<<

“One-Pot” Oligomeric A 2 + B 3 Approach to Branched Poly(arylene ether sulfone)s: Reactivity Ratio Controlled Polycondensation

A thesis submitted in partial fulfillment of the requirements for the degree of

Master of Science

By

ANDREA M. ELSEN

B.S., Wright State University, 2007

2009

Wright State University

WRIGHT STATE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

June 19, 200 9

I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY Andrea M. Elsen ENTITLED “One-Pot” Oligomeric A2 + B3 Approach to Branched Poly(arylene ether sulfone)s: Reactivity Ratio Controlled Polycondenstation BE ACCEPTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Master of Science .

______Eric Fossum, Ph.D. Thesis Director

______Kenneth Turnbull, Ph.D. Department Chair

Committee on Final Examination

______Eric Fossum, Ph.D.

______Kenneth Turnbull, Ph.D.

______William A. Feld, Ph.D.

______Joseph F. Thomas, Jr., Ph.D. Dean, School of Graduate Studies

Abstract

Elsen, Andrea M. M.S., Department of , Wright State University, 2009. “One-Pot” Oligomeric A 2 + B 3 Approach to Branched Poly(arylene ether sulfone)s: Reactivity Ratio Controlled Polycondensation

The synthesis of fully soluble branched poly(arylene ether)s via an oligomeric A 2 + B 3 system, in which the A 2 are generated in situ, is presented. This approach takes

advantage of the significantly higher reactivity toward nucleophilic aromatic substitution

reactions, NAS, of B 2, 4-Fluorophenyl sulfone, relative to B 3, tris (4-Fluorophenyl) phosphine oxide. The A 2 oligomers were synthesized by reaction of Bisphenol-A and B 2, in the presence of the B 3 unit, at temperatures between 100 and 160 °C, followed by an increase in the reaction temperature to 180 °C at which point the branching unit was incorporated. The presence of branching was confirmed via 31 P NMR spectroscopy and the thermal properties of the were evaluated utitilizing TGA and DSC analyses.

iii

Table of Contents

1. Introduction…………………………………………………………………………….1

Linear polymers………………………………………………………...…2

Dendrimers………………………………………………………………...5

Branched poymers………………………………………………………...9

Poly(arylene ether sulfone)s……………………………………………..14

Current project…………………………………………………………...17

2. Experimental…………………………………………………………………………..19

Synthesis of 4,4-bis -(4-methylphenoxy)diphenyl sulfone, 1b ……………20

Synthesis of mono-substituted tris -(4-Fluorophenyl)phosphine

oxide, 3a ………...………………………………………..20

Representative procedure for model reactions…………………………...20

Representative procedure………………………………..21

General procedure for reverse precipitations……………………….…....21

Preparation of GC/MS calibration curve………………………………...22

Polymerization of A 2 + B 2 +B 2’ , linear analog………………22

Synthesis of A 2 capped oligomers, 2a…………………………………...22

Synthesis of branched poly(arylene ether sulfone)s, 6g …………………23

3. Results and Discussion………………………………………………………………..24

Nucleophilic aromatic substitution………………………………………24

Reactivity determination…………………………………………………24

GC/MS calibration curve………………………………………………...26

Simulation of in situ oligomeric A 2 formation…………………………..30

iv

Synthesis of branched poly(arylene ether sulfone)s……………………..33

Incorporation of B 3 ……………………………………………38

Estimation of degree of branching, DB …………………………………..39

Thermal Analysis………………………………………………………...44

Conclusions………………………………………………………………47

4. References……………………………………………………………………………..48

v

List of Figures

I. Introduction:

Figure 1 . Chain Entanglements……………………………………………………...…....1

Figure 2 . Intrinsic or inherent viscosity versus M n for a generic linear ……...…2

Figure 3 . Young’s Modulus versus temperature plot for a generic linear polymer…...... 3

Figure 4 . Poly(amido )……………………………..………………………….…...6

Figure 5 . Generic scheme of a branched polymer…………...…………………………...9

Figure 6 . Comparison of mechanical properties versus degree of branching for linear,

dendritic and branched polymers………………………………………………………...10

Figure 7a . Comparison of linear, dendritic and branched polymer viscosity versus degree

of branching……………………………………………………………………………...11

Figure 7b . Comparison of linear, dendritic and branched polymer viscosity versus

Mw………………………………………………………………………………………..11

II. “One-Pot” Oligomeric A 2 + B 3 Polymerization of Poly(arylene ether sulfone)s:

Reactivity Ratio Controlled Polycondensation

Figure 8 . 4-Fluorophenyl sulfone, tris-(4-Fluorophenyl)phosphine oxide and bis-(4-

Fluorophenyl)phenyl phosphine oxide…………………………………………………..25

1 Figure 9 . The 300 MHz H NMR spectrum (CDCl 3) of 1b ……………………………..27

13 Figure 10 . The 75 MHz C NMR (CDCl 3) spectrum of 1b …………………………….28

31 Figure 11 . The 121 MHz P NMR (CDCl 3) spectrum of 3a mixture…………………..30

Figure 12 . GPC overlay of 6b and 6b 1…………………………………………………..34

Figure 13 . Overlay of “one-pot” polymers, 6a-f………………………………………..35

Figure 14 . Overlay of MeOH soluble, insoluble and crude polymer……………………36

vi

Figure 15 . Overlay of 6d , 6f and 6g ……………………………………………………..38

1 Figure 16 . The 300 MHz H NMR spectrum (CDCl 3) of 6c ……………………………39

Figure 17 . Potential B 3 structural units…………………………………………………40

31 Figure 18 . The 121 MHz P NMR (CDCl 3) spectrum of 6d mixture…………………..41

Figure 19 . Overlay of 31 P NMR spectra of 6a-h………………………………………...43

Figure20 Overlay of TGA traces under N 2………………………………...……………45

Figure 21 . Overlay of DSC spectra for samples 6a-h…………………………………...46

vii

List of Schemes

I. Introduction:

Scheme 1 . Divergent and convergent methods for synthesis of dendrimers…………....8

Scheme 2 . Various syntheses of hyperbranched polymers...... 13

Scheme 3 . Generic synthesis of …………...………………………………15

Scheme 4. Generic synthesis of the oligomeric A 2 + B 3 system………………………...16

Scheme 5 . Synthesis of branched PAES via the oligomeric A 2 + B 3 system……………17

II. “One-Pot” Oligomeric A 2 + B 3 Polymerization of Poly(arylene ether sulfone)s:

Reactivity Ratio Controlled Polycondensation

Scheme 6 . NAS mechanism……………………………………………………………..34

Scheme 7 . Synthesis of 1b ……………………………………………………………….27

Scheme 8 . Synthesis of phosphoryl based compounds 3a-c…………………………….29

Scheme 9 . Model Reaction Scheme……………………………………………………..31

Scheme 10 . Conversion of 3 to 2a as an AB 2 ………………………………...32

Scheme 11 . Branched PAEs polymer synthesis…………………………………………33

Scheme 12 . Synthesis of PAEs via “two-pot” literature method………………………..37

Scheme 13 . Synthesis of 6h ……………………………………………………………...37

viii

List of Tables

I. “One-Pot” Oligomeric A 2 + B 3 Polymerization of Poly(arylene ether sulfone)s:

Reactivity Ratio Controlled Polycondensation

Table 1. Reactivity data of 4-Fluorophenyl sulfone, tris(4-Fluorophenyl)phosphine oxide and bis-(4-Fluorophenyl)phenylphosphine oxide……………………………………….26

Table 2 . GC/MS data of model reactions………………………………………………..32

Table 3 . Reaction conditions utilized for the preparation of branched PAEs…………...33

Table 4 . Mn and PDI values of PAEs…………………………………………………...44

Table 5 . Incorporation of B 3 monomer………………………………………………….39

Table 6 . Degree of branching values…………………………………………………….42

Table 7 . TGA and DSC data for 6a-h…………………………………………………...45

ix

Acknowledgements

I would first like to thank Dr. Eric Fossum for his guidance, support and encouragement throughout my career at Wright State University as well as the faculty and staff of the

Chemistry department. Also, many thanks to my husband, Nick, my parents, Jack and

Donna, and to my sister, Laura. This could not have been accomplished without their endless love and support.

x

INTRODUCTION

Linear Polymers

Linear polymers are the most basic of polymers, consisting of only a simple back bone with two end groups. While linear polymers all have the same global structure, their individual structure will influence the conformation they exhibit. For example, poly (p- phenylene terephthalamide) will exhibit a rigid-rod like structure. 1 This means the polymer is in a defined rod-like shape and very little movement is allowed. On the other hand, polymers with a structure which has flexible bonds, as in poly(ether)s, will allow for behavior, meaning there is no defined pattern within the structure of the polymer. While movement in rigid rod polymers is restricted, random coil polymers can participate in chain entanglements, where the chains become intertwined with one another or themselves ( Figure 1 ). Specifically, these chain entanglements bestow

mechanical properties to the polymer.

Figure 1 . Chain Entanglements.

Intermolecular forces can also play a role in the ability of polymers to participate in chain entanglements. While the structure of the polymer may not be rigid on its own,

1

when enough time is allowed, the polymer chains can align their intermolecular forces to form a crystalline material. For example, poly( terephthalate) can either be amorphous or crystalline depending on the time allotted for cooling. If cooled slowly, crystallites will form and the polymer’s amorphous material will be “tied down” by the crystalline material until the crystallites melt. In this semi-crystalline material, only the small portions between crystallites can have chain entanglements, significantly reducing the total number of chain entanglements within the polymer and therefore also changing the mechanical properties.

While many factors influence the ability of a polymer to participate in chain entanglements, the number-average molecular weight (M n) of the polymer is the most important. The lowest M n value a polymer can have and still participate in chain entanglements is known as the Critical Point of Entanglement (M c). The M c for a

specific polymer can be found by plotting viscosity versus M n (Figure 2).

Mc

Mn

Figure 2. Intrinsic or Inherent viscosity versus M n for a generic linear polymer.

2

Below M c, the viscosity is determined through Equation 1 where K L is a constant for low degree of polymerization and Z is the number of monomer units in the polymer backbone. Within this range the viscosity scales with M n to the 1.0 power.

1.0 η = K LZ (1)

Above M c the viscosity is determined through Equation 2. Again, Z is the number of

monomer units in the polymer backbone, while K H represents a constant for high degree

of polymerization. Also note the difference in power with which viscosity scales with

1 Mn, changing from 1.0 to 3.4. This was derived using scaling concepts by de Gennes.

3.4 η = K HZ (2)

An example of a mechanical property is viscoelasticity behavior, which is the ability to relieve various stresses through the capacity of these chain entanglements to adopt different conformations. The viscoelastic behavior of a polymer can be described by

Young’s Modulus.

Figure 3. Young’s Modulus versus temperature plot for a generic linear polymer.

3

Young’s Modulus measures a polymer’s response to an applied stress. This is described in Equation 3 as the stress on the polymer over the strain by the polymer and is simply a way to measure the stiffness of the polymer.

E = ε/σ (3)

For a typical amorphous or semi-crystalline polymer there are five regions within

Young’s Modulus ( Figure 3). Region 1 is the Glassy Region, which is found at low temperatures where polymers exhibit high modulus. High modulus means the material is very hard and brittle. Throughout this region, movement within the polymer is restricted to vibrations and very small rotations. As the temperature is increased, the polymer goes through the Glass Transition (T g), which is shown as region 2. At this point the polymer

gains enough energy to show signs of cooperative motion in the backbone resulting in a

considerable drop in the modulus of the polymer. Next, the modulus flattens as the

polymer continues to warm and enters region 3, the Rubbery Plateau. Within this region

the polymer exhibits elastic behavior. The material has enough energy that it may be

stretched out, causing the chains to straighten. When the stress is removed, the chains

relax back into entropically favorable positions. Region 4, called the Rubbery Flow

Region acts similarly to region 3 when the stress is applied on a short time scale,

however, if the stress is applied over long periods of time, the polymer will not be able to

snap back to its original shape like it would in region three. The final region is the Liquid

Flow Region. Here, at very high temperatures, the polymer has enough energy that the

chains can move relatively quickly past one another allowing for a liquid-like behavior.1

While the physical properties of linear polymers are very attractive for many

applications, most do have one quality that is undesirable: high viscosity. Typically, the

4

viscosity of linear polymers is very high and this becomes a major problem when trying to process the polymer after synthesis. Often, the polymer must be diluted with solvent to obtain a processable material; therefore, it is sought-after to find a similar material that maintains the physical properties, but with a lower viscosity than traditional linear polymers.

Dendrimers

A second class of polymers, dendrimers, differs significantly from linear polymers. Dendrimers are unique because at every monomer unit they are perfectly branched. The structure of dendrimers consists only of dendritic units and chain ends which affords their globular, spherical shape.2 The first dendrimer, Polyamidoamine

(PAMAM), was synthesized by Dow in 1986 ( Figure 4 ). 3 Dendrimers are typically

prepared by either a divergent or convergent approach. As used in the synthesis to

produce PAMAM, the divergent method grows the dendrimer outward by starting with a

core and adding a “layer” of monomer with each reaction to every functional

group at the current chain ends. While this method has proven to be very successful in the

production of many dendrimers including poly(amidoamines) 4, polyethers 5, and

poly(arylamines) 6 there are several downfalls with the method. For example, late in the synthesis, the number of functional groups on the dendrimer required to react with the monomer for proper generation growth is quite high.7 It becomes quite difficult to ensure

that all reactions have occurred, creating an imperfect dendrimer.

5

NH2

N H O

H2N NH2 NH2 H2N O NH2 HN N NH O NH HN O NH O O O NH

N O H N N N N H N O O HN O NH N NH2 H H2N O HN O N HN

N H2N O

HN N NH O N NH2 O O NH O HN O O NH N N N H N NH O O NH2 O O O N HN HN H2N H HN

N O N H2N Chain End HN

H2N

HN O O Dendritic Unit HN

NH2

NH2

Figure 4 . Poly(amidoamine).

An alternative approach, the convergent method, was first introduced by Fréchet and Hawker in 1990.8-10 It begins with the chain ends and works inward toward the center

6

of the molecule to finally attach all “dendritic wedges” to a core molecule forming the completed dendrimer. This method often requires only two reactions for each generation growth 7 thus allowing for a cleaner synthesis and, potentially, a more perfect structure.8,

11, 12 Unfortunately, this method limits the number of generations for the dendrimer as the focal point of the reaction becomes sterically crowded. However, regardless of which method is used to synthesize dendrimers, it is a very tedious process as both require several steps as well as extensive purification after each generation is formed ( Scheme

1).13

7 Divergent Convergent

2x

6x

Scheme 1 . Divergent and convergent methods for synthesis of dendrimers

8

Due to the dense surfaces of dendrimers, they cannot have chain entanglements

with other dendritic . Therefore, dendritic materials generally exhibit poor

mechanical properties and engineering applications are often limited to blending agents

while biological applications range from molecular delivery agents to unimolecular

micelles.14 An interesting characteristic of dendrimers is their intrinsic viscosity. Unlike linear polymers, whose viscosity continues to rise with increasing molecular weight, a dendrimer’s viscosity typically goes through a maximum as the density of the surface increases and is always lower than its linear counterpart.15

Branched Polymers

Branched polymers, are a hybrid of linear and dendritic polymers, both structurally and physically. In Figure 5 a generic structure of a branched polymer is shown and it is evident that these polymers consist of linear, dendritic and teminal units.

Terminal Unit

Linear Unit

Dendritic Unit

Figure 5. Generic representation of a branched polymer

9

Branched polymers are unique because their properties are directly related to the

degree of branching ( DB ). The degree of branching, DB is defined by the number of

linear ( L), terminal ( T), and dendritic ( D) units present in each polymer (Equation 5).

D + T DB = (5) D + T + L

So by changing the number of branching units within the polymer the physical properties should also change. Figure 6 depicts the relationship between mechanical properties and

DB . Linear polymers have no branching (DB = 0) and their mechanical properties are

very good. On the other hand, dendrimers have perfect branching ( DB = 1) while their

mechanical properties are very poor. Branched polymers are unique in that they can be

prepared with varying DB values, thus, it should be possible to tailor their mechanical

properties. For example, if a highly branched polymer is made, it should exhibit more

dendritic-like qualities, while a very lightly branched polymer should exhibit qualities

closer to those of linear polymers.

Figure 6. Comparison of mechanical properties versus degree of branching for linear, dendritic and branched polymers.

10

Along with this, viscosity is another factor that varies with DB . As seen in Figure

7a , linear polymers can have a very high viscosity while dendrimers typically have a very low viscosity. The viscosity of branched polymers can be tuned by either increasing or decreasing the number of branching units. Figure 7b also shows the differences between

linear, dendritic and branched polymers’ viscosity with increasing molecular weight.

Branched polymers fall just below their linear analogs and above their dendritic

counterparts.

(a) (b)

Mw

Figure 7 . (a) Comparison of linear, dendritic and branched polymer viscosity versus

degree of branching and (b) comparison of viscosity versus M w.

As a result of these tunable qualities, branched polymers have the potential to bridge the gap between linear polymers and dendrimers by possessing most of the desirable qualities from each. For example, linear polymers have good mechanical properties but high viscosity, in other words poor processability, while dendrimers have poor mechanical properties but low viscosity, meaning easier processability. Branched polymers can possess the good mechanical properties of linear polymers as well as the

11

ease of processing and increased solubility 16 that exist within dendrimers. However, their polydispersity index (PDI) is quite large in comparison with those of linear and dendritic polymers. 17

Control over the DB within a branched polymer is the key to tuning its properties.

One straightforward route to tailor DB is to control the length between the branching points. Longer linear pieces between branches will allow for chain entanglements which in turn give mechanical properties similar to those of linear polymers, while the branching units maintain the low viscosity associated with dendrimers.

18, 19 There are many methods to make hyperbranched polymers including AB 2, AB

20, 21 22 17, 23, 24 + AB 2 , ABB’, and finally A 2 + B 3 ( Scheme 2). The products of AB n, or AB 2 polymerization reactions are essentially non-perfect dendrimers with very irregular

23 structures. AB 2 monomers are often difficult to find commercially and therefore must

25 be synthesized in what is usually a multi-step process. Another downside to the AB 2 synthesis is the lack of control over the degree of branching that occurs throughout the

26 polymerization. The ABB’ or the AB + AB 2 method is better in that the degree of

branching can be more easily controlled, but again the monomers are difficult to find

16 commercially and also afford irregular structures. Two classes of A 2 + B 3 system exist; monomeric and oligomeric. The A 2 + B 3 method not only allows for better control over the degree of branching but the monomers are often commercially available. 23 While the

A2 + B 3 systems have significant advantages over the other AB n-based methods, there is

one major problem: gelation. 25

12

Scheme 2. Various syntheses of hyperbranched polymers: A) AB 2, B) AB + AB 2 C)

ABB’ and D) A 2 + B 3

An equation published by Flory in 1941 27 describes the point in conversion ( αc) at which point the polymers will form a gel A 2 + B 3 system.

p 2 α = rp 2 = B (4) c A r

Here pA and pB represent the number of B and A functional groups, respectively, that have reacted and r is the ratio of A to B groups. For a 1:1 ratio of A 2:B 3 (2A:3B functional groups) αc is reached at P A = 0.85 and with an A:B functional group ratio of

13

1:1, PA = 0.72. Flory’s equation utilized three assumptions: 1) No intramolecular cyclization was occurring 2) All functional groups had equal reactivity throughout the polymerization, and 3) “A” functional groups could only react with “B” functional groups. 25

In practice, Flory’s final assumption is almost always followed. However, it

might be possible to delay or avoid gelation completely by purposefully allowing

intramolecular cyclization to occur. For example, reaction under dilute conditions will

promote intramolecular cyclization 28 . Another way to avoid crosslinking is to have

unequal reactivity of the functional groups. For example, if one B functional group is

less reactive than another of its type then a more linear polymer will be formed with

minimal branching.

Poly (arylene ether sulfone)s

Polysulfones are engineering that possess excellent thermo and

oxidative stability as well as low dielectric constants 29 , with glass transition temperatures

above 185°C. Polysulfones were introduced commercially in 1965 and are used in hair

dryers, structural foams, electronics and cookware 30, 31 . Astrel ®, Victrex ® and Udel ® are

all examples of commercial polysulfones. Astrel ® is synthesized through the Friedel-

Crafts condensation of biphenyl with oxybis(benzenesulfonylchloride). On the other

hand, Victrex ® and Udel ® use the more common method of polycondensation of

bis(chlorophenyl) sulfone and an assortment of bisphenols under nucleophilic aromatic

substitution conditions.30, 31 While there are several methods to synthesize polysulfones most utilize some variant of the synthetic procedure outlined in Scheme 3 .

14

O

HO Ar OH Cl S Cl

O

O

*O Ar O S *

O

Scheme 3. Generic synthesis of polysulfones.

Branched PAES have been synthesized previously via the condensation of an AB 2

19 monomer, 3,5-difluoro-4’-hydroxydiphenylsulfone, an A 2 + B 3 polymerization of 4,4’- difluorodiphenyl sulfone with phloroglucinol,32 as well as 1,1,1-tris(4-hydroxyphenyl)

ethane with 4,4’-difluorodiphenyl sulfone.33 Unfortunately, all of these methods lead to

polymers with relativly low M n values and also lack any control over the DB . A method has since been developed which allows for control over the DB while maintaining the commercial availability of the monomers; oligomeric A2 + B 3. Examples of materials

34 prepared via oligomeric A 2 + B 3 systems include poly(ether )s, poly (urethane

urea)s,35 and poly(arylene ether sulfone)s 16 .

Typically, the oligomeric A 2 + B 3 method to synthesize branched polymers is accomplished in two steps. The first step is to make the A 2 oligomers by reacting an

excess of A 2 monomer with B2 monomer ( Scheme 4).

15

Scheme 4. Generic Synthesis of the Oligomeric A 2 + B 3 system.

16 The length of the oligomer is controlled by the ratio between the A 2 and B 2 monomers.

Equations 6 and 7 describe how this is possible.

N r = B (6) N A

As seen in Equation 6, the value r describes the ratio of moles of “A” groups (N A) to “B” groups (N B). This value of r, inserted into Equation 7, affords the number of repeat units within the oligomer.

1+ r U = (7) x 1− r

Therefore, by simply altering the ratio of A 2 monomers to B 2 monomers the length of the oligomeric material is changed. Once the oligomers of a desired length have been synthesized they are isolated and reacted with the branching B 3 monomer in the second step to form the branched polymer with controlled length between branch points. While gelation is still an issue with this method of synthesis, a modified version of Flory’s equation is used.27

16

p A pB ρ α c = (8) 1− p A pB 1( − ρ)

The values pA and pB again represent the number of B and A functional groups respectively that have reacted and ρ is the ratio of B groups belonging to branching units

to the total number of B units.

The current project describes a reactivity ratio controlled process for the synthesis

of branched poly(arylene ether sulfone) polymers (PAES) via the oligomeric A 2 + B 3 system, but provides the desired materials in one pot, without the isolation of the A 2 oligomers. This can be achieved by selectively reacting an A 2 monomer, Bisphenol-A

(2) with B 2 monomers through the difference in reactivity between the B 2 and B 3 monomers.

OH

O O NMP F S F F P F H C CH 3 3 O 1.5 eq K2CO3 ∆ 1 2 F 3 OH

Scheme 5 . Synthesis of branched PAES via the oligomeric A 2 + B 3 system.

In this case, the more reactive B 2 monomer is 4-fluorophenyl sulfone ( 1) and the less

reactive B 3 monomer is tris-(4fluorophenyl)phosphine oxide ( 3). Carter has shown through 13 C and 19 F NMR spectroscopy that a sulfone group affords a higher reactivity

than the phosphoryl by almost 250:1. 36

16 Following the literature procedure introduced by Lin et al, the A 2 and B 2 monomer molar ratio will afford PAES oligomers with approximately 15 repeat units

17

which corresponds to an M n ~ 4000 g/mol. Having this molecular weight for the

oligomers is important because it has been shown that the M c for PAES is 2000-3000 g/mol 16 . After oligomer synthesis, the temperature of the reaction mixture will be

increased to permit the B 3 unit to react and provide the desired branched polymer.

The one pot approach will provide a number of advantages over the currently available methods including 1) No requirement for separation and purification of oligomeric material and 2) A one-pot synthesis is very important as it saves time, money and solvent use. Even though monomer 1 is more expensive than its counterpart, 4-

chlorophenyl sulfone, the benefits of a one step reaction should far outweigh the

increased cost of the monomer.

18

EXPERIMENTAL

General.

All reactions were performed under a atmosphere. All transfers were done with syringes as needed. 1H, 13 C and 31 P NMR were acquired on a Bruker Avance

300 MHz NMR operating at 300, 75.5 and 121 MHz respectively. Samples were dissolved in chloroform-d. GPC data were acquired utilizing a Viscotek Model 300 TDA

GPC equipped with refractive index, viscosity and light scattering detectors operating at

70 °C. Polymer Laboratories 5 µm PL gel mixed C columns were used with N-methyl pyrrolidinone (NMP) (0.5% LiBr) as the eluent and a Thermoseparation Model P1000

Pump operating at 0.8mL/min. GC/MS data were collected on a Hewlett Packard 6890

Series GC System with an Agilent 6890 Series Injector and Hewlett Packard 5973 Mass

Selective Dectector. TGA and DSC data were collected on TA instruments TGA Q500 and DSC Q200 models, respectively, at a heating rate of 10 °C/min.

Materials.

NMP was purchased from Sigma Aldrich and was distilled from CaH 2 under a nitrogen atmosphere prior to use. Tetrahydrofuran (THF) and Methanol (MeOH) were used as received from Sigma Aldrich. ACS certified was used as received from

Fischer Scientific. Tris (4-fluorophenyl)phosphine oxide was prepared according to

literature procedure.37 Bis (4-fluorophenyl)phenylphosphine oxide was provided by

DayChem Laboratories. 4-Fluorophenyl sulfone and Bisphenol-A were purchased from

Sigma Aldrich and recrystallized from hexanes and ethanol/water, respectively, followed by drying in vacuo prior to use.

19

Synthesis of 4,4’-Bis-(4-methylphenoxy)diphenyl sulfone , 1b

In a 10 mL round bottomed flask equipped with a condenser, stir bar and gas inlet were placed 500 mg (1.97 mmol) of 4-fluorophenyl sulfone, 1, 467 mg (2.2 eq, 4.32

mmol) of 4, 897mg (1.5 eq, 3.21 mmol) of K2CO 3 and 6.2 mL of NMP. The reaction

mixture was then heated to 180 °C and held for 1.5 h. The reaction mixture was cooled

to room temperature and slowly poured into a large excess of acidic (pH ~ 2) water. The

resulting tan solid was isolated via vacuum filtration to afford 550 mg of 1a (65%). A sample was taken for GC/MS analysis, which showed complete conversion to 1a . 1H

NMR (CDCl 3, δ): 2.4 (s, 6 H CH 3), 6.9-7.0 (m, ArCH), 7.2 (d, ArCH), 7.85 (d, ArCH).

13 C NMR (CDCl 3, δ): 21 (CH3), 117.5 (ArCH), 120.5 (ArCH), 129 (ArCH), 131 (ArCH),

134.5 (ArCCH3), 135 (ArCSO 2), 152.5 (ArCO), 162 (ArCO).

Synthesis of Mono-substituted tris-(4-fluorophenyl phosphine oxide, 3a

In a 5 mL round bottomed flask equipped with a condenser, stir bar and gas inlet were placed 300 mg (0.30 mmol) of tris -(4-fluorophenyl) phosphine oxide, 3, 127 mg

(1.2 eq. 0.361 mmol) of p-cresol, 4, 225 mg (1.5 eq, 0.54 mmol) of K 2CO 3, and 3 mL of

NMP. The resulting mixture was heated to 140 °C and held there for 20 h at which point the reaction mixture was cooled to room temperature and slowly poured into a large excess of acidic (pH ~ 2) water. The resulting brown solid was isolated via vacuum filtration. A sample was taken for GC/MS, which showed the presence of 17.1 % 3, 82.8

1 % 3a , and 0.1 % di-substituted monomer, 3b . H NMR (CDCl 3, δ): 2.4 (s, CH 3), 6.9-7.15

31 (m, ArCH), 7.1-7.3 (m, ArCH), 7.5-7.75 (m, ArCH) P NMR (CDCl 3, δ): 27.0, 27.5,

28.0, 28.4.

20

Representative Procedure for Model Reactions

A 10 mL round bottomed flask equipped with a stir bar, condenser and gas inlet was charged with 80 mg (0.24 mmol) of 3, 414 mg (3.83 mmol) of 4, 426 mg (1.67

mmol) of 1, 1.6 g of K2CO3 (1.5 eq, 5.83 mmol) and 5.2 mL of NMP. The reaction mixture was heated at 100 °C for 13 h. The progress of the reaction was monitored by removing aliquots for GC/MS analysis and the resulting data are summarized in Table 2.

Variations on this procedure can also be found in Table 2.

Representative Polymerization Procedure

A 10 mL round bottomed flask equipped with a stir bar, condenser and gas inlet was charged with 75 mg (0.22 mmol) of 3, 410 mg (1.80 mmol) of 2, 400 mg (1.57

mmol) of 1, 750 mg (1.5 eq, 2.70 mmol) of K2CO 3 and 5 mL of NMP. The reaction

mixture was heated at 160 °C for 4 h after which the temperature was increased to 180 °C

for an additional 5 h. The reaction mixture was cooled to room temperature and slowly

poured into a large excess of water. The resulting white solid was isolated via vacuum

filtration. Subsequent reactions were carried out at a series of temperatures as listed in

1 Table 3 . The corresponding GPC data can be found in Table 4. H NMR (CDCl 3, δ): 1.7

31 (s CH 3), 6.9-7.1 (m, ArCH), 7.2-7.3 (m, ArCH), 7.5-7.7 (m, ArCH), 7.85 (d, ArCH). P

NMR (CDCl 3, δ): 27.4, 27.9, and 28.4.

General Procedure for Reverse Precipitations

Polymer 6e (250 mg) was dissolved in the minimum amount of THF required.

While the solution was stirred vigorously, MeOH was added dropwise until the highest

Mn polymer precipitated out of solution while the lower M n materials remained soluble in

21

the MeOH. The high M n polymer was isolated via filtration followed by drying in vacuo .

The MeOH layer was evaporated to isolate the low M n material as an off-white solid.

GPC traces for the samples isolated from this process are shown in Figure 14 .

Preparation of GC/MS Calibration Curve

16 mg (0.021 mmol) of model compound 3a (also containing 3, 3b , and 3c ) and 9 mg (0.021 mmol) of model compound 1b were weighed directly into a GC/MS vial and diluted with 1.5 mL of acetone. The mixture was then analyzed through GC/MS.

Polymerization of A 2 + B 2 + B’ 2 monomers, linear analogue.

A 10 mL round bottomed flask equipped with a stir bar, condenser and gas inlet was charged with 71 mg (0.22 mmol) of 3d, 410 mg (1.80 mmol) of 2, 400 mg (1.57

mmol) of 1, 750mg (1.5 eq 2.78 mmol) of K 2CO 3, and 5 mL of NMP. The reaction mixture was heated at 120 °C for 11 h after which time the temperature was increased to

180 °C for an additional 5 h. The reaction mixture was cooled to room temperature and slowly poured into a large excess of water. The resulting tan solid was isolated via vacuum filtration. The corresponding GPC data can be found in Table 4. 1H NMR

31 (CDCl 3, δ): 1.7 (AlCH) 6.9-7.1 (m, ArCH), 7.25 (d, ArCH), 7.85 (d, ArCH). P NMR

(CDCl 3, δ): 28.2, 28.6.

Synthesis of A 2 capped oligomers, 2a.

A 12 mL round bottomed flask equipped with a stir bar, Dean-Stark trap, water condenser and gas inlet was charged with 345 mg (8 eq, 1.52 mmol) of 2, 300 mg (7eq,

1.18 mmol) of 1, 654 mg (1.5 eq 2.43 mmol) of K2CO 3, 3.7mL of toluene and 3.7 mL of

NMP. The reaction mixture was heated at 140 °C for 2 h after which time the toluene was removed and the temperature was increased to 170 °C for an additional 6 h. The reaction

22

mixture was cooled to room temperature and slowly poured into a large excess of acidic

(pH ~ 2) water. The resulting brown solid was isolated via vacuum filtration and dried in an 80 °C drying pistol for 24 h. The corresponding GPC data can be found in Table 4.

1 H NMR (CDCl 3, δ): 1.7 (CCH3), 6.75 (d, ArCH), 6.85-7.1 (m, ArCH), 7.15-7.3 (m,

ArCH), 7.85 (d, ArCH).

Synthesis of Branched Poly(arylene ether sulfone)s, 6g.

A 10 mL round bottomed flask equipped with a stir bar, Dean-Stark trap, water condenser and gas inlet was charged with 400 mg (0.106 mmol) of 2a, 35 mg (0.106 mmol) of 3, 44mg (1.5 eq, 0.16 mmol) of K2CO 3, 1.1 mL of toluene and 1.1 mL of NMP.

The reaction mixture was heated at 140 °C for 2 h after which time the toluene was

removed and the temperature was increased to 170 °C for an additional 6 h. The reaction

mixture was cooled to room temperature and slowly poured into a large excess of water.

The resulting brown solid was isolated via vacuum filtration and dried in an 80°C drying

1 pistol for 24 h. The corresponding GPC data can be found in Table 4. H NMR (CDCl 3,

δ): 1.7 (CH 3), 6.8 (d, ArCH), 6.9-7.1 (m, ArCH), 7.3-7.4 (m, ArCH), 7.5-7.7 (m, ArCH),

31 7.85 (d, ArCH). P NMR (CDCl 3, δ): 28.2, 28.6.

23

RESULTS AND DISCUSSION

Nucleophilic aromatic substitution, NAS, is a very versatile and important synthetic pathway utilized in polycondensation reactions. The mechanism for NAS begins with a reversible step in which the nucleophile attacks the electrophilic site on the aromatic ring.

This forms a resonance stabilized anionic intermediate known as the Meisenheimer complex.

In the second step, the leaving group is displaced to restore aromaticity to the ring ( Scheme

6). NAS is typically aided by the presence of strong electron withdrawing groups located in the R 1 positions, which help stabilize the Meisenheimer complex through inductive and resonance effects. The most common leaving groups are fluoride or chloride and they serve the additional purpose of helping to stabilize the anionic intermediate.

X X R1 X - Nuc X R1 Nuc

R1 = -SO2-, -CO-, -PhPO- X = Cl, F

- X X R1 Nuc X X R1 Nuc

X X

X R1 X R1 Nuc Nuc

Scheme 6. NAS Mechanism.

Reactivity Determination.

In order to achieve the main goal of this project, the synthesis of branched poly(arylene ether)s via a “one-pot” oligomeric A 2 + B 3 methodology, the concept of

reactivity controlled polycondensation was applied. The desired outcome requires the

24

selective reaction of bisphenol-A (A 2), with (4-fluorophenyl) sulfone, B 2, in the presence of tris -(4-fluorophenyl)phosphine oxide (B 3), 3, to generate the oligomeric A 2 component in situ , followed by incorporation of the B 3 component at a higher reaction temperature. Thus, the first phase of the project was to determine the relative reactivity of the B 2 and B 3 monomers, under typical NAS conditions.

Carter published a thorough study in which the relative reactivity, toward NAS, for a wide range of aryl difluoride monomers, was evaluated by NMR spectroscopy as well as model reactions 36 . It was determined that the chemical shifts of the ipso carbons can be used

to determine the relative reactivity of monomers toward NAS reactions. For example, the

more downfield a peak appears in the spectrum, the less electron density is found on that

atom and thus, it is more likely to participate in NAS. A similar correlation was observed

19 with the F NMR data. Figure 8 shows the structures of the B 2 and B 3 components while

the chemical shifts of the ipso carbon atom and atom, respectively, as well as the

relative reactivity data are tabulated in Table 1.

O O O 1 F S F 3 F P F 3d F P F O

F

Figure 8. 4-Fluorophenyl sulfone, 1, tris –(4-Fluorophenyl) phosphine oxide, 3, and its B 2 analogue, bis –(4-Fluorophenyl)phenyl phosphine oxide, 3d .

25

Table 1 . Reactivity data of 4-Fluorophenyl sulfone, Bis –(4-Fluorophenyl) phenyl phosphine oxide and bis -(4-Fluorophenyl) phenyl phosphine oxide.

Relative Reactivity a 13 C NMR Shift 19 F NMR Shift

1 57140 165.31 ppm -104.08 ppm

3 N/A 165.16 ppm N/A

3d 230 165.05 ppm -106.71 ppm

(a) relative to 2,3-Bis (4-Fluorophenyl) quinoxaline.

While tris -(4-Fluorophenyl) phenyl phosphine oxide, 3, ( Figure 8) was not evaluated by Carter, some indication of its relative reactivity can be inferred from the corresponding B 2 analogue, bis -(4-Fluorophenyl) phenyl phosphine oxide, 3d . The 13 C data for 3 are provided in Table 1. 19 F NMR data are unavailable but should be comparable to 3d, which shows that there is a significant difference in reactivity between monomers 1 and 3d and 1 should react approximately 250 times faster than 3d . The relative reactivity ratios of the compounds in

Table 1 are correlated best with 19 F NMR spectroscopic data. Aryl fluorides have a wider

range of chemical shifts than found in 13 C NMR allowing for a clearer understanding of

reactivity.

GC/MS Calibration Curve

The actual reactivity of B 2 and B 3, under the conditions utilized in this project, were

determined via a series of model reactions, the products of which were analyzed by gas

chromatographic/mass spectroscopic, GC/MS, analysis. In order to ensure accurate data

from the GC/MS analyses, a calibration curve was first constructed with independently

prepared authentic samples. Compound 1b was synthesized according to the route outlined in

Scheme 7. A slight excess of p-cresol, 4, was reacted with 1 for 1.5 hours in NMP at a reaction temperature of 180 °C, to afford 1b , which represents the oligomeric A 2 component.

26

CH3 O O F S F NMP H3C O S O CH3 1.2 eq K CO O 2 3 O o OH ∆ 180 C 1 1b 4

Scheme 7. Synthesis of 1b

NMR spectroscopy was used to confirm the structure of 1b . The 1H and 13 C NMR

spectra for 1b showed that no other products were formed and also confirmed no starting

material was present. Figure 9 shows the 1H NMR for compound 1b and as expected there is

a singlet (a) at 2.4 ppm, which corresponds to the aliphatic protons, while the aromatic

protons (b-e) ranged from 6.8-7.8 as multiplets. The signals assigned to c and d appear as an

overlapping multiplet.

a

e c, d

b

1 Figure 9. The 300 MHz H NMR spectrum (CDCl 3) of 1b .

27

The 13 C NMR spectrum also showed the expected number of signals. The aliphatic

carbon (a) appeared near 20 ppm. Peaks b-i represented the aromatic carbons, four of which

are quaternary and therefore appear much less intense than the rest.

CDCl 3 g d

h c

a

f b i e

13 Figure 10 . The 75 MHz C NMR (CDCl 3) spectrum for 1b .

The synthesis of the model compounds derived from 3 was carried out as depicted in Scheme

7. Reaction of 3 with a slight excess (1.2:1) of 4 for 4 hours in NMP at a reaction temperature of 140 °C resulted in a mixture of 3, 3a , 3b , and 3c .

28

CH O 3 O F P F NMP F P F 1.2 eq K2CO3 ∆ 140oC OH 3 3 F 4 F O

F P O CH3

3a

F

O

H3C O P O CH3

3b

F O

H3C O P O CH3

3c

O

CH3

Scheme 8. Synthesis of phosphoryl based compounds ( 3a -c).

According to GC/MS data the reaction mixture contained only the compounds 3, 3a and 3b . However, 31 P NMR showed the presence of these three products as well as the tri-

substituted B 3 monomer, 3c , in the ratio of 1.6:15.2:10.2:1 as shown in Figure 11 . The peak

assignments were based on literature reports of A 2 + B 3 polycondensation reactions utilizing

3 as the branching unit. 17

29

3a

3b

3 3c

31 Figure 11 . The 121 MHz P NMR (CDCl 3) spectrum of monomer 3 reaction mixture.

Approximately equimolar amounts of 1b and 3a were used to create a calibration

curve for the GC/MS data, which should show a 53:47 mixture of the two compounds if

properly calibrated. However, the calibration curve showed a 60:40 mixture of 1b :3a ,

revealing an under representation of 3a and over representation of 1b . Subsequent GC/MS

analyses were corrected accordingly.

Simulation of in situ Oligomeric A 2 Formation .

Monomers 1 and 3 were reacted competitively at various temperatures with the

“model” nucleophile, p-cresol, 4, which should possess similar nucleophilic properties to bisphenol-A, 2, but is not able to polymerize because there is only one functional group per molecule. This affords the opportunity to determine the relative reactivity of the monomers under our specific reaction conditions without any competing polymerization occuring. At

30

specific time intervals during the course of the model reactions aliquots were removed for

GC/MS analysis.

O O CH3 F S F F P F NMP O

1.5 eq K2CO3 1 3 4 ∆ OH F

O O R O S F R O S O R O O 1a 1b

O O F P O R R O P O R

3a 3b

F F

O

R O P O R R = CH3

3b

O R

Scheme 9. Model Reaction Scheme

Table 2 shows the GC/MS data of the model reactions performed at reaction temperatures ranging from 100 to 200 °C. The reaction was considered complete when all of

1 was converted to 1b . While the NMR data indicated that 1 reacts at a much faster rate than

3, the unknown was if 1 could be substituted twice to form 1b before 3 would begin to react, thus allowing for the exclusive formation of A 2 oligomers under polymerization conditions.

31

Table 2 . GC/MS data for model reactions.

Rxn Ratio T (°C) T (h) % 1a % 1b 3% % 3a %3b % 3c 5a 16:7:1 100 13 0 84.2 14.7 1.1 0 0 5b 16:7:1 120 11 13.2 43.9 41.7 1.2 0 0 5c 16:7:1 140 7 0 78.5 18.2 3.3 0 0 5d 16:7:1 160 3.5 0 82.6 14 3.4 0 0 5e 16:7:1 180 1 0 81.36 14.6 4 0 0 5f 16:7:1 200 0.5 0 77.5 15.3 7.2 0 0

While it was not possible to convert 1 to 1b without any conversion of 3, the data clearly show that only 3a was produced and no 3b . This means that under polymerization conditions, when 3 would react, it should do so only to form an AB 2 type oligomer ( Scheme

10 ) and therefore the B 3 component should be a terminal unit or end group on the A 2 oligomer.

CH3 O O HO OH F P F F S F O CH3

2 3 1 F

CH3 O CH3 O HO O S O O P F

CH3 O CH3 n

2a F

Scheme 10 . Conversion of 3 to 2a as an AB 2 oligomer.

As expected, the data also show that as the temperature increased the time required to fully convert 1 to 1b decreased. Unfortunately, at higher temperatures, the amount of 3a that was

produced also increased. For example, the reaction at 100 °C showed a ratio for 1b:3a of

107:1, while at 180 °C the same ratio was 27:1. While this was not ideal, since again no 3b

was produced, this simply meant more AB 2 oligomer would form within the polymerization.

32

O CH3 O NMP F S F HO OH F P F O 1.5 eq K2CO3 CH3 ∆ T1 1 2 3 ∆ T2 F

CH3 O CH3 O H O O S O O P F m CH3 O CH3 n

6a-f F

Scheme 11. Branched PAES polymer synthesis.

Synthesis of Branched Poly(arylene ether sulfone)s .

The reactivity ratio data gleaned from the model reactions were then applied to the

polymer synthesis phase of the project. Scheme 11 shows the synthesis of branched

poly(arylene ether)s. Polymer synthesis was based on the model reaction times and

temperatures; for each T 1 temperature the time, t1, was the same as was required to convert 1 to 1b in the model reactions. With the exception of 6f , which was held constant at 200 °C, T2 for all of the polymerization reactions was 180 °C. Table 3 summarizes the conditions utilized for each polymerization reaction.

Table 3 . Reaction conditions utilized for the preparation of branched PAES samples.

6a 6b 6c 6d 6e 6f

t1 (h) 13 11 7 4 1 0.5

T1 (°C) 100 120 140 160 180 200

t2 (h) 5 5 5 5 5 4

T2 (°C) 180 180 180 180 180 200

As the M c values for polysulfones are in the range of 2000-3000 g/mol, it was ideal

for the A 2 oligomer to possess a M n value near or above this value, thus ensuring the possibility for chain entanglement. After the allotted t1 time, a sample was taken for GPC

33

characterization and while most samples were above the threshold, the M n values for 6a and

6b oligomeric species were slightly below the desired value ( Table 4). Therefore t 1 was

increased by a factor of 1.8 ( 6a 1 and 6b 1 in Table 4) to allow for continued oligomer growth.

Table 4 . M n and PDI values of PAES .

Oligomer Polymer HMw LMw a a a a Mn PDI Mn PDI % Mn PDI % Mn PDI 6a 1,100 1.1 16,600 13 84 24,100 6.4 86 3,100 2.7 6a1 1,700 1.2 13,500 2.4 80 17,400 2.1 76 3,500 1.9 6b 2,000 1.3 16,800 11.7 80 77,400 4 64 7,800 7.5 6b1 3,900 1.5 14,000 3.3 93 24,900 2.5 77 12,700 2.8 6c 2,500 1.6 17,000 8.4 92 84,700 3.2 75 12,300 3.1 6d 6,023 1.7 17,500 14.1 82 41,000 6.8 83 3,800 2.4 6e 4,300 1.9 17,100 25.6 82 28,400 16.3 80 3,000 2 6f 4,800 1.7 18,147 16.7 88 85,000 5 78 6,800 2.7 6g 4,700 1.6 12,800 10.5 91 47,000 3.8 56 8,100 2.6 6h 2,800 1.3 9,800 2.5 98 17,200 1.6 53 7,600 1.6

( a) g/mol

The increased time for oligomer synthesis in 6b 1 proved to be successful as the M n value,

3,900 g/mol, was above the critical M c value. An overlay of the GPC traces for 6b and 6b 1 is

shown in Figure 12 . On the other hand, 6a 1, had a slight increase in molecular weight to

1,700 g/mol, but did not rise above M c.

b a

Figure 12 . GPC overlay of (a) 6b and (b) 6b1 .

34

------6a ------6b ------6c ------6d ------6e ------6f

Figure 13 . Overlay of “one-pot” polymers 6a-f.

The final polymers ( 6a-f) were all fully soluble in a variety of solvents including

NMP, chloroform, and THF. All of the samples had Mn values near 17,000 g/mol and PDIs

ranging from 8 to 25, with the exception of 6a 1 and 6b 1, which were slightly lower. Branched

polymers synthesized from A 2 + B 3 polycondensation reactions are known to have very large

PDI values, which result from a broad spectrum of molecules with varying weights. To

lower the PDI of these polymers they were subjected to a purification step called reverse

precipitation, which involves dissolving the samples in THF and slowly adding MeOH to

precipitate the higher molecular weight material selectively. Reverse precipitation allows for

the removal of lower molecular weight material, which remains soluble in MeOH, while the

larger molecules precipitate out of solution. The effects of this process can easily be seen

when comparing the GPC trace of a ‘crude’ polymer, which has simply been precipitated

35

from water with one that has been reverse precipitated. Figure 14 shows GPC traces of the

‘crude’ polymer as well as the MeOH soluble and insoluble samples.

b a c

Figure 14 . Overlay of the (a) MeOH soluble and (b) MeOH insoluble as well as (c) crude

polymer.

To provide a basis for comparison of the physical properties of the branched PAES materials

prepared via the “one-pot” polymerization method, two additional samples were produced,

the first being a branched PAES synthesized via the literature “two-pot” method, 6g ( Scheme

16 12) . This polymerization process gave A 2 oligomers with similar M n and PDI values to

those from the “one-pot” polymerizations. However, the final branched polymer, fully

soluble, did have a slightly lower M n value at 12,800 g/mol and a PDI, 10.5, which was

within the same range as observed for samples 6a-f. The reverse precipitation of 6g exhibited

similar results to the one-pot” polymerizations as the M n increased significantly while the

PDI decreased.

36

Step One

O CH3 NMP/Toluene F S F HO OH 1.5 eq K CO O CH 2 3 3 ∆ 1.140oC o 1 2 2.170 C

CH3 O CH3 HO O S O O H 2b

CH3 O CH3 n

Step Two

CH3 O CH3 O HO O S O O H F P F CH3 O CH3 n

2b 3 F

CH O CH O NMP/Toluene 3 3 H O O S O O P F 1.5 eq K2CO3 m CH3 O CH3 n ∆ 1.140oC o 2.170 C 6g F

Scheme 12. Synthesis PAES via the “two-pot” literature method.

The second polymer, a linear analog, 6h , was made to compare the effects of

branching present in the “one-pot” branched polymers on their thermal properties. As

depicted in Scheme 13 , 3d was used in place of 3 so that no branching could occur. The M n and PDI values of the oligomeric material of 6h were similar to those found in 6c , though the

Mn of 6h polymer was significantly lower at 9,800 g/mol, however, as expected, the PDI for

this sample was considerably lower, 2.5. The MeOH insoluble material from reverse

precipitation possessed a M n value of 17,000 g/mol with a PDI of 1.6.

O CH3 O NMP F S F HO OH F P F 1.5 eq K2CO3 O CH3 ∆ 1.120oC 11h o 1 2 3d 2.180 C 5h

Scheme 13. Synthesis of 6h , linear analog to 6a-f.

37

Figure 15 shows an overlay of 6d along with the two polymer samples made for comparison,

6g-h. Sample 6d showed the highest M n value followed by 6g and 6h .

a b c

Figure 15 . Overlay of (a) 6d , (b) 6g and (c) 6h.

Incorporation of B 3 Monomer

1H NMR was used to analyze how much of the branching unit, 3, was incorporated

into the polymer. If all of monomer 3 was incorporated into the polymer, the ratio between

aromatic and aliphatic protons should be approximately 2.75:1 and if none of 3 was incorporated into the polymer, the ratio of aromatic to aliphatic protons becomes 2.5:1. The integration from the 1H NMR spectra showed in all polymer samples that 3 was in fact

included as the values ranged from 2.65 in 3b 1 to 3.11 in 3d ( Table 5). An example of this

1H NMR integration is shown in Figure 16 . The arrow in Figure 16 points to the peaks

1 corresponding to the B 3 monomer, which are not present in the H NMR spectra of the A 2 oligomeric species.

38

Aliphatic

Aromatic

* † † * * *

1 Figure 16 . The 300 MHz (CDCl 3) H NMR spectrum (CDCl 3) of 6c (* represents signals from residual NMP and † represents signals from residual THF)

Table 5 . Incorporation of B 3 monomer.

6a 6a1 6b 6b1 6c 6d 6e 6f 6g 6h

HAr :H aliph 2.82 2.83 2.99 2.65 2.83 3.11 2.69 2.78 2.93 2.92

Estimation of the Degree of Branching, DB.

In a branched polymer there are three possibilities of substitution for the branching unit; terminal with one substitution, linear with two substitutions and dendritic with three substitutions ( Figure 17 ). The number of each unit found within a polymer allows for an estimation of the degree of branching ( DB ) using Equation 5, in which D = dendritic , T = terminal , and L = linear, respectively.

39

Terminal Linear Dendritic

O O O O P F O P O O P O

F F O

Figure 17 . Potential B 3 structural units.

D + T DB = ×100 (5) D + T + L

Integration of the 31 P NMR spectra was used to determine the relative amounts of each type of branching unit present in the polymer samples.17 An example of a 31 P NMR

spectrum is shown in Figure 18 with integration. These integration values were then placed

into equation 5 to determine the local DB of 44%. The DB values for all polymer samples are found in Table 6. All of the “one-pot” polymers possessed DB values between 44 and

48%, with the exception of 6a 1, which showed no dendritic units. 6h does not have a DB as there is no branching unit within the polymer.

40

28.26 27.83 27.40

Linear

Terminal

Dendritic

31.0 30.5 30.0 29.5 29.0 28.5 28.0 27.5 27.0 26.5 26.0 25.5 25.0 24.5 24.0 ppm

0.34 1.00 0.46

31 Figure 18 . The 121 MHz P NMR (CDCl 3) spectrum of polymer 6d .

An important concept to determining the DB is the fact that each time a branching unit is formed; one more terminal unit is also formed. This means, in the absence of any cyclization, that the dendritic and terminal units should have the same integration value and therefore Eq 5 can be modified so that only dendritic and linear units need to be counted,

Equation 9.

2D DB = ×100 (9) 2D + L

41

Table 6 . Degree of Branching. ( a) determined using Eq 5, ( b) determined using Eq. 9, ( c)

determined using Eq 10, ( d) determined using Eq 12.

DB DB DB G DB G a b c d D L T (%) (%) LA2 (%) (%) 6a 0.51 1.00 0.41 48 51 15 5 6 6a1 0.00 1.00 1.00 50 0 15 6 0 6b 0.36 1.00 0.49 46 42 15 5 4 6b1 0.35 1.00 0.61 49 41 15 6 4 6c 0.46 1.00 0.48 48 48 15 6 5 6d 0.34 1.00 0.46 44 40 15 5 4 6e 0.42 1.00 0.39 45 45 15 5 5 6f 0.33 1.00 0.48 45 40 15 5 4 6g 0.87 1.00 0.21 52 64 8 11 16 6h 0.00 1.00 0.86 46 0 15 5 0

Table 6 shows that 6g , the “two-pot” sample, exhibits a higher DB across the board but the value was significantly higher when using the equations which only take dendritic and linear units into account. The amount of dendritic units found in 6g in relation to terminal units implies that a significant amount of intramolecular cyclization has occurred as these two values should be similar. Dilute reaction conditions are often used to avoid gelation but this also enhances intramolecular cyclization. As 6g was produced under more dilute conditions than the “one-pot” samples the DB data are expected. An overlay in Figure 19 shows the small change in each type of branching unit for the “one-pot” polymers, 6a-f.

However, samples 6g-h had slightly different chemical shifts in comparison to 6a-f. The relatively high degree of cyclization found in 6g may have an effect on the chemical shift and as 6h has a different structure than found in the other samples, this chemical shift is to be expected.

42

6a

6b1

6c

6d

6e

6f

6g

6h

Figure 19 . Overlay of 31 P NMR spectra of 6a-h.

The previous methods to determine DB only account for the local degree of

branching, relative to the B 3 monomer. In an oligomeric A 2 + B 3 system the standard

equations do not account for the linear portions of the polymer. Therefore a modified

equation was used to account for these segments, equation 10 where D B3 , T B3 , and L B3 are the

16 dendritic, terminal and linear units respectively for the B 3 monomer. L A2 accounts for the

linear units of the A 2 oligomers.

DB3 + TB3 DB G = ×100 (10) DB3 + TB3 + LB3 + LA2 where LA2 is defined as,

[A2 ] LA2 = × n (11) [B3 ]

43

[A 2] and [B 3] are the molar concentrations of A 2 oligomers and B 3 monomers, respectively,

and n is the number of repeat units within the oligomeric units.

While there are two methods to calculate local DB, the first utilizing dendritic, linear and terminal units and the second utilizing only the dendritic and linear units, the same is true for global DB . Equation 10 uses all three branching possibilities while Equation 12 uses only

dendritic and linear units. Global DB, DBG, values are provided in Table 6 showing that all

of the polymers possess DBG values between 5-6%.

2DB3 DB G = ×100 (12) 2DB3 + LB3 + LA2

Thermal Analysis.

Thermal Gravimetric Analysis (TGA) is a method to determine the thermal stability

of a material and the results are often reported as the decomposition temperature (T d). Four

Td values were measured for each sample corresponding to 5 and 10% losses under both nitrogen and air. Polysulfones typically are used as high temperature thermoplastics and therefore have a very high T d. An overlay of the TGA curves for the “one-pot” polymer

samples is shown in Figure 20 , which shows the decomposition temperatures for each sample under a nitrogen atmosphere. All samples showed very high T d temperatures with

most samples decomposing above 450 °C. A comparison of T d temperatures for the “one

pot” samples and the linear analog, 6h , showed very similar temperatures. However, the

“two pot” sample, 6g , had a significantly lower T d value of 440 °C for 5% loss in N 2.

44

Figure 20 . Overlay of TGA traces under a N2 atmosphere at a heating rate of 10 °C/min.

Table 7 . TGA and DSC data for polymer samples 6a-h.

Td (5 %) Td (10 %) Td (5 %) Td (10 %) Tg o o o o o N2 ( C) N2 ( C) air ( C) air ( C) ( C) 6a 480 491 448 454 195 6a1 465 471 464 475 198 6b 459 469 454 459 199 6b1 472 482 460 477 200 6c 456 467 458 460 201 6d 480 489 452 458 198 6e 463 471 447 452 193 6f 455 465 441 445 199 6g 440 453 438 449 211 6h 470 486 472 489 187

Differential scanning caliorimetry (DSC) was used to determine the glass transition

temperature (T g) of each high molecular weight polymer sample. The data in Table 7 show

45

“one-pot” samples have T g values ranging from 193-201 °C. Overall, for these samples there

is an increase in T g with increasing T 1 temperatures until it reaches a maximum at 6c then begins to decrease as the T 1 value continues to rise. Branching is known to affect T g in such a way that increased branching typically results in an increased T g. The “two-pot” sample,

6g , shows a much higher T g value than 6a-f, which leads to the conclusion that this polymer

sample has more branching or that the length of the linear segment between branch points is

shorter, thus restricting cooperative molecular motion. The opposite holds true for the linear

analog, which has no branching and the lowest T g value.

Figure 21 . Overlay of DSC traces for “one-pot” samples 6a-h.

46

Conclusions.

Branched poly(arylene ether)s were successfully synthesized through a “one-pot” oligomeric A 2 + B 3 method. Oligomers of a desired length were prepared in the presence of a B 3 branching unit, after which a temperature increase allowed for the synthesis of fully soluble branched polymers. The polymers exhibited high T d values and T g values similar to

commercial PAES systems, both desirable qualities for high temperature thermoplastics. The

“one-pot” system reduces solvent waste as well as overall reaction time. Future work

includes testing mechanical properties of the polymers with dynamic mechanical analysis

(DMA) as well as reaction scale-up studies.

47

References

1. Sperling, L.H., Introduction to Physical . Fourth ed. 2006,

Hoboken: John Wiley & Sons, Inc.

2. Hawker, C.J. and Frechet, J.M.J., Monodispersed Dendritic with

Removable Chain Ends: a Versatile Approach to Globular with

Chemically Reversible Polarities. J. Chem. Soc. Perkin Trans. 1, 1992: p. 2459-

2469.

3. Tomalia, D.A., et al., Dendritic Macromolecules - Synthesis of Starburst

Dendrimers. Macromolecules, 1986. 19 (9): p. 2466-2468.

4. Tomalia, D.A., et al., A New Class of Polymers - Starburst Dendritic

Macromolecules. Polymer Journal, 1985. 17 : p. 117-132.

5. Padias, A.B., et al., Starburst Poly(ether) Dendrimers. Journal of Organic

Chemistry, 1987. 52 : p. 5305-5312.

6. Hall, H.K. and Polis, D.W., Starburst Poly(aryl amine)s and Their Semi-

Conducting Complexes as Potentially Elecroactive Materials. Polymer Bull.,

1987. 17 : p. 409-416.

7. Uhrich, K.E. and Frechet, J.M.J., Synthesis of Dendritic via a

Convergent Growth Approach. J. Chem. Soc. Perkin Trans. 1, 1992: p. 1623-

1630.

8. Hawker, C.J. and Frechet, J.M.J., A New Convergent Approach to Monodisperse

Dendritic Macromolecules. J. Chem. Soc., Chem. Commun., 1990. 19 : p. 1010-

1013.

48

9. Hawker, C.J. and Frechet, J.M.J., Preparation of Polymers with Controlled

Architecture - A New Convergent Approach to Dendritic Macromolecules.

Journal of the American Chemical Society, 1990. 112 : p. 7638-7647.

10. Wooley, K.L., Hawker, C.J., and Frechet, J.M.J., Hyperbranched

Macromolecules Via a Novel Double-Stage Convergent Growth Approach.

Journal of the American Chemical Society, 1991. 113 : p. 4252-4261.

11. Hawker, C.J. and Frechet, J.M.J., Control of Surface Functionality in the

Synthesis of Dendritic Macromolecules using the Convergent Growth Approach.

Macromolecules, 1990. 23 : p. 4726-4729.

12. Wooley, K.L., Hawker, C.J., and Frechet, J.M.J., Polymers with Controlled

Molecular Architecture - Control of Surface Functionality in the Synthesis of

Dendritic Hyperbranched Macromolecules using the Convergent Approach. . J.

Chem. Soc. Perkin Trans. 1, 1991: p. 1059-1076.

13. Brauge, L., et al., First Divergent Strategy Using Two AB2 Unprotected

Monomers for the Rapid Synthesis of Dendrimers. Journal of the American

Chemical Society, 2001. 123 : p. 6698-6699.

14. Supramolecular Polymers . Second ed, ed. Ciferri, A.. 2005, Boca Raton: CRC

Press.

15. Wooley, K.L., et al., Physical Properties of Dendritic Macromolecules: A study of

Glass Transition Temperature. Macromolecules, 1993. 26 : p. 1514-1519.

16. Lin, Q., et al., Highly Branched Poly(arylene ether)s via Oligomeric A2 + B3

Strategies. Macromolecular Chemistry and Physics, 2006. 207 : p. 576-586.

49

17. Czupik, M. and Fossum, E., Manipulation of the Molecular Weight and

Branching Structure of Hyperbranched Poly(arylene ether phosphine oxide)s

Prepared via an A2 + B2 Approach. Journal of Polymer Science: Part A: Polymer

Chemistry, 2003. 41 p. 3871-3881.

18. Baek, J.-B. and Harris, F.W., Hyperbranched Polyphenylquinoxalines from Self-

Polymerizable AB(2) and A(2)B Monomers. Macromolecules, 2005. 38 : p. 297-

306.

19. Himmelberg, P. and Fossum, E., Development of an Efficient Route to

Hyperbranched Poly(arlene ether sulfone)s. Journal of Polymer Science: Part A:

Polymer Chemistry, 2005. 43 : p. 3178-3187.

20. Sennet, L., Fossum, E., and Tan, L.S., Branched poly(arylene ether ketone)s with

tailored thermal properties: Effects of AB/AB2 ratio, core (B3) percentage, and

reaction temperature. Polymer, 2008. 49 : p. 3731-3736.

21. Markoski, L.J., Thompson, J.L., and Moore, J.S., Synthesis and Characterization

of Linear-Dendritic Aromatic Etherimide : Tuning Molecular

Architecture to Optimize Properties and Processability. Macromolecules, 2000.

33 (15): p. 5315-5317.

22. Wang, K.L., Jikei, M., and Kakimoto, M.A., Synthesis of Soluble Branched

Polyimides Derived from an ABB ' Monomer. Journal of Polymer Science Part a-

Polymer Chemistry, 2004. 42 (13): p. 3200-3211.

23. Lin, Q. and Long, T.E., Polymerization of A 2 with B 3 Monomers: A Facile

Approach to Hyperbranched Poly(aryl ester)s. Macromolecules, 2003. 36 : p.

9809-9816.

50

24. Kricheldorf, H.R., Vakhtangishvili, L., and Fritsch, D., Synthesis and

Functionalization of Poly(ether sulfone)s Based on 1,1,1-Tri(4-

hydroxyphenyl)ethane. Journal of Polymer Science: Part A: Polymer Chemistry,

2002. 40 : p. 2967-2978.

25. Hao, J., Jikei, M., and Kakimoto, M.A., Preparation of Hyperbranched Aromatic

Polyimides via A 2 + B 3 Approach. Macromolecules, 2002. 35 : p. 5372-5381.

26. Schmaljohann, D. and Voit, B., Kinetic Evaluation of Hyperbranched A 2 + B 3

Polycondensation Reactions. Macromolecular Theory and Simulations, 2003.

12 (9): p. 679-689.

27. Flory, P.J., Molecular Size Distribution in Three Dimensional Polymers I.

Gelation. Journal of the American Chemical Society, 1941. 63 : p. 3083 - 3090.

28. Kricheldorf, H.R., The Role of Self-Dilution in Step-Growth Polymerizations.

Macromolecular Rapid Communications, 2008. 29 : p. 1695-1704.

29. Concise Polymeric Materials Encyclopedia , J.C. Salamon, Editor. 1999, CRC

Press LLC: Boca Raton.

30. Carraher, Charles Jr., Polymer Chemistry 5th ed. 2000, New York: Marcel

Dekker, Inc.

31. Margolis, J.M., Engineering Handbook . 2006, McGraw-Hill: New York.

32. Kricheldorf, H.R., et al., Multicyclic Poly(ether sulfone)s of Phloroglucinol

Forming Branched and Cross-Linked Architectures. Macromolecules, 2003. 36 :

p. 4337-4344.

33. Kricheldorf, H.R., L. Vakhtangishvili, and D. Fritsch, Synthesis and

Functionalization of Poly(ether sulfone)s based on 1,1,1-Tris(4-

51

hydroxyphenyl)ethane. Journal of Polymer Science: Part A: Polymer Chemistry,

2002. 40 : p. 2967-2978.

34. Unal, S. and Long, T.E., Highly Branched Poly(ether ester)s via Cyclization-Free

Melt Condensation of A2 Oligomers and B3 monomers. Macromolecules, 2006.

39 : p. 2788-2793.

35. Unal, S., et al., A New Generation of Highly Branched Polymers: Hyperbranched,

Segmented Poly(urethane urea) . Macromolecules, 2004. 37 : p. 7081-

7084.

36. Carter, K.R., Aryl Fluoride Monomers in Nucleophilic Aromatic Substitution

: Evaluation of Monomer Reactivity by I9F NMR

Spectroscopy. Macromolecules, 1995. 28 : p. 6462-6470.

37. Bernal, D.P., et al., Fluoride-Terminated Hyperbranched Poly(arylene ether

phospine oxide)s via Nucleophilic Aromatic Substitution. Journal of Polymer

Science Part A-Polymer Chemistry, 2002. 40 : p. 1456-1467.

52