictions. For example, even in the absence of 
side reactions such as chain transfer and termination, butadiene may be incorporated into 
a polymer backbone via chain polymerization in numerous regiochemical and 
steriochemical modes of propagation 37 which are certainly not adequately represented by 
the single generic structure depicted in Figure 5-1. The monomer may be incorporated 
into the chain in 1,4 (both cis and trans) and 1,2 fashion and sequences of 1 ,2 repeats may 
also display varying degrees of tacticity. An example of a highly efficient near perfect 
polymerization is Zeigler-type polymerization of butadiene with Neodymium catalysts by 
which ultra-high molecular weight polybutadiene, i.e., on the order of 10 6 g/mol may be 
prepared with 1,4 content as high as 99.5% (> 98% cis). 78 This is truly an amazingly well 
controlled reaction as any organic chemist would be thrilled to conduct transforms 
yielding crude product mixtures that are 99.5% pure. However, in this case, the 0.5% 1,2 
units cannot be removed from the product mixture as they are covalently bound to the 
pure portion of the product. By comparison with small molecule chemistry, while the 
bromination of butadiene may afford a mixture of products where 1,2 and 1,4 addition 
compete, distillation, chromatography, or a combination may separate the desired 
product. 



92 



While pure polymer microstructure is not a prerequisite for technical utility, the 
availability of polymers with progressively greater chemical homogeneity has allowed 
deeper insight into structure/property relationships. With judicious choice of catalyst and 
monomer, ADMET polymerization has been shown to produce polymers with pure 
microstructure within spectroscopic detection limits. The combination of the clean 
ADMET polymerization with mild chemoselective olefin hydrogenation has been shown 
to be an effective method for preparing model polymers that are polymer uniform. The 
realization of the facile one-pot procedure for ADMET/hydrogenation reported in 
Chapter 4 served to streamline this process for producing model polymers. The 
remainder of this chapter describes the use of this process to prepare models for 
copolymers of ethylene with vinyl monomers including vinyl acetate, vinyl chloride, 
styrene, and acrylates and subsequent preliminary thermal analysis of these polymers. A 
brief description of the commercial polymers for which these polymers serve as models is 
first given in order that the reader may grasp the significance of the results reported here. 



Polymerization: Products from different modes of addition inseparable 

1,4 1,2 




Addition 



Polymerization 




Bromination: Products from different modes of addition easily separable 




Figure 5-1. Macromolecular versus small molecule chemistry of butadiene. 



93 



Ethylene/Polar Monomer Copolymers 

Polyethylene has been produced commercially since 1942 79 and has grown to be 
the world’s largest volume commercial polymer. 80 A multitude of grades exist, 
depending on the method of manufacture, with varying molecular weights and levels and 
types of branching granting a broad scope of properties. Variation of the polymer 
constitution via copolymerization of ethylene with various substituted ethylenes 
dramatically increases the range of new materials. 

Commercial products in the class of ethylene/polar monomer copolymers include 
as comonomers vinyl acetate, acrylates, acrylic acid, etc. Inclusion of vinyl acetate or 
acrylates, for example, yields products with decreased crystallinity and increased room 
temperature flexibility, impact resistance, and optical clarity relative to ethylene 
homopolymer. Variation of the comonomer content also greatly effects properties 
allowing application of EVA in such diverse end uses as packaging films, gloves, 
squeeze toys, and hot-melt adhesive formulations. 37 

Ethylene/polar monomer copolymers are produced commercially by the same 
high-pressure free radical processes utilized for low density polyethylene. As a result of 
the nature of the free radical polymerization, the polymer microstructures are ill-defined 
with uncontrolled branching and broad polydispersities due to chain transfer reactions. 
Branching can occur via hydrogen-atom extraction (chain transfer) by the propagating 
chain end both intra- and intermolecularly. The former leads to short chain branching 
(SCB) while the latter leads to long chain branching (LCB) as shown schematically in 
Figure 5-2. 



94 



Intramolecular: Short Chain Branching 




Intermodular: LongChain Branching 




CpH 



2 n 4 





Figure 5.2. Origin of SCB and LCB in free radical polymerization. 



Some polar monomers contain hydrogen atoms which are more easily abstracted 
and therefore copolymers containing these have a higher degree of branching. For 
example, both the methine proton of the vinyl acetate repeat unit and the methyl protons 
of the acetoxy group in either the monomer or the repeat unit provide sites for additional 
branching. 

The microstructures of these copolymers are further complicated by statistical 
incorporation of the comonomers into the polymer backbone. The comonomers may be 
incorporated in different amounts according to their relative concentrations and 
reactivities. The sequence distributions may also vary between alternating, blocky, and 
random as governed by the product of the monomer reactivity ratios ri and r 2 as defined 
below. For a system consisting of two monomers, Mi and M 2 , kn is the rate constant for 
a chain ending in M| adding to monomer M|, k) 2 is the rate constant for a chain ending in 
Mi adding to monomer M 2 , and so on. 

k\ | k 2 2 

' ~ k 2 ~ k 

ACj2 ^21 







95 



For ri = r 2 s 0, the monomers tend to be incoiporated in an alternating fashion 
where each Mj is followed by M 2 and vice versa. When ri>l, r 2 > 1, the chain end has a 
higher tendency to add to the monomer from which it was derived resulting in blocks, or 
extended sequences, of each monomer. In the case of i'i r 2 = 1, the polymerization is 
termed an ideal polymerization and the two monomers are incorporated in a random 
fashion. This is generally the case for ethylene vinyl acetate, ethylene/acrylate and some 
other commercial ethylene polar monomer pairs in the typical temperature/pressure 
regimes used in commercial free radical copolymerizations. 

Among commercial polar monomer copolymers with differing comonomers there 
is broad variability in the molecular weights, PDI's, branch content and identity, and 
sequence distributions. Model copolymers with less variable microstructure should 
greatly ease the evaluation of structure/property relationships. The preparation of such 
models will be described in this chapter. 

Model Polymers 

“The term ‘model polymers’ in its modern meaning now refers to samples with 
often complex architectures in which fluctuations in size, structure, composition, or 
functionalization are minimal.” 83 A polymer chemist, mindful of the limitations of 
modern synthetic chemistry, penned this definition including the key word “minimal”. 
The production of model polymers requires 1) a mechanistic route that is free of 
competing chemistry and is long lived, 2) each and every monomer unit must be 
incorporated in precisely the same fashion, and 3) reaction must be site-specific in that 



96 



intended spectator functionality in the monomer remains as such. In short, only one 
mechanistic event may occur. 

Typical routes used to prepare model polymers include living chain 
polymerizations utilizing ionic, radical, and group transfer chain techniques which have 
been optimized to limit chain transfer, termination, and other side reactions to maximize 
control of the primary structure. These modifications have allowed the construction of 
polymers with narrow polydispersity and of block copolymers with relatively uniform 
block lengths. However, it must be noted that even for polymers with perfect 
microstructure and extremely narrow polydispersity, PDI < 1.03, one has but to observe a 
MALDI spectrum to recognize that the products are still chemical mixtures of 

84 

homologous polymers. 

The synthesis of discrete macromolecules has only been achieved by stepwise 
synthesis where the building blocks are connected one at a time and the resulting product 
is isolated and purified after each connection. 88 Now prolific examples of this repetitive 
approach to macromolecular engineering are the three dimensional dendrimers. 86 This 
repetitive approach has been utilized to prepare discrete polyalkenamers (Figure 5.3) 
identical in repeat structure to polydisperse samples which might be prepared through the 
ADMET or ROMP of 1 , 1 3-tetradecadiene and cyclododecene, respectively. The “dimer” 
is prepared by Wittig reaction after which the product is isolated and divided into two 
portions. One portion is hydrolyzed to the aldehyde and the other is converted to the 
Wittig salt. These may be reacted together and the sequence of events is repeated, each 
time doubling the degree of polymerization. 



97 



Dimer 




Figure 5-3. Repetitive approach to synthesis of polyalkenamers. 

Preparation of Model Polymers via Metathesis 

For the purposes here, the definition for model polymers provided at the 
beginning of the previous section shall be modified. These shall include polymers 
synthesized by alternate methods to obtain a primary structure similar to existing 
polymers but greater chemical homogeneity than is obtainable by traditional means. 



Polybutadiene 




Butadiene/ethylene copolymer 




Vinyl Acetate/butadiene/ethylene terpolymer 




Figure 5-4. Synthesis of model polybutadiene and butadiene copolymers. 



98 



A few examples, illustrated in Figure 5-4, serve to demonstrate the utility of 
metathesis polymerizations in preparing models for polybutadiene and its copolymers. 
Purely linear, branch-free polybutadiene with 100% 1,4 microstructure may be prepared 
by the ADMET of 1 ,5-hexadiene or the ROMP of cyclobutene, 1,5-cyclooctadiene, 1,5,9- 
cyclododecatriene, etc. Polymerization of monomers with a greater number of methylene 
spacers between the double bonds produces models for sequence-ordered copolymers of 
ethylene and butadiene where the exact frequency of unsaturated sites (butadiene repeat 
unit) is governed by the number of methylene spacers in the monomer. A strictly 

oo 

alternating butadiene isoprene copolymer has also been prepared in this manner as 
shown in Figure 5-4. It must be noted that due to the unsymmetrical nature of the 
monomer, the isoprene units may be incorporated in 1,4 or 4,1 fashion; the methyl 
substituent may be in the 2 or 3 position of the isoprene unit. In an analogous example 
utilizing ROMP, Hillmyer 61 prepared a model butadiene, ethylene, vinyl acetate 
terpolymer by the ROMP of 5-acetoxy cyclooctene. In both of the last two examples, the 
number of carbons separating the carbon with a pendant substituent varies between 6, 7, 
and 8. This introduces the important caveat that the production of polymers by these 
methods with perfect repeating sequences requires the use of symmetric monomers. 

Model Polymers via Metathesis Polymerization and Hydrogenation 

The only variable in the repeat structure of ADMET homopolymers is the 
cis/trans ratio of the backbone olefinic bonds. Coupling of ADMET polymerization with 
a clean, quantitative olefin hydrogenation technique constitutes a viable synthetic method 
to produce polymers with perfect repeating sequences. This was first demonstrated 89 



99 



whereby polyoctenamer was fully saturated utilizing stoichiometric diimide reduction to 
a perfectly linear, branch-free model for the largest volume industrial polymer, 
polyethylene (Figure 5-5). The key to this successful, clean hydrogenation was the 
addition of a bulky amine base, tripropyl amine, to prevent side reactions induced by a 
known side product, p-toluene sulfonic acid. A polyethylene with narrow polydispersity 
was later produced 90 by hydrogenation of a polybutadiene obtained via the ROMP of 
cyclobutene. 




TsNHNH 2 , Pr 3 N 
o-xylene, reflux 




Figure 5-5. Production of perfectly linear polyethylene via ADMET/hydrogenation. 



This method was later extended to prepare models to study branching in 
polyethylene. 51 Symmetrical dienes like that shown in Figure 5-6 were condensed by 
ADMET and then hydrogenated by the diimide reduction. The microstructure of the 
resulting polymers is essentially the same as polyethylene with regularly spaced methyl 
branches, where precision in the sequence distribution of methyl substituents is granted 
by the nature of the ADMET reaction. The frequency of the methyl branches is 
prescribed by the value of “n” which may easily be varied during monomer synthesis. 
These polymers and others like them may serve as useful models to study polymer 
crystallization, as more in depth studies follow. 



100 




Figure 5-6. The synthesis of model “polyethylene” with regular methyl branches. 



Similar polymers have been prepared via hydrogenation of alternating 
butadiene/a-olefin copolymers obtained by Ziegler-type polymerization. 91 While the 
frequency and chemical identity of branching is highly controlled, the only possible 
frequency of branching is one branch on every 5 lh carbon. Attempts to prepare models 
with lower counit frequency by chain techniques results in random placement. This 
introduces a second important caveat in these modeling studies: Model polymers with 
“long ” linear aliphatic segments of precise lengths are most conveniently prepared by 
condensation chemisty. Further, ADMET condensation chemistry is perhaps currently 
the most general method, allowing the macromolecular engineer to dictate the precise 
length of the hydrocarbon segment as well as the incorporation of a broad range of 
functionality into the backbone or as pendant groups. 

A simple retrosynthetic analysis (Figure 5-7) of the generic model polymer 
teaches this point. The target polymers have repeat units with precisely defined aliphatic 
segments separated by functional groups “Z”. Utilizing classical polycondensation 
chemistry the polymer can be prepared from the bifunctional monomers shown where 
“X” and “Y” are reactive functional groups which condense to form the desired 
functionality “Z”. This approach will be sufficient for polymers such as polyesters (Z = 



101 



C0 2 ; X, Y = C0 2 H, OH) where the associated chemistry has been well delineated and 
applied commercially for many years. Due to the relation between conversion and M n 

in condensation chemistry, the types of functionality “Z” that may be substituted for ester 
is limited to only those groups that may be constructed in very high yield. 

To illustrate the value of the ADMET/hydrogenation approach in preparing these 
type models, we need only to attempt to design a synthetic route for the same type 
polymer but substitute any functional group “Z” for which polycondensation chemistry 
has not been proven so effective. For example, should it be desired to prepare a polyether 
(Z = oxygen), low molecular weight oligomers might be produced by reaction of a long 
chain dialkoxide (X = OLi) with a dihalide(Y = Br), but the limiting yields of this 
reaction would not allow production of high molecular weight polymers. 




Path B 



Olefin 

hydrogenation 



Path A 



■< 

Classic 

Condensation 




m m 



Path B 



ADMET 





m = (n/2) -1 



Figure 5-7. Classic condensation chemistry versus ADMET/hydrogenation to prepare 
model polymers with long, defined aliphatic segments. 



Instead, the ether moiety could be incorporated into a long chain diene by the 
Williamson ether synthesis and then this monomer condensed to high molecular weight 
by ADMET; hydrogenation then provides the desired polyether. Regardless of the 



102 



identity of Z, the polymerization proceeds to high conversion, provided that Z is 
compatible with the metathesis catalyst. This approach should be quite general as 
ADMET chemistry has been proven effective for the polymerization of dienes containing 
a wide variety of functionality. 36 Some examples of polymers prepared by 
ADMET/hydrogenation are shown in Figure 5-8. 

Poly( p-phenylene octylenef 6 Ethylene/ethylene oxide blocks 41 ' 49 Polyalcohol 50 



OH 




Figure 5-8. Examples of model polymers produced by ADMET/hydrogenation. 



Design and Synthesis of Symmetrical Diene Monomers 
for Ethylene/Polar Monomer Model Polymers 

The preparation of ADMET model ethylene/polar monomer copolymers hinges 
on the availability of highly pure dienes with relevant functionality. A series of 
symmetrical dienes with central pendant functionality were prepared with varied 
functionality and number of methylene spacers separating the functionality from the 
olefin groups. These were obtained in sufficient purity that contaminants could not be 
detected by NMR, GC, and HPLC. 

Central starting materials in nearly all synthetic schemes used to prepare ADMET 
monomers have been co-alkenyl bromides. They are versatile reagents in that they can 
serve as alkylating agents (electrophiles) and as nucleophiles when converted to Grignard 
reagents allowing for the synthesis of a broad range of dienes. These starting materials 



103 



are commercially available with varied number of spacers between the terminal alkene 
and bromo functionality. In all previous ADMET studies, it was not the focus to prepare 
polymers with “very long” aliphatic segments. However, as the target polymers reported 
here are to have long aliphatic segments it was necessary to prepare longer dienes, 
through chain extension of commercially available co-alkenyl materials. Decenyl and 
undecenyl bromide are the longest commercially available starting materials but were 
found to be contaminated by regioisomers (internal olefin) which could not be removed 
in a practical manner. The consequence of using these starting materials would be 
unsymmetrical placement of the double bonds about the central pendant group of the 
diene monomers and consequently irregularity in the periodicity of the polymer. 

Commercially available synthons for these starting materials are decenol, 
undecenol, and undecenoic acid, which are free of detectable regioisomeric impurities. 
Many means are available in the literature for the conversion of alcohols to bromides 
however a mild method must be chosen here to avoid reaction with the terminal olefin. 
The least expensive and simplest techniques, utilizing PL n /Br 2 , PBr n , or HBr as the Br 
source can result in the unwanted side reactions of hydrobromination and bromination of 
the terminal olefin. The co-unsaturated alcohols may easily be converted to the bromide 
by either of two very mild methods, both in good to excellent yields, as shown in Figure 
5-9. In the first method, the alcohol is tosylated and then converted to the bromide after 
only simple aqueous workup without the need for further purification. The second 
method utilizing PPh 3 /CBr 4 achieves the conversion in comparable yields in one step. 
Both reactions are highly atom-inefficient but are facile and typically afford the bromides 
in high isolated yield and purity after distillation. 



104 




5A SJt 



Figure 5-9. Syntheses of co-unsaturated carboxylic acids and bromides, a: Ph 2 P,CBr 4 . b: 
LAH. c: TsCl, pyr. d: LiBr, acetone, e: ®Mg, ®C0 2 .f : ®Mg, ® CuCl, (3- 
propiolactone. 



These alkenyl bromides were chain-extended by reaction of their Grignard salts 
with C0 2 or (3-propiolactone. In the second case, the addition of catalytic Cu(I) favors 
nucleophilic attack at the methylene adjacent to the ring oxygen over attack at the 
carbonyl carbon. These synthetic fatty acids may not be purified on a large scale by the 
classical method of extraction into basic aqueous solution as they are of sufficient length 
to act as surfactants producing intractable emulsions under typical isolation conditions. 
They instead are purified by distillation or flash chromatography. These acids were 
reduced to the alcohols with excess LAH and then converted to the corresponding chain- 
extended bromides via the tosylate as above. Conversion of the bromides to the chain 
extended alcohols, transforms e + b or f+ b in Figure 5-8, could also each be achieved in 
one step by substituting formaldehyde and oxirane, respectively, for C0 2 and (3- 
propiolactone. However, the co-unsaturated acids are useful starting materials for 
preparing additional dienes as will be discussed later. 



105 



Synthesis of Acetoxy-Functional Monomers 

A series of symmetrical dienes with central pendant acetoxy groups (Figure 5-10) 
was synthesized with the goal of preparing model ethylene/vinyl acetate copolymers. 
The shortest acetate-functional diene ( 5.12 , n=8) was prepared by a four-step procedure, 
due to the very low cost of starting material, and the ease of the steps. The first three 
steps were conducted with aqueous extraction as the only workup. The secondary 
alcohol was purified by flash chromatography as acetylation without prior purification 
produced a mixture with impurities that could not be practically separated from the target 
acetate monomer. The remaining secondary alcohols (n = 9, 10 and 12) were prepared in 
good yields in one step via reaction of in situ formed alkenyl Grignard reagents with 
methyl formate, after which acetylation provided the desired monomers. 








5Ji 8 = n 
5J> 9 

5.10 10 

5.11 12 

Figure 5-10. Syntheses of diene monomers with central pendant acetoxy groups, a: 0.5 
eq KH. b : LiCl, DMSO, H 2 0, A. c: LAH. d: ®Mg, ©0.5 eq HC0 2 Me. e: 
AcCl or Ac 2 0, pyr. 



5.12 8 = n 

5.13 9 

5.14 10 
5T5 12 



106 



Chloro- and Phenyl-Substituted Dienes 

Intermediates shown in Figure 5-10 were also converted to symmetrical dienes 
with central pendant chloro and phenyl substituents (Figure 5-11) from which 
ethylene/vinyl chloride and ethylene/styrene model copolymers may be prepared. The 
reported mildest method for converting alcohols to chlorides is a variation of the 
Mitsunobu reaction, utilizing PPh 3 , DEAD and ZnCE where the nucleophile is chloride 
ion supplied by the zinc salt. Multiple attempts to prepare the chloride monomer 5.16 
by this route proved to be low-yielding (< 40% crude yield). Combination of the 
products from each of the reactions yielded a practical quantity of crude monomer (~lg). 
However, an impurity was present in detectable (GC, NMR) amount which proved 
difficult to separate even by HPLC. The combination of excess Ph 3 P in refluxing CCI4 
did however prove effective in converting the secondary alcohol to the chloride in good 
yield. Some elimination was apparent in the ’H NMR spectrum (small broad multiplet 
near 5.4 ppm) of the crude product but, surprisingly, the eliminated product could be 
separated and the pure chloride (100% GC) was obtained after simple flash 
chromatography. 

Dienes substituted by a phenyl ring can be used to prepare model ethylene/styrene 
copolymers. To this end a phenyl-substituted monomer was easily obtained from the 
ketone diene ( 5 . 7 ) also shown in Figure 5-11. The phenyl ring is attached via 
nucleophilic attack by phenyl lithium on the carbonyl carbon and the resulting benzylic 
alkoxide is deoxygenated in situ under Birch reduction-like conditions. 93 



107 




Ph 3 P,CCI 4 

reflux 




1) PhLi 

► 

2) NH 3 , XSLi 

3) NH 4 CI 



Ph 




5.17 



Figure 5-11. Preparation of chloro- and phenyl-substituted dienes. 



Monomers with Pendant Alkoxy-Carbonyl Moieties 

Dienes with central pendant alkoxy-carbonyl groups (Figure 5-12) were 
synthesized as monomers for model ethylene/acrylate copolymers. One of the most 
useful organic transforms for constructing carbon skeletons of this type is the dialkylation 
of enolate ions with alkyl halides. This method has previously been shown to be 
effective in the synthesis of symmetrical diene intermediates for use as ADMET 
monomers. 51 This route, was exploited to prepare the methyl substituted dienes (Figure 
5-5) discussed earlier in this chapter. The diene skeleton is constructed by twice 
alkylating the enolate of ethyl acetoacetate, generated in situ by deprotonation with t- 
butoxide base. Deacylation by simple retro-Claisen reaction yields the desired ester- 
functionalized dienes like the target monomers shown in Figure 5-12. To convert to 
other esters (different R), would simply require trans-esterification or the use of a 
different starting acetoacetate ester. 

Review of the literature indicated that this process for constructing the diene 
skeleton would not be effective for producing the desired larger dienes as dialkylation of 
[3-keto esters with very long alkyl halides has been shown 94 to be a very low yielding 




108 



reaction. As an alternative, previous researchers turned to the dianions of carboxylic 
acids as carbon nucleophiles. These may be generated by reaction of a carboxylic acid. 



with a-hydrogens, with two equivalents of LDA at 0° > T > -15°C and are stable in THF 
solution even at 50°C for several hours. 95 Their use is not complicated by Claisen 
condensation type reactions, multiple sites for C-alkylation, and very low temperatures 
are not required for their generation or alkylation. 



Acrylate-like functionality 

1 ) 2 LDA, -1 5°C 
► - 

n+1 2 ) 

' 'n 

Methyl methacrylate-like functionality 
1) 2 LDA, -15°C 



C0 2 H 



n n 

5T8 n=8 
5T9 n=10 



C0 2 H 



K 2 C0 3 , Rl 

acetone 

reflux 



5.18 



2) XS Mel 






K 2 C0 3i Mel 
» 

acetone 

reflux 



5.23 



1) 2 LDA, -15°C 
.Br 






8 



0 1) 2 LDA, -15°C 

OH ** 



C0 2 H 



2)^Jf 



Br 



r 

8 



8 



C0 2 R 



n n 



5^20 n=8, R=Me 
L21 n=8, R=Et 
5^22 n=10, R=Me 



GO2MG 

5.24 



5.25 



Figure 5-12. Synthesis of dienes with pendant alkoxycarbonyl functionality. 



The needed long chain co-unsaturated carboxylic acids were available in our 
laboratories, as they are intermediates in the synthesis of alkenyl bromides described 
earlier in this chapter (Figure 5-9). The synthetic scheme for preparing ester- 
functionalized dienes is outlined in Figure 5-12. Construction of the diene skeleton 



109 



proved to be high yielding and the products are easily purified by simple flash 
chromatography. The acid groups of these dienes were O-alkylated in near quantitative 
yields by base promoted esterification with excess methyl or ethyl iodide, which has been 
shown in the literature to be a particularly effective method for sterically hindered 
esters. 96 Classic Fischer esterification with sulfuric acid catalysis was not utilized due to 
the susceptibility of the olefinic groups to acid-catalyzed hydration or isomerization. 

The acid dienes may also be alkylated at the a-position, via the same enolate 
chemistry, producing monomers with methacrylic acid-like functionality. It was 
reported 96 that trisubstituted acids, produced by alkylation of the enolates of disubstituted 
acids, could be easily purified by simple aqueous extraction. That is, the starting material 
could be separated from the product by simple preferential extraction of the disubstituted 
acid into basic aqueous solution. Hoping to capitalize on this presumed facile method, 
the acid diene 5.18 was methylated in the a-position. However, the starting material and 
the product, 5.23 , could not be separated as described by aqueous extraction. As 
expected due to the very small structural difference between the starting material and the 
product, 5.23 could not be adequately purified by practical techniques. Only after 
repeated small-scale HPLC purification could a satisfactory sample be obtained. A more 
fascile approach to 5.23 involves twice alkylating the enolate of propionic acid with 9- 
decenyl bromide. The much larger disparity in molar mass and polarity of the starting 
materials and product greatly eased purification. Base-promoted O-alkylation of 5.23 
then affords the desired precursor with methyl methacrylate-like functionality. The 
polymer which results from this monomer is yet to be prepared. 



An unsymmetrical acid-functional diene, 5.26 , was constructed by this method 
and then deoxygenated to the corresponding methyl-substituted monomer, 5.27 , for 
comparison with its symmetrical counterpart prepared previously and to be published 
elsewhere.'^ Ic 



Propylene-like functionality (unsymmetrical) 




1) 2 LDA, -15°C 

* 

2) > Br 
' '9 




1) LAH 

2) TsCfpyr 
1 

3) LAH 




'7 ' '9 
5.27 



Figure 5-13. Synthesis of unsymmetrical methyl substituted monomer. 



Preparation and Characterization of ADMET 
Ethylene/Polar Monomer Copolymers 

Each of the model polymers was prepared by the one-pot homogeneous 
ADMET/heterogeneous hydrogenation process described in Chapter 4. The only 
exception, a model ethylene/vinyl chloride copolymer, required the use of a different 
hydrogenation system to be discussed later. Monomer and catalyst (400: 1 ) were 
combined inside an argon-purged dry box and the reaction was then conducted at 45- 
65°C (48hrs) under reduced pressure utilizing schlenk line techniques. The resulting 
unsaturated polymers were combined with silica (100 times the weight of ruthenium 
catalyst) and olefin hydrogenation was conducted in toluene at 90°C under 120 psig H 2 
for 24 hours to insure maximum reduction. The catalyst residue/silica composite was 
filtered from solution and the solvent evaporated under reduced pressure to yield the 
colorless, viscous or white solid polymers in virtually quantitative yield. The samples 
were not reprecipitated except where noted. Molecular weights as measured by GPC 



Ill 



versus polystyrene standards consistently fell within the range 2 - 5 x 10 4 g/mol with 
PDI's - 2. 



ADMET Ethylene/Vinyl Acetate Model Copolymers 

Ethylene/vinyl acetate (EVA) copolymers constitute the world’s largest volume 
ethylene copolymer, due in no small part to the technical significance of the post- 

-IT 

polymerization hydrolysis to ethylene vinyl alcohol copolymers.' They are prepared by 
the traditional free-radical techniques employed for low-density polyethylene. This is a 
special case of an ideal copolymerization, termed Bernoullian, where the reactivity ratio 
product rir 2 = 1.0 and ri = r 2 = 1.0. Monomer incorporation is truly random and the 
copolymer constitution is the same as the feed ratio, allowing the preparation of EVA’s 
with a broad range of compositions. The ready availability of EVA's with varying VA 
content has made this copolymer system the most ideal available one for studying 
copolymer composition/property and morphology relationships and as a result much 
physical data has accumulated for these copolymers. 97 The abundance of physical data 
concerning random EVA's should prove most useful for later comparison with the model 
periodic ADMET EVA's reported here. 

A series of ADMET EVA's was prepared with pendant acetate groups separated 
by 18, 20, 22, and 26 backbone carbons. The latter 3 polymers yield tough films and 
fibers from solution and from the melt and optically transparent or translucent articles 



may be obtained from the melt. 



1 12 











Figure 5-14. 'H NMR spectra for P5.12 and HP5.12 , 



1.0 



113 




220 200 180 160 140 120 100 80 60 40 20 0 ppm 




Figure 5-15. I3 C NMR spectra for P5.12 and HP5.12 . 



The ‘H and l3 C NMR spectra for P5.12 and HP5.12 are shown in Figures 5-14 
and 5-15. Examination of the olefin regions indicates that the method for hydrogenation 
is exhaustive as indicated by complete disappearance of signals at -5.3-5. 4 ppm ('H) and 
at -130 ppm ( I3 C). The diminutive 'H signals at 4.9 and 5.8 ppm (terminal vinyl groups 
of unsaturated polymer) and at 0.8 ppm (corresponding methyl end groups of saturated 
polymer) indicate high molecular weight. Hydrogenation greatly simpliflies the 13 C 
spectrum as a result of the degeneration of chemical environments distal to the acetate 
group. All spectra for the remaining EVA models are identical, regardless of ethylene 
run length, aside from changes in the ‘H integrals corresponding to decreasing acetate 
content. 

The pattern of peaks in the 13 C spectra of the model EVA is characteristic of all 
the model polymers regardless of the pendant functionality. In addition to peaks 
attributed to the pendant group and the methine carbon, the only other peaks are a 
partially resolved cluster of peaks at - 30 ppm ("f" in Figure 5-15) and signals slightly 
downfield and upfield relative to "f" for the carbons a and (3, respectively, to the methine 
carbon. The cluster of peaks "f" may be partially resolved to 3-5 peaks and corresponds 
to backbone carbons (polyethylene segments) more than twice removed from the methine 
carbon. The GPC traces of P5.12 and HP5.12 are overlaid in Figure 5-16 to illustrate 
constancy in retention time and polydispersity. The unchanged PDI suggests the absence 
of any reactions such as chain scission or other means of molecular weight reduction 
during hydrogenation. The fact that the retention time remains the same is not conclusive 
evidence that DP is unaffected, only that hydrodynamic volume is similar. It might 
however be expected that the conversion of isolated double bonds, separated by long 



115 



flexible segments to single bonds would not drastically affect hydrodynamic properties. 
Should this assumption hold, then it can also be assumed that identical M p from GPC 
truly indicates unchanged DP. Also of note, PDI for all of these polymers will remain ~2 
as opposed to much broader and variable PDI's (3 - 7) for commercial EVA’s. 




Figure 5-16. GPC traces for P5.12 and HP5.12 . M n = 4 x 10 4 g/mol, PDI = 1.8. 



Thermal Analysis of EVA Model Copolymers 

The physical properties of EVA copolymers can be tailored by varying the VA 
content to fit applications as diverse as hot melt adhesives, packaging films, and rubbery 
toys. Melting point depression upon inclusion of a noncrystallizable repeat unit into a 
crystallizable backbone is a general phenomenon and EVA copolymers are no exception, 
where the degree of depression is proportional to the comonomer content. 



1 16 



This phenomenon was treated theoretically by Flory 98 and described by the well- 
known Equation 5-2. The equation describes the observed melting point ( T m ) as a 

function of the mole fraction of noncrystallizable groups (Xb) where T° is the theoretical 

maximum melting point of the crystalline homopolymer and A H° is the heat of fusion 

per crystallizable repeat unit. While this equation adequately describes the generally 
observed trends in melting point depression, it underestimates the magnitude of 
depression due to oversimplifying assumptions." Among other things, the theory 
assumes crystals of infinite dimensions, ignores the chemical identity of the non- 
crystallizable unit, and assumes crystallization of very short segments. Subsequent 
revisions of the theory provide closer fit to experimental data but these versions still fall 
short. 



y<o 



EL y 
A Hi h 



Eqn 5-2 



Not surprisingly, the same trend in T m (DSC) with increasing VA content was 
observed with these model EVA polymers. As expected, each of the polymers displayed 
narrow melting transitions relative to commercial EVA’s due to greater chemical 
homogeneity. Typical DSC heating traces, with varying thermal pretreatment are shown 
in Figure 5-17 along with a trace reported elsewhere for a commercial EVA (28% VA 
wt/wt, not to scale); overlaid to demonstrate the stark contrast in melting behavior. The 
melting transition for this polymer spans 50°C, typical of semicrystalline random 
copolymers indicating a broad range of crystalline morphologies ascribed to chemical 



inhomogeneity. 



117 



/ Commercial EVA (28 wt % VA) \ 




Figure 5-17. Annealing studies for HP5.13 (24.4 wt % VA). Annealing conditions are 
indicated above each trace. 



The samples were heated to 100°C to erase thermal history and then scanned at 
5°C/min, cooling cycle first, to locate thermal transitions. On the cooling cycle, all 
polymers displayed a first order, sharp exotherm, generally spanning <15°C. On the 
heating cycle, a broader endotherm with a shoulder on the high temperature side was 
observed (bottom curve, Figure 5-17). The difference between the peak temperatures for 



118 



the two major transitions, assigned as T c and T m respectively, was < 20°C. The onset of 
the T m is unclear due to observed premelting indicated by the slope of the baseline 
leading up to the peak, which may induce error in the measured values for AH m . 

Annealing was found to affect the positions and shapes of the melting endotherms 
in typical fashion 101 (Figure 5-17). Samples were isothermally crystallized from the melt 
at various temperatures (T a ) below T m , cooled to 15°C and then scanned at 5°C/min. 
Some observations regarding the effect of annealing are: 

• Peak T m and AH,„ increase with increasing T a and the peak sharpens. 

• Baseline preceding T m changes (dotted lines included to aid visualization). 

• For T a = 34°C, AFI„,decreases and broad endotherm below T a appears. 

While interpretation of thermal analyses such as these in relation to polymer 
morphology without corroboration with other techniques (e.g., X-ray, microscopy, etc.) 
can be perilous, some general statements are in order. Annealing between T m and T c 
provides energy to allow amorphous polymer to organize into crystals, allows crystals to 
organize into larger crystals, and allows defective crystals to become more perfect. The 
outcome is generally an increase in T m and/or AH m as seen in the results above. The 
slope of the baseline preceding T m for the unannealed sample may be the result of 
melting (possibly with simultaneous recrystallization) of a population of smaller or less 
perfect crystals formed during the thermal scan. In the annealed samples, a greater 
percentage of this population has most likely been more highly organized and melts at the 
higher temperature. Annealing at T a =34°C shifts the T m to a still higher value but 
provides enough thermal energy to maintain more of the polymer in the liquid state. 
Poorly formed crystals can then arise during the cooling prior to scanning and these 



crystals may give rise to the broad exotherm preceding the higher temperature, narrow 
peak. 




Figure 5-18. DSC trace of HP5.15 with endotherm prior to T m on heating indicating 
cold-crystallization, a: Annealed from the melt (49°C, 5 hrs) 

The DSC trace for HP5.15 reveals thermal behavior not observed with any other 
polymer reported in this dissertation (Figure 5-18). The endothermic peak on the heating 
cycle just prior to the T m indicates cold crystallization. This phenomenon, first 
theoretically treated by Wunderlich. 102 is most effectively observed after a crystalline 
polymer is quenched from above T m to below T g , minimizing crystallization. On heating 
to a given temperature between T m and T g , crystallization ensues. The rate of 
crystallization for HP5.15 may be significantly slower than for the other EVA's, 
diminishing the effective crystallization which occurs during the cooling scan. On the 



120 



subsequent heating cycle, more of the amorphous material then crystallizes rapidly just 
below T ra . Annealing just below T m as before for the other polymers all but eliminates 
the sharp endotherm in the subsequent heating scan. 

The melting points (unannealed) of the polymers are plotted versus weight % VA 
together with the values reported 103 for the commercially available Elvax® series (Figure 
5-19). Like the commercial resins, the relationship is linear (R 2 = 0.99) in the region 
studied. The values are consistently lower for the model polymers and the absolute value 
for the slope is approximately 3 times that of the commercial resins. It is interesting to 
note the relationship between the values of the y-intercepts, corresponding to 0%VA. 
The intercept for the commercial resins is 1 14°C which is very close to the reported T m = 
110°C for the parent low density polyethylene prepared under the same conditions. 
Although it may be purely coincidental, the intercept for the model EVA’s (145.6°C) is 
almost exactly the same as the theoretical value for perfectly linear 
polyethylene( 145. 5°C). 




Figure 5-19. Plot of T m vs. weight % VA. ▲ ADMET EVA copolymers (5°C/min). 
OElvax® series commercial EVA (10°C/min) 103 . 



121 



Upon heating to approximately 300°C, acetic acid is thermally eliminated from 
the polymer backbone and this is the source of the upper limit for processing and end use 
temperatures of commercial EVA's. Measurement of weight loss due to this reaction by 
TGA has been shown to be a useful technique for measuring the acetate content of EVA 
copolymers. 104 A representative TGA plot (HP5.12 ) is shown in Figure 5-20 together 
with the tabulated predicted and measured % weight loss due to thermal elimination of 
acetic acid for each polymer. The transition between the weight loss due to thermal 
elimination of AcOH and the onset of further decomposition was not seen as a distinct 
plateau nor did it occur at the exactly same temperature for each run. The end of the first 
step was taken as the inflection point “A” (Figure 5-20) digitally calculated for each plot. 
Average values for each of the model EVA's are tabulated showing good correlation 
between the measured acetate content and that calculated from the known repeat unit. 



100 



80 



60 



JZ 

CT) 

CD 



40 



20 



AcO 



Poly> 




.Poly 






Poly^ 




- AcOH 
^Poly 1 


AY 

r 


\ 

\ 

\ 

\ 



Polymer 


Wt % AcOH 


AY (%) 


HP5.12 


18.50 


18.7 +/- 0.3 


HP5.13 


17.03 


17.1 +/- 0.5 


HP5.14 


15.78 


15.7 +/- 1.0 


HP5.15 


13.75 


14.1 +/- 0.3 


1 

100 


200 


1 

300 






Point "A" 



400 



Temperature (°C) 



500 



Figure 5-20. TGA trace for HP5.12 showing typical two-stage weight loss. Values 
reported for AY are average of 3 runs. 



122 



ADMET Ethylene/Alkyl Acrylate Copolymers 

Commercial ethylene/methyl acrylate (EMA) copolymers generally contain 
approximately 20 weight % methyl acrylate. 37 The monomer reactivity ratios for methyl 
acrylate and ethylene differ substantially (ri=ll, r 2 = 0.2, Mi = methyl acrylate) 
compared to the EVA system(ri=l .02, r 2 =0.97, M, = vinyl acetate) at 60°C. A plot of 
copolymer composition (Fi) vs feed composition (fj) using these values for both systems 
is shown in Figure 5-21. The high initial slope of the line for EMA indicates that very 
small changes in feed ratio have a significant effect on copolymer composition in the 
range of copolymer composition (10-40 weight % comonomer) typical for useful 
properties in EVA. In this range, a change of 1% in f) results in approximately 5% 
change in F(. 

EMA’s exhibit similar physical properties to EVA’s which is not surprising 
considering that the pendant groups in the polar repeat units are merely isomers, differing 
only in the orientation of the ester group. Technical bulletins put forth by Chevron make 
numerous comparisons between their EMAC® series of ethylene/methyl acrylate 
copolymers with commercial EVA copolymers. Reported melting points for the 
EMAC® resins are within 10°C of the analogous EVA (Elvax®, DuPont) resins with 
approximately the same weight percent comonomer. 

The DSC heating curves for HP5.20 are HP5.22 are shown in Figure 5-22 along 
with the analogous ADMET EVA’s for comparison. In agreement with the commercial 
relatives, DSC thermal data for the ADMET EVA’s and EMA’s do not show drastic 
differences in peak melting temperatures. However, distinction is seen in the shape of the 



123 



endotherms where the EMA peaks appear narrower and do not exhibit a high temperature 
shoulder as seen in the latter. 




Figure 5-21. Dependence of copolymer composition (F|) on comonomer feed 
composition (fi) for EMA and EVA. 



Without corroborative evidence, we can only speculate at this point as to the 
cause of this difference. While no attempts were made to measure second order 
transitions here, it has been shown that commercial EVA’s and EMA’s exhibit 
significantly different second order transitions attributed to flexible side-group 
reorientation. Numerous studies, empirical 101 and theoretical , 106 reach the general 
consensus that the cause is greater hindrance of rotation of the pendant ester group of MA 
relative to VA repeat units about the bond by which each is attached to the main chain. 



124 





0 10 20 30 40 50 60 

Temperature ( °C) 



Figure 5-22. DSC heating curves for ADMET EMA’s and EVA’s (2 nd scan 5°C/min) 



Comparison of representative partial 'H NMR spectra (Figure 5-23) of the EVA 
and EMA models in the region of the methine protons reveals the differing chemical 
environments exerted by the two groups on the proximal backbone atoms. The methine 
proton of the EVA gives rise to the expected pentet due to splitting by four equivalent 
geminal protons. However, in the case of the EMA, these four protons are not equivalent 
due to anisotropy induced by the attached carbonyl, supporting theoretical calculations 
indicating greater hindrance to rotation. Perhaps the differing mobilities of these two 
pendant groups affects crystal packing in these periodic polymers giving rise to the 
observed differences in melting behavior. 



125 



O' 



O 



(> l 

'InVf. 



18 



hJ 



vV vy^’ 

5.00 4.90 



(VCX 



(> 



18 



w 



A/ 



^ 

4.80 4.70 PPm 2.50 2.40 



2.30 



2.20 2.10 PPm 



Figure 5-23. Partial 'H NMR spectra of ADMET EVA and EMA showing differing 
multiplicities of methine proton signals. 



A significant difference exists between the thermal stabilities of commercial 
EVA’s and EMA’s, with important technical implications, where the lower thermal 
stability of the former limit their maximum processing temperatures to 232°C, nearly 
100°C lower than for the latter. TGA measurements bear out the expected higher thermal 
stability of the ADMET EMA’s with Tj (10% Wt loss) = 445°C, and with an onset 
approximately 100°C greater than for the EVA’s. 

The larger ethoxycarbonyl group of HP5.21 (ethylene/ethyl acrylate, EEA) causes 
a significant change in the profile of the DSC melting endotherm (Figure 5-24) relative to 
the analogous EMA. Two maxima are seen and the position of the higher temperature 
maxima is within 1°C of the peak seen for the analogous EMA model. Annealing at 
3°C(lhour) shifts the lower temperature maxima by nearly 4°C toward the higher 
temperature peak, creating greater overlap and eliminating visible premelting as 
evidenced by flattening of the slope of the baseline prior to the endotherm. It might be 
assumed that incorporation of the larger ethoxy carbonyl group into some type of ordered 



126 



array is more difficult than the smaller methoxy carbonyl. Although based solely on 
speculation, the observed melting behavior of the EEA and EMA might suggest that 
similar morphologies are formed in both with more of a less perfect phase in the EEA 
giving rise to the lower temperature peak. Annealing just below this lower temperature 
peak shifts it to higher temperature possibly indicating crystal perfection toward that 
giving rise to the higher temperature peak. 





Figure 5-24. DSC heating traces for ADMET EMA( HP5.20 ) and EEA( HP5,21) . 
A: EEA, B: EEA annealed 3°C(1 hr), C: EMA. 



ADMET Ethylene/Styrene Model Copolymer 

Due to the extremely high reactivity of styrene relative to ethylene, copolymers 
with even moderate levels of ethylene incorporation are unattainable by free-radical 



127 



copolymerization. Alternately, traditional Ziegler-Natta copolymerization of ethylene 
and styrene typically produced polymers which contain less than 1% styrene. 107 Recent 
reports disclose the use of homogeneous half-sandwich and metallocene catalysts, 
activated by MAO, to successfully prepare ethylene/styrene copolymers with varied 
levels of styrene content. 108 

In order to compare the effect of a regularly spaced phenyl group on 
crystallization, an ADMET ethylene/styrene copolymer ( HP5.17 ) was prepared. Much 
larger melting point depression is seen relative to the analogous EVA copolymer with 18 
carbons between pendant groups. Figure 5-25 shows the DSC heating trace with two 
broad but apparently separate endotherms with low AH relative to other polymers. 
Annealing for 5 hours between the two peaks resulted in a relatively broad peak at 
approximately -6°C with several maxima following the peak temperature. The obvious 
assumption in this case is that the greater steric bulk of the phenyl ring has a drastic effect 
on the packing behavior of the polymer chains. 




Figure 5-25. DSC traces for ADMET E/STY ( HP5.17) 



128 



ADMET Ethylene/Vinyl Chloride Copolymer 

Chlorinated polyethylene (CPE) may be prepared from polyethylene via various 
chlorination techniques resulting in polymers with are formally ethylene/vinyl chloride 
(EVC) copolymers. 109 These materials have significant technical possibilities and have 
proven convenient subjects for microstructure/morphology studies. 110 In particular, 
morphological studies have been aimed at determining whether the chloride groups are 
included into crystal lattices as defects or excluded in amorphous regions. In an elegant 
study by Wegner, 1 ll)J EVC's were prepared by several chlorination techniques from PE 
and partially hydrogenated polyoctenamers. The different methods produced polymers 
with varying incorporation of isolated and vicinal chloro substituents. X-ray studies in 
concert with DSC measurements indicated that the chloro groups were incorporated as 
defects in the crystals. As the chloride content of the polymers increased, the lateral 
dimensions of the unit cell were shown to increase relative to the unit cell of pure 
polyethylene. Polymers with vicinal chloro groups showed greater melting point 
depression as a result of the larger defects. Alternately, EVC's with a broad range of 
chlorine content have also been prepared by partial reduction of PVC with trialkyl tin 
hydrides and similar results were obtained from X-Ray and DSC measurements. 1 IOc 

The DSC melting points of random EVC's were compared to analogous 
ethylene/propylene (EP) copolymers as a function of counit content and the two data sets 
fell on the same line. ll0b The researchers concluded that the melting behavior was the 
same due to the similar size of the chloro and methyl groups. The smaller van der Waals 
radius 111 of chloride(1.75 A) relative to methyl (2.0 A) may be offset somewhat by the 
longer C-Cl bond length (1.78 A) relative to C-C(1.54 A). 



129 



In order to compare the effect of a periodic chloro group to the other pendant 
groups in this study, an ADMET ethylene/vinyl chloride (EVC) copolymer was prepared. 
The EVC could not be prepared by the one-pot method used for all the other polymers in 
this study. While the ADMET reaction proceeded in high conversion utilizing the 
ruthenium catalyst, the heterogeneous hydrogenation step never proceeded to greater 
than 70% ('H NMR). It is postulated that the chloride acts in some fashion to poison the 
heterogeneous hydrogenating species. However, hydrogenation over Pd/C at 5 atm H 2 in 
toluene proved quantitative within spectroscopic detection limits. The reduction was 
conducted at 40°C as the polymer begins to precipitate from room temperature toluene 
during the course of the reaction as evidenced by the formation of gelatinous particles in 
the mixture. Subsequent observations showed the final product to be poorly soluble in 
toluene, benzene, methylene chloride, and 1,2-dichloroethane but freely soluble in 
chloroform and THF at room temperature. 

Based purely on steric arguments, a model ethylene/vinyl choride copolymer with 
pendant chloride every 19 th carbon would be expected to have a higher melt than any of 
the analogues reported here with the same ethylene run length. This was indeed found to 
be the case as HP5.16 shows a T m = 77°C (AH m = 78 J/g). In stark contrast to the studies 
involving random EVC and EP, llob the analogous ADMET EP polymer with a pendant 
methyl every 19th carbon melts 21°C lower (Tm = 56°C) 51c than this ADMET EVC. 

TGA shows stepwise weight loss due to thermal elimination of HC1 prior to 
catastrophic decomposition. The elimination of HC1 occurs at a lower temperature than 
for loss of AcOH from EVA, consistent with the relative stabilities of secondary alkyl 



130 



chorides and acetates. The measured weight loss due to elimination of HC1 (1 1.54 %) is 
in reasonable agreement with that predicted from the known structure (12. 1 1 %). 




°L 



100 200 300 400 500 600 

Temperature ( °C ) 

Figure 5-26. TGA trace of HP5.16 , Onset for catastrophe decomposition = 445°C. 

Conclusions 

The method of homogeneous ADMET/heterogeneous hydrogenation has been 
demonstrated as an effective and facile alternative to two step procedures for the 
generation of models for ethylene polar monomer copolymers. The necessary dienes are 
easily prepared by typical organic transforms whereby the type of substituent in the 
polymers and their frequency in the backbone is regulated during the monomer synthesis. 
Access to periodic polymers with much longer ethylene run lengths than previously 



131 



reported was made possible through the synthesis of homologated co-alkenyl bromides 
and carboxylic acids used in the diene synthesis. 

Comparison of polymer melting behavior of a series of polymers with identical 
ethylene run lengths but differing pendant groups R, found correlation between the 
degree of melting point depression and the steric bulk of R with the highest degree of 
depression for R = phenyl and the lowest for R = Cl. The trend is graphically depicted in 
Figure 5-27. In addition to precisely defined run lengths, all the model ethylene polar 
monomer copolymers regardless of comonomer, have approximately the same molecular 
weights and PDI's, decreasing the number of variables which must be considered for 
adequate analysis of physical data. Enormous potential for obtaining deeper insight into 
structure property relationships of commercial ethylene/polar mononomer copolymers via 
more applied property studies of these models is at hand. 




Figure 5-27. Peak melting temperatures of ADMET ethylene/polar monomer 
copolymers as a function of chemical identity of the regular pendant 
group. Value for R = Ph is the average of two observed peaks. 



REFERENCES 



E Kobayashi, S. Catalysis in Precision Polymerization , John Wiley and Sons: West 
Sussex, 1997. 

2. Ivin, K.J.; Mol, J.C. Olefin Metathesis and Metathesis Polymerization, Academic 
Press: San Diego, 1997. 

3. Grubbs, R.H., Chang, S. Tetrahedron 1998, 54, 4414. 

4. Grubbs, R.H; Marsella, M.J.; Maynard, H.D. Angew. Chem. Int. Ed., Eng. 1997, 
36, 1101. 

5. Wessermel, K.; Arpe, H.-J. Industrial Organic Chemistry, 2 nd Ed, VCH 
Publishers, Inc: New York, pp 85-88. 

6. Herrison, J.L.; Chauvin, Y. Makromol. Chem. 1970 , 141. 

7. Streck, R. J. Mol. Cat. 1998 , 46, 305. 

8. Pariya, C.; Jayaprakash, K.N.; Sarkar, A. Coord. Chem. Rev. 1998 , 168, 1-48. 

9. a) Calderon, N.; Ofstead, E.A.; Judy, W.A. J. Poly. Sci A-2. 1967 , 5, 2209. b) 
Calderon, N.; Chen, H.Y.; Scott, K.W. Tetr. Lett. 1967 , 3327. c) Hughes, W.B. 
Organomet. Chem. Synth. 1972 , 7, 341. c) Truett, W.L.; Johnson, D.R. J. Am. 
Chem Soc. 1960 , 82, 2337. 

10. a) Cho, I.; Bae, S.S. Polym. J.. 1991 , 23, 1283. b) Cho, I.; Jeong, S.W. Kor. 
Polym. J. 1993 , 7, 1. c) Hockev, H. J. Mol. Cat. 1991 , 65, 95. d) Mol, J.C. J. Mol. 
Cat.. 1982 , 95. e) Feast, W.J.; Winter, J.N. J. Chem. Soc. Chem. Comm.. 1985 , 
202. f) Bencze, L. et al J. Mol. Cat. 1991 , 65, 21 . 

11. a) Moulijin, J.A.; Mol, J.C. J. Mol. Cat. 1988, 46, 1. b) Mol J.C. Chemtech 1983, 
13, 258. 

12. Fischer, E.O; Maasbol, A. Angew. Chem. Int. Ed., Eng. 1964 , 3, 580. 

13. Casey, C.P.; Burkhardt, T.J. J. Am. Chem Soc. 1974 , 96, 7808. 

14. Katz, T.J.; Lee, S.J.; Acton, N. Tetr. Lett. 1976, 4247. 



132 



133 



15. Schrock, R.R. Acc. Chem. Res. 1990 , 23, 158. 

16. Fu, G.C.; Grubbs, R.H. J. Am. Chem Soc. 1993 , 115, 3800. 

17. Komiya, Z.; Schrock R.R. Macromolecules 1993 , 26, 1393. 

18. Deonceau, A.; Noels, A.F.; Saive, E. J. Mol. Cat. 1992 , 76, 123. 

19. Noels, A.F. Demonceau, A. in Proceedins of the NATO Advanced Study Institue 
on Metathesis Polymerization of Olefins and Polymerization of Alkynes, 
Imamoglu, Y. Ed., Kluwer Academic Publishers: Netherlands, 1998. 

20. France, M.B.; Paciello, R.A.; Grubbs, R.H. Macromolecules 1993 , 26, 4739. 

21. Doyle, M.P. in Catalysis of Organic Reactions, Augustine, R.L. Ed., Marcel 
Dekker: New York, 1985. 

22. Grubbs, R.H.; Miller, S.J. Acc. Chem. Res. 1995, 28, 446. 

23. Grubbs, R.H.; Nguyen, S.T. J. Am. Chem. Soc. 1993 , 115, 9858. 

24. Schwab, P.E.; France, M.B.; Ziller, J.W.; Grubbs, R.H. Angew. Chem. Int. Ed. 
Engl. 1995 , 34, 2039. 

25. Fraser, C.; Grubbs, R.H. Macromolecules 1995 , 28, 7248. 

26. Grubbs, R.H.; Lynn, D.M.; Kanaoka, S. J. Am. Chem. Soc. 1996 , 118, 784. 

27. Fu, G.C.; Nguyen, S.T.; Grubbs, R.H. J. Am. Chem. Soc. 1993 , 115, 9856. 

28. Hillmyer, M.A., Nguyen, S.T.; Grubbs, R.H. Macromolecules 1997 , 30, 718. 

29. Shon, Y-S., Lee, T.R. Tetr. Lett. 1997 , 38, 1283. 

30. Wolfe, P.S. Investigation of the Boronate and Phosphorane Functionality in 
ADMET Polymerization , Ph.D. Dissertation, University of Florida, 1997. 

31. Hanson, P.R.; Stoianova, D.S. Tetr. Lett. 1998 , 39, 3939. 

32. Nicolaou, K.C.; He, Y.; Vourloumis, D.; Vallberg, H.; Yang, Z. Angew. Chem. 
Int. Ed. 1996 , 35, 2399. 

33. Yang, Z.; He, Y.; Vourloumis, D.; Vallberg, H.; Nicolou, K.C. Angew. Chem. Int. 
Ed. 1997 , 36, 166. 

34. Kim, S.H.; Figueroa, I.; Fuchs, P.L. Tetr. Lett. 1997 , 38, 2601. 



134 



35. Wagener, K.B.; Boncella, J.M.; Nel, J.G. Macromolecules 1991 , 24, 2649. 

36. Wagener, K.B.; Tindall, D.; Pawlow, J.H. in Alkene Metathesis in Organic 
Synthesis, A. Furstner, Ed., Springer-Verlag: New York, 1998. p. 183. 

37. Odian, G.G. Principles of Polymerization, 3 ld Ed., John Wiley and Sons, Inc: New 
York, 1991. 

38. Barry, A.J.; Beck, H.N. in Inorganic Polymers, F.G.A. Stone and W.A.G. 
Graham, Eds, Academic Press: New York, 1962, Chapter 5. 

39. Zeuch, E.A.; Hughes, W.B.; Kubicek, D.H.; Kittleman, E.T. J. Amer. Chem. Soc. 
1970, 92, 528. 

40. Dall’Asta, G.; Stiglian, G.; Greco, A.; Motts, L Chim. Ind. 1973 , 55, 142. 

41 . Wagener, K.B.; Lindmark-Hamberg, M. Macromolecules 1987 , 20, 2949. 

42. Nubel, P.O.; Lutman, C.A.; Yokelson, H.B. Macromolecules 1994. 27, 7000. 

43. Wagener, K.B.; Brezezinska, K.; Wolfe, P.S.; Watson, M.D. Macromol. Chem. 
Phys. 1996 , 197, 2065. 

44. Wagener, K.B.; Brzezinska, K. Macromolecules 1991 , 24, 5273. 

45. Wagener, K.B.; Wolfe, P.S.; Gomez, F.J. Macromolecules 1997 , 30, 714. 

46. Steiger, D.; Ehrenstein, M.; Weder, C.; Smith, P. Macromolecules 1998 , 31, 1254. 

47. Qiao, J.; Baker, G.L. Polym. Prep. (Amer. Chem. Soc., Polym. Chem.) 1997, 38(1), 
633. 

48. Qiao, J.; Baker, G.L. Polym. Prep. (Amer. Chem. Soc., Polym. Chem.) 1997, 38(2), 
239. 

49. Qiao, J.; Baker, G.L. Polym. Mater. Sci. Eng. 1995 , 73, 515. 

50. Valenti, D.J.; Wagener, K.B. Macromolecules 1998 , 31, 2764. 

51. a) Wagener, K.B.; Valenti, D.J. Macromolecules 1997, 30, 6688. b) Valenti, D.J. 
Acyclic Diene Metathesis Polymerization(ADMET) Polymerization: The Synthesis 
of Well-Defined Model Polymers for Polyolefin Materials, Ph.D. Dissertation, 
University of Florida, Gainesville, 1997. c) Smith, J.A. Brzezinska, K.; Valenti, 
D.J.; Wagener, K.B. manuscript in preparation. 



135 



52. a) Ast, W.; Hummel, K. Naturwiss. 1970 57, 545. b)Ast, W.; Hummel, K. 
Makromol. Chem. 1973 , 166, 39. c) Hummel, K. Pure Appl. Chem. 1982 , 54, 351. 
d) Hummel, K.; Steltzer, F.; Heiling, P.; Wedam, O.A.; Griesser, H. J. Mol. Cat. 
1980 , 8, 253. e) Hummel, K.; Heiling, P.; Honig, H. Monatshefte Chem. 1980 , 
111, 575. f) Hummel, K.; Hubmann, E.; Dosinger, M.; Pongratz, T. Eur. Polm. J. 
1991 , 27, 303. g) Hummel, K.; Hubmann, E. Makromol. Chem., Macromol. Symp. 

1990 , 40, 139. h) Hummel, K.; Lechner, H.; Zekoll, H.; Chemilli, R. Makromol. 
Chem. 1987 , 188, 2841. i) Hummel, K.; Kiattanavith, N.; Marti, K.; Marti, C. J. 
Mol. Cat. 1992 , 76, 239. j) Steltzer, F.; Hobisch, G.; Pongratz, T.; Hummel, K. J. 
Mol. Cat. 1988 , 46, 433. k) Hummel, K.; Paschen, A. Polym Bull. 1990 , 24, 391. 
1) Hummel, K.; Hubmann, E.; Pongratz, T. Eur. Polym. J. 1988 . 24, 141. m) 
Steltzer, F.; Hummel, K.; Graimann, C.; Hobisch, J.; Marti, M.G. Makromol. 
Chem. 1987 , 188, 1795. n) Calderon, N.; Ofstead, E.A. Makromol. Chem. 1972 , 
154, 21. o) Wagener, K.B.; Marmo, J.C. Macromolecules 1993 , 26, 2137. p) 
Wagener, K.B.; Marmo, J.C Macromolecules 1995 , 28, 2602. q.) Marmo, J.C. 
"ADMET Depolymerization: The Synthethis of Perfectly Difunctional Telechelics 
", Ph.D. Dissertation, University of Florida, 1994. r) Chung, T.C.; Chasmawala, 
M. Macromolecules 1992 , 25, 5137. 

53. a)Crowe, W.E.; Zhang, Z.J. J. Am. Chem. Soc. 1993 , 115, 10998. b)Crowe, W.E.; 
Goldberg, D.R. J. Am. Chem. Soc. 1995 , 117, 5162. c)Crowe, W.E.; Goldberg, 
D.R.; Zhang, Z.J. Tet. Lett. 1996 , 37, 2117. 

54. Bradshaw, C. P. C.; Howman, E.J.; Turner, L. J Catalysis. 1967 , 7, 269. 

55. a) Mol, J. C. J Molecular Catalysis 1991 , 65, 145. b) Bosma, R. H. A.; Van den 
Aardweg, F.; Mol, J.C. J. Chem. Soc. Chem. Comm 1981 , 1 132. c) Zuech, E.A.; 
Hughes, W.B.; Kubicek, D.K.; Kittleman, E.T. J. Am. Chem. Soc. 1970 , 92, 528. 
d)Boelhouwer, C.; Mol, J.C. J. Am. Oil chem. Soc. 1984 , 61, 425. e)Banks, R.L.; 
Banasiak, D.S.; Hudson, P.S.; Norell, J.R. J. Mol. Catal. 1982 , 75, 21. f) K.J. Ivin 
and J.C. Mol, in Olefin Metathesis and Metathesis Polymerization , Academic 
Press, San Diego, 1997, chap 7,9, and 16. 

56. Wagener, K.B.; Puts, R.D.; Smith, D.W. Makromol. Chem., Rapid Commun. 

1991 , 72,419. 

57. Schwab, P.; Grubbs, R.H.; Ziller, J.W. J . Am. Chem. Soc. 1996 , 118, 100. 

58. Abendroth, H.; Canji, E. Makromol. Chem. 1975 , 176, 775. 

59. Thorn-Csanyi, E.; Perner, H. Makromol. Chem. 1979 , 180, 919. 

60. Wagener, K.B.; Marmo, J.C. Macromol. Rap. Commun. 1995, 67, 557. 



136 



61. Hillmyer, M. A. The Preparation of Functionalized. Polymers by Ring-Opening 
Metathesis Polymerization, Ph.D. Dissertation, California Institute of Technology, 
1995. 

62. Grubbs, R.H. CHEMTECH 1977 , 512. 

63. Nguyen, S.T.; Grubbs, R.H. J. Organomet. Chem. 1995, 497, 195. 

64. Murrell, L.L..; Advanced Materials in Catalysis, Academic Publishers: New 
York, 1977. 

65. Collman, J.P.; Hegedus, L.S.; Norton, J.R.; Finke, R.G. Principles and 
Applications of Organotransition Metal Chemistry, University Science Books: 
Mill Valley, 1987, p. 560. 

66. Barlow, M.G. J. Organomet. Chem. 1970, 21 , 215. 

67. McLain, S.J.; Sancho, J.; Schrock, R.R. J. Am. Chem. Soc. 1979, 101 , 5451. 

68. Schrock, R.; McLain, S.; Sancho, J. Pure and Appl. Chem. 1980, 52, 729. 

69. Datta, S.; Fischer, M.B.; Wreford, S.S. J. Organomet. Chem 1980, 188, 353. 

70. Wilkinson, G. Nature 1965, 208, 1203. 

71. Wilkinson, G. Chem. Comm. 1967, 7, 305. 

72. McLain, S. Proceedings PMSE 1997, 76, 246. 

73. Garrou, P.E. Chem. Rev. 1985, 85, 171. 

74. Knifton, J.F. J. Org. Chem. 1976, 41, 1200. 

75. Pennelle, J.; et al Proceedings PMSE 1997, 76, 158. 

76. Staudinger, H. From Organic Chemistry to Macromolecules, John Wiley and 
Sons, Inc: New York, 1970, p 93. 

77. Armstrong, F.B. Biochemistry 3 rd Ed, Oxford University Press: New York, 1989. 

78. Porri, L.; Giarrusio, A. in Comprehensive Polymer Science Vol 4, Pt 2, G. Allen 
Ed., Pergamon Press: New York, 1989. Ch. 5. 

79. Stanley, H.M. in Polyethylene and its Industrial Derivatives, S.A. Miller, Ed., 
Ernstbenn: London, 1969. Ch. 1. 



137 



80. Levowitz, E.L. in Modern Plastics Encyclopedia ’99, McGraw-Hill: New York, 
1999. pB-3. 

81. James, D.E. in Encyclopedia of Polymer Science and Engineering 2 nd Ed., Vol 6. 
Wiley-Interscience: New York, 1986. 

82. Brandrup, J.; Immergut, E.H. Eds. Polymer Handbook. Wiley-Interscience: New 
York, 1989. 

83. Gnanou, Y. J.M.S. Rev. Macromol. Chem. Phys. 1996, C36, 77. 

84. Bahr, U.; Deppe, A.; Karas, M.; Hillenkamp, F.; Giessmann. U. Anal. Chem. 

1992, 64, 2866. 

85. Moore, J.S. in Synthesis of Polymers, A.D. Schliiter, Ed. Wiley-VCH: Weinheim, 
1999. Ch 1. 

86. Jannsen, H.M.; Meijer, E.W. in Synthesis of Polymers, A.D. Schliiter, Ed. Wiley- 
VCH: Weinheim, 1999. Ch 12. 

87. Brooke, G.M.; Burnett, S.; Mohammed, S.; Proctor, D.; Whiting, M.E. J. Chem. 
Soc. Perkin Trans. 1996, 1 , 1635. 

88. Wagener, K.B.; Konzelman, J. Macromolecules 1996, 29, 7657. 

89. O’Gara, J.E.; Wagener, K.W.; Hahn, S.F. Makromol. Chem., Rapid Commun. 

1993, 14, 657. 

90. Wu, Z.; Grubbs, R.H. Macromolecules 1994, 27, 6700. 

91. Arnold, M.; Wegner, T. Macromol. Chem. Phys. 1995, 196, 3797. 

92. Ho, P-T.; Davies, N. J. Org. Chem. 1984, 49, 3027. 

93. Hall, S.S.; McEnroe, F.J. J. Org. Chem. 1975, 40, 271. 

94. Belletire, J.L.; Fry, D.F. J. Org. Chem. 1987, 52, 2549. 

95. Pfeffer, P.; Silbert, L.S.; Chirinko, J.M. J. Org. Chem. 1972, 37, 45 1 . 

96. Moore, G.G.; Foglia, T.A.; McGraham, T.J. J. Org. Chem. 1979, 44, 2425. 

97. Zutty, N.L.; Faucher, J.A.; Bonotto, S. in Encyclopedia of Polymer Science and 
Technology H.F. Mark, Ed. John-Wiley and Sons: New York, 1967. p 387-43. 

98. Flory, P.J. Trans. Farad. Soc. 1955, 390(51), 848. 



138 



99. Wunderlich, B. Macromolecular Physics Vol 3. Academic Press: New York, 
1980. 

100. Bistac, S.; Kunemann, P.; Schultz, J. Polymer 1998, 39, 4875. 

101. Wunderlich, B. Macromolecular Physics Vol 2. Academic Press: New York, 
1980. 

102. Wunderlich, B. J. Chem. Phys. 1958, 29, 1395. 

103. Thermal Properties of Elvax® Measured by Differential Scanning 
Calorimetry(DSC). http://www.dupont.com/industrial-polymers/elvax/H-49653- 
l/H-49653-l.html. Accessed 2/15/98. 

104. Marin, M.L.; Jimenez, A.; Lopez, J. J. Thermal Anal. 1996, 47, 247 . 

105. a) Smith, G.D.; Boyd, R.H. Macromolecules 1991, 24, 2725. b) Smith, G.D.; 
Boyd, R.H. Macromolecules 1991, 24, 2731. c) Smith, G.D.; Boyd, R.H. 
Macromolecules 1992, 25, 1326. 

106. a) Boyd, R.H.; Buerger, D.E. Macromolecules 1989, 22, 2694. b) Boyd, R.H.; 
Buerger, D.E. Macromolecules 1989, 22, 2699. c) Boyd, R.H.; Buerger, D.E.; 
Liu, F.; Devereaux, R.W. Macromolecules 1992, 25, 703. 

107. Mani, R.; Burns, C.M. Macromolecules 1991, 24, 5476. 

108. a) D’Aniello, C; Decandia, F.; Oliva, L.; Vittoria, V. J. Appl. Poly. Sci. 1995, 58, 
1701. b) Sernetz, F.G.; Mulhaupt, R. Macromolecules, 1997, 30, 1562. c) 
Venditto, V.; DeTullio, G.; Izzo, L.; Oliva, L. Macromolecules 1998, 31, 4027. 

109. Young, W.L.; Blanchard, R.R. in ref. 81. p 495+. 

1 10. a)Wegner, G.; Gutzler, F. Coll. & Poly. Sci. 1980, 258, 776. b) Bowmer, T.N.; 
Tonelli, A.E. Polymer 1985, 26, 1 195. c) Gomez, M.A.; Tonelli, A.E.; Lovinger, 
A.J.; Schilling, F.C.; Cozine, M.H.; Davis, D.D. Macromolecules 1989, 22, 4441. 
d) Pourahmady, N.; Bak, P.I. J. Macromol. Sci., Pure Appl. Chem. 1992, A29(ll), 
959. 



Bondi, A. J. Phys, Chem. 1964, 68, 441 . 



BIOGRAPHICAL SKETCH 



Mark Deas Watson was born on February 23, 1969 in Pascagoula, Mississippi. 
Shortly thereafter, he moved with his parents and five older siblings as they reestablished 
residence in their hometown, Hattiesburg, MS. Upon graduation from Oak Grove High 
School in 1987, he entered the University of Southern Mississippi enrolling in the 
Department of Polymer Science. He conducted undergraduate research in various facets 
of solution phase properties of surfactants and polymeric viscosity modifiers under the 
guidance of Prof. Robert Y. Lochhead. He also participated in a cooperative education 
program with Schering-Plough in Memphis, Tennessee where he spent three alternating 
semesters in the Dr. Scholl’s Materials and Devices R & D. After receiving his B.S. 
degree in August 1992, he remained at USM for one year conducting research in the 
Shelby F. Thames group in the area of formulation and physical testing of adhesives and 
coatings. He began his graduate studies in organic/polymer chemistry at the University 
of Florida in August 1993 under the guidance of Professor Kenneth B. Wagener. During 
a brief visit home in March 1996, he met his wife-to-be, Kim, whom he married in 
August of the following year. Mark received his Ph.D. in May 1999, and began post- 
doctoral research at the Max-Planck-Institut fur Polymerforschung in Mainz, Germany in 
the research group of Prof. Dr. Klaus Mullen. 



139 



I certify that I have read this study and that in my opinion it conforms to acceptable 
standards of scholarly presentation and is fully adequate, in scope and quality, as a 
dissertation for the degree of Doctor of Philosophy. 

1 /^C 0 IS)- ) — 

Kenneth B. Wagener, Chairman ' 

Professor of Chemistry 

I certify that I have read this study and that in my opinion it conforms to acceptable 
standards of scholarly presentation and is fully adequate, in scope and quality, as a 




I certify that I have read this study and that in my opinion it conforms to acceptable 
standards of scholarly presentation and is fully adequate, in scope and quality, as a 
dissertation for the degree of Doctor of Philosor 




Randolph S. Dut&n 
Associate Professor of Chemistry 

I certify that I have read this study and that in my opinion it conforms to acceptable 
standards of scholarly presentation and is fully adequate, in scope and quality, as a 
dissertation for the degree of Doctor of Philosophy. 



C 



Kirk S. Schanze 
Professor of Chemistry 

I certify that I have read this study and that in my opinion it conforms to acceptable 
standards of scholarly presentation and is fully adequate, in scope and quality, as a 
dissertation for the degree of Doctor of Philosopf 




Anthony B. btenrtan 
Associate Professor of Materials 
Science and Engineering 

This dissertation was submitted to the Graduate Faculty of the Department of 
Chemistry in the College of Liberal Arts and Sciences and to the Graduate School and 
was accepted as partial fulfillment of the requirements for the degree of Doctor of 
Philosophy. 



May, 1999 



Dean, Graduate School