The Curing and Degradation Kinetics of Sulfur Cured EPDM Rubber A

Home , Curing (chemistry), EPDM rubber, Vulcanization

The Curing and Degradation Kinetics of Sulfur Cured EPDM Rubber

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

ROBERT J. WEHRLE B.S., Northern Kentucky University, 2012

Wright State University 2014 WRIGHT STATE UNIVERSITY

GRADUATE SCHOOL

August 29, 2014

I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY Robert Joseph Wehrle ENTITLED The Curing and Degradation Kinetics of Sulfur Cured EPDM Rubber BE ACCEPTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE.

Eric Fossum Ph.D. Thesis Director

David A. Grossie, Ph.D. Chair, Department of Chemistry

Committee on Final Examination

Eric Fossum, Ph.D.

William A. Feld, Ph.D.

Steven B. Glancy, Ph.D.

Kenneth Turnbull, Ph.D.

Robert E. W. Fyffe, Ph.D. Vice President for Research and Dean of the Graduate School Abstract

Wehrle, Robert J. M.S, Department of Chemistry, Wright State University, 2014. The Curing and Degradation Kinetics of Sulfur Cured EPDM Rubber.

Ethylene‐propylene‐diene (EPDM) rubbers containing varying amounts of diene were cured with sulfur using either a moving die rheometer (MDR) or a rubber process analyzer (RPA). The effect of removing curatives and how the curing reaction changed was explored. Kinetic data was extracted from the rheology plots and reaction rate constants were determined by two separate ways: manually choosing points of interest or by a computer model.

iii

TABLE OF CONTENTS

Page 1. Introduction 1 1.1 EPDM Overview 1

1.2 Preparation of EPDM 2

1.2.1 Ziegler‐Natta Catalysts 2

1.2.2 Metallocene Catalysts 4

1.3 Cross‐link Chemistry 5

1.3.1 Peroxide Cure 5

1.3.2 Sulfur Cure 6

1.3.3 Cross‐link Sites 8

1.3.3.1 Polymer Branching 9

1.3.4 Rubber Ingredients 10

1.3.4.1 Non‐curative Ingredients 10

1.3.4.2 Curative Ingredients 11

1.4 Kinetics 12

1.5 Instrumentation 13

2. Experimental 15 2.1 Materials 15

2.2 Instrumentation 16

2.3 Rubber Mixing 16

Table of Contents (continued)

Page 2.4 Rheology Testing 17

2.5 Curve Fitting 17

3. Results and Discussion 18 3.1 Ziegler‐Natta vs. Metallocene Catalysts 18

3.2 ENB Incorporation 20

3.3 Filler‐less Formulations 21

3.4 Curative Effects 23

3.5 Degradation Effects and Factors 32

3.6 Curve Fitting to Predict Reaction Rates 42

4. Conclusion 48

5. References 48

LIST OF FIGURES

Page Figure 1: A typical EPDM polymer………………………………………………………………………………….2

Figure 2: A typical metallocene catalyst…………………………………………………………………………4

Figure 3: An example of four entangled polymer chains………………………………………………..5

Figure 4: A peroxide cross‐link……………………………………………………………………………………….6

Figure 5: A mature sulfur cross‐link ……………………………………………………………………………….6

Figure 6: Three main dienes used in EPDM. A: DCPD B: VNB C: ENB…………………………..8

Figure 7: The ways the dienes can be incorporated into the polymer. A: Through the norbornene ring B: Outside the norbornene ring………………………………………………………….8

Figure 8: An example of LCB…………………………………………………………………………………………..9

Figure 9: Di(p‐octylphenyl)amine …………………………………………………………………………………10

Figure 10: A: MBT B:TMTD………………………………………………………………………………………….11

Figure 11: A: ZDBDC B: TDEDC…………………………………………………………………………………….12

Figure 12: A diagram of how the moving die rheometer and rubber process analyzer works…...………………………………………………………………………………………………………………………13

Figure 13: A torque vs. Time plot. Area A: induction zone Area B: cure zone Area C: overcure zone Line 1: marching Line 2: plateau Line 3: reversion/degradation……….15

Figure 14: Comparison of the metallocene and Ziegler‐Natta rubbers using MH‐ML vs.

ENB content...... …………………………………………………………………………………………………………19

LIST OF FIGURES (continued)

Page Figure 15: Alkene region of the 300 MHz 1H NMR spectra of A: EPDM‐5, B: EPDM‐1, C:

EPDM‐3. The peaks at 5.44 ppm and 5.18 ppm correspond to the external and internal pathways, respectively……...…………………………………………………………………………………………21

Figure 16: Comparison of MDR (or RPA) data for identical formulations with and without

carbon black...………………………………………………………………………………………………………………23

Figure 17: The curing of VL1710Z1 (1.8% ENB) at different temperatures…………………….25

Figure 18: RPA traces of VL1710Z1 and VL1710ZA1 (no MBT) at 170°C...... …………………..26

Figure 19: RPA traces of VL1710Z1 and VL1710ZB1 (no TMTD) at 170°C...……………………27

Figure 20: RPA traces of VL1710Z1 and VL1710ZC1 (no ZDBDC) at 170°C …………………….28

Figure 21: RPA traces of VL1710Z1 and VL1710ZD1 (no TDEDC) at 170°C...... ……………….29

Figure 22: Derivative plots of VL1710ZA1‐VL1710ZD1 at 170°C. VL1710ZA1 (no MBT),

VL1710ZB1 (no TMTD), VL1710ZC1 (no ZDBDC), and VL1710ZD1 (no TDEDC)………………30

Figure 23: RPA traces of VL1710Z1 and VL1710ZE1 (no sulfur) at 170°C……………………….31

Figure 24: VL1710ZE1 (no sulfur) at different temperatures...………………………………………31

Figure 25: Curing of VL1710ZI2 at 200°C for 20 minutes……………………………………………….34

Figure 26: TGA traces of (blue) raw polymer (EPDM‐5), (green) cured and (red) uncured samples of VL1710Z6. Samples were run under an air atmosphere …………………………….35

Figure 27: DCS traces of raw polymer (EPDM‐5) (blue), cured (green) and uncured (red)

samples of VL1710Z6...... ………………………………………………………………………………………………36

vii

LIST OF FIGURES (continued)

Page Figure 28: An example of how the intersect points were made. The slope of the curing process was taken around the highest rate and the slope of the degradation was taken from where the MH was reached until the end…………………………………………………………….37

Figure 29: Torque vs. temperature of the intersect points of the various formulations at different temperatures…………………………………………………………………………………………………38

Figure 30: The Arrhenius plots for the curing process of VL1710Z1, VL1710Z3, VL1710Z5 and VL1710Z6 ………………………………………………………………………………………………………………39

Figure 31: Arrhenius plots for the degradation process of VL1710Z1, VL1710Z3,

VL1710Z5 and VL1710Z6.……………………………………………………………………………………………..39

Figure 32: Arrhenius plots for the ratio of curing vs. degradation of VL1710Z1, VL1710Z3,

VL1710Z5, and VL1710Z6……………………………………………………………………………………………..40

Figure 33: Activation energy required for both processes for varying ENB content………41

Figure 34: Fitted curve data for VL1710Z1 at various temperatures. Blue: RPA traces

Red: Theoretical data obtained from curve fitting software Green: The absolute difference between the theoretical and experimental curves………………………………………42

Figure 35: Fitted curve data for VL1710Z3 at various temperatures. Blue: RPA traces

Red: Theoretical data obtained from curve fitting software Green: The absolute difference between the theoretical and experimental curves………………………………………43

viii

LIST OF FIGURES (continued)

Page Figure 36: Fitted curve data for VL1710Z5 at various temperatures. Blue: RPA traces

Red: Theoretical data obtained from curve fitting software Green: The absolute difference between the theoretical and experimental curves...... …………………………………44

Figure 37: Fitted curve data for VL1710Z6 at various temperatures. Blue: RPA traces

Red: Theoretical data obtained from curve fitting software Green: The absolute

difference between the theoretical and experimental curves………………………………………45

Figure 38: Arrhenius plot of curing for all four formulations………………………………………..46

Figure 39: Arrhenius plot of degradation for all four formulations...... …...……………………46

Figure 40: Activations energy required for both process...... ………………………………………..47

LIST OF TABLES

Page Table 1: The standard minimal EPDM formulation……………………………………………………….17

Table 2: All the formulations in this set include the same ingredients except for the polymer used ………………………………………………………………………………………………………………18

Table 3: Polymers used in each formulation and their ENB content...………………………….19

Table 4: The filler‐less formulation…..…...... ……...…………………………………………………………22

Table 5: Basic formulations for curative studies…………………………………………………………..24

Table 6: Basic formulations for degradation studies…………………………………………………….32

LIST OF EQUATIONS

Page Equation 1: General kinetic model A: initial polymer B: cross‐linked rubber C: degradation product…………………………………………………………………………………………………….12

Equation 2: The rate equation for B…………………………………………………………………………….13

Equation 3: Kinetic equation used to describe curing behavior……...……………...…………..13

LIST OF SCHEMES

Page Scheme 1: The catalytic process of α‐olefin polymization via a Ziegler‐Natta catalyst system……………………………………………………………………………………………………………………………3

Scheme 2: General peroxide curing mechanism ……………………………………………………………6

Scheme 3: General sulfur curing mechanisms………………………………………………………………..7

xii

1 Introduction

One of the most widely used elastomers is ethylene propylene diene monomer rubber or EPDM rubber 1. In order to decrease the cure time for EPDM rubber the cure temperature can be increased, however, an increase in temperature can also lead to degradation. This work will focus on studying a kinetic relationship between the sulfur curing of EPDM rubber as well as a degradation process that occurs simultaneously. In addition, experiments will be designed that will allow for the development of a basic understanding of how the different curatives affect the two processes.

1.1 EPDM Overview

Rubber has been an important material in consumer and industrial settings ever since Charles Goodyear first vulcanized natural rubber in 1844 2. Since that time many different rubber types were created for different purposes: styrene‐butadiene, isobutylene‐isoprene, acrylonitrile butadiene, fluoroelastomer, etc. Ethylene propylene diene monomer (EPDM) rubber was first introduced in the United States in 1962 and since has been widely used 1, having an extensive repertoire of applications.

Propylene Ethylene

Diene

Figure 1: A typical EPDM polymer.

These applications range from automotive seals and gaskets to roofing and sheeting 3.

EPDM excels in these fields because its polymeric backbone is completely saturated

(Error! Reference source not found.), which in turn gives added resistance to oxygen,

ozone, electricity, heat, and UV 4. Due to EPDM having a hydrocarbon structure, it can be used in the presence of polar solvents, but is attacked by nonpolar solvents, such as toluene.

1.2 Preparation of EPDM Polymers

EPDM polymers are typically synthesized by a chain growth process utilizing a transition metal catalyst1. The type of catalyst used can have a dramatic impact on the properties of the polymer formed, such as molecular weight and tacticity. There are two main classes of catalysts that dominate the production of EPDM polymers, Ziegler‐

Natta and metallocene catalysts.

1.2.1 Ziegler‐Natta Catalysts

In 1953 Karl Ziegler came up with a process to polymerize ethylene. A year later

Giulio Natta discovered that this process can be used to make isotactic polypropylene and other olefin polymers such as EPDM 1. Much of the EPDM today is produced by use of a heterogeneous Ziegler‐Natta catalyst. This is done by polymerizing the α‐olefin monomers in a low boiling hydrocarbon diluent with titanium tetrachloride and triethylaluminum as the catalyst system. As purposed by Cossee 5, chain growth

happens at the activated titanium species by the olefin associating its π‐electrons with an open orbital of titanium (Scheme 1). The double bond of the ethylene then breaks forming bonds with titanium and the growing polymer chain. Next migration happens reopening the active site of titanium.

Scheme 1. The catalytic process of α‐olefin polymization via a Ziegler‐Natta catalyst system

Cl CH3 Cl Et Al CH3 Et Al CH2 CH2 Cl CH2 CH2 Ti Cl CH2 Cl Cl Ti CH2 Cl Cl Cl Cl Activated catalyst Ethylene with an open orbital Ethylenes pi electrons complex with titanium's open orbital

CH3

CH2 Cl CH3 Cl H2C Et Al C CH2 H Et Al CH2 2

Cl CH2 Cl Ti Ti Cl Cl Cl Cl Cl Cl The electrons rearrange in the system Rearrangement of the growing chain due to breaking of the pi bond reopens the catalytic site for additional growth

1.2.2 Metallocene Catalysts

Metallocene catalysts, developed after the creation of Ziegler‐Natta catalysts,

show higher catalytic activity than their conventional

predecessors, which is believed to be due to the homogenous

nature of the metallocene catalyst 6. The catalysts utilize group

IV transition metals (Ti, Zr, and Hf) as seen in Figure 2, which

Figure 2. A typical are then activated, typically by methylalumoxane. Depending metallocene catalyst on the steric influences placed on the system from the aromatic ligands, the tacticity can be controlled 6, 7. Since the aromatic ligands take up space, large bulky groups limit the ability of prochiral monomers to orient themselves in the catalytic site; this decrease in orientation gives rise to polymers being constructed in a regular manner, which leads to isotactic polymer chains. If there is less hindrance at the catalytic site, there is more space for the prochiral monomers to enter which leads

to variety of orientations. When there are multiple orientations of prochiral monomers then the polymer chain exhibits atacticity. These different tacticities can affect the physical properties of the polymer, such as the glass transition temperature, crystallinity, and solubility 8.

are then activated, typically by methylalumoxane. Depending on the steric influences

placed on the system from the aromatic ligands, the tacticity can be controlled 6, 7.

Since the aromatic ligands take up space, large bulky groups limit the ability of prochiral monomers to orient themselves in the catalytic site; this decrease in orientation gives rise to polymers being constructed in a regular manner, which leads to isotactic polymer chains. If there is less hindrance at the catalytic site, there is more space for the prochiral monomers to enter which leads to variety of orientations. When there are multiple orientations of prochiral monomers then the polymer chain exhibits atacticity.

These different tacticities can affect the physical properties of the polymer, such as the glass transition temperature, crystallinity, and solubility 8.

1.3 Cross‐link Chemistry

Without being cross‐linked EPDM by itself has limited mechanical properties.

to form polymeric radicals. From this point two polymer radicals combine to yield a carbon‐carbon bond, the cross‐link 4 H2 H2 (Scheme 2). Peroxide cured rubbers are used for high C C temperature applications, in part due to the strong carbon‐ carbon cross‐links (Figure 4). Figure 1: A peroxide cured cross‐link

Scheme 1: General peroxide curing mechanism

RO OR 2RO

RO + ROH +

1.3.2 Sulfur Cure

Since the discovery of vulcanization, sulfur‐curing

systems have been widely used to cure rubber. Unlike peroxide

curing systems, however, sulfur curing systems require a source

of unsaturation in the polymer. Curing by use of sulfur alone is

a slow process, so curatives were developed to aid in curing, as

well as to tailor specific properties of the final product. In sulfur Figure 2. A mature sulfur cross‐link curing systems, sulfur is first typically activated by zinc oxide

6 and an accelerator. From this point it is accepted that the now formed sulfurating agent abstracts an allylic hydrogen and attacks where the double bond once was on the polymer forming a cross‐link precursor 4, 9. The cross‐link precursor then attacks another

polymer forming an initial polysulfidic cross‐link. The final matured cross‐link, as seen in

Figure 5, is formed when the sulfur content between the polymers has been reduced by desulfuration (Scheme 3).

Scheme 3. General sulfur curing mechanisms

Accelerator Sx Zn Sy Accelerator + + HSy Accelerator + Zn S

Sx Accelerator

+ S Sx 1-3

Sx Accelerator

The presence of accelerators, activators, etc. also gives rise to competing side reactions that lead to degradation during the curing process, which reduces the final mechanical properties of the rubber 9, 10. When compared to peroxide curing, the

cross‐links that are formed are weaker, so applications that avoid high temperatures are

utilized, such as automotive belts and electrical wire insulation. With many reactions occurring at once, each with its own influence on the curing process, a mechanistic understanding what is happening remains difficult.

1.3.3 Cross‐linking Sites

With the presence of unsaturation in the diene monomer EPDM can be cured using a sulfur system. One of three dienes are typically incorporated in EPDM: A) dicyclopentadiene (DCPD), B) 5‐vinylidene‐2‐norbornene

(VNB), or C) 5‐ethylidene‐2‐norbornene (ENB) as seen in Figure 6. The three main dienes used in EPDM. A: Figure 6. DCPD B:VNB C:ENB

Due to it being four times more reactive than ENB, VNB has had limited use in EPDM’s

11. Because the double bond is in the terminal position with VNB, the catalyst system used to make the polymer can readily access this source of unsaturation and cause undesirable branching or premature

ENB VNB DCPD cross‐linking, which can lead to the polymer forming a gel and therefore A becoming unprocessable. If VNB is

incorporated into EPDM it is at low amounts, which then lead to B unsatisfactory cross‐link density and less than desired mechanical properties 11. Figure 7. The ways the dienes can be Fortunately, ENB does not experience incorporated into the polymer. A: Through the norbornene ring B: Outside the these problems. Since the unsaturation norbornene ring

of ENB is more hindered, there is less of a chance for premature cross‐links to form.

However, since there are two initial points of unsaturation present in the dienes, it is possible that they can be integrated into the polymer at either the external or internal double bond (Figure 7). Different modes of incorporation can have a direct effect on branching in the polymer backbone leading to changes in glass transition temperature, crystallinity, and solubility.

1.3.3.1 Polymer Branching

A structure that EPDM polymers can include is long Polymeric backbone chain branching (LCB). LCB happens as a result of the diene substituent having a free unit of unsaturation to which a catalyst system can add 11. Due to this possibility, the LCB section branches can reach a certain length enabling them to Figure 8. An example of LCB become entangled. With this added entanglement, it is possible for lower diene containing polymers to act more like higher diene containing polymers 12. A problem that may arise from LCB is that branching may become too widespread and lead to gelation, although with proper tailoring of the catalyst system gelation will not occur. LCB in EPDM can lead to microstructures that would otherwise not be possible with linear EPDM, such as a comb type structure.

1.3.4 Rubber Ingredients

There are a number of ingredients that go into making a rubber, and each has its own role. Despite the overwhelming number of items a rubber formulation can include, a general knowledge of the purpose of each ingredient can provide some insight as to the final properties of the cured rubber.

1.3.4.1 Non‐curative Ingredients

Non‐curatives are ingredients that, as their name implies, will not participate directly in the curing of EPDM rubber. These ingredients include the processing aides, antioxidants, and fillers. Processing aides are petroleum based oils that are added to help when the polymer is cured inside a mold. The oils prevent the rubber from sticking

H N to the mold surface, which could otherwise cause

defects in the product or hinder its removal from the C8H17 C8H17 Figure 9. Di(p‐ octylphenyl)amine mold. The processing aide that was used for this study was a paraffinic, petroleum oil. Antioxidants are compounds that protect the final

rubber product from degradation caused by heat, oxygen, ozone, UV radiation, etc. The antioxidants act by scavenging the radicals that are caused by the sources of degradation. Di(p‐octylphenyl)amine (Figure 9), the antioxidant used in this study, makes use of the nitrogen present in its structure as a radical scavenger which forms NO

compounds. Fillers, carbon black in the case of this study, is beneficial to rubber manufacturing in that it helps reinforce the final rubber product making it more durable

and giving higher mechanical properties. However, the addition of carbon black increases the viscosity of the rubber during the mixing process, causing shear heating.

This shear heating can have undesirable effects such as premature cross‐linking or, as will be discussed later, increased degradation.

1.3.4.2 Curative Ingredients

The curatives are one of the more critical ingredients of rubber manufacturing; they are what allow rubber to cure. For sulfur cured systems, these ingredients include

accelerators, activators, and a sulfur source. Accelerators N A SH S were developed to speed up the sulfur curing process and act

S by forming an adduct, with sulfur making it a more active B S N N S species. 2‐Mercaptobenzothiazole, MBT, the primary S Figure 10. A: MBT B: accelerator in this study, is believed to be the main TMTD sulfurating agent in the curing process (Figure 10A). With the addition of zinc oxide,

MBT is able to form a complex that is actively able to integrate elemental sulfur in this complex forming an active sulfurating species. The process of sulfur incorporation in

MBT, however, is still not yet understood. Tetramethylthiuram disulfide, TMTD, is another accelerator used in this study (Figure 10B), which is thought to act in a similar fashion to MBT, but it is also able to act as a sulfur donor. Efficient vulcanizations, or vulcanizations that do not utilize elemental sulfur, require a large amount of TMTD to form cross‐links by sulfur donation. Like the other accelerators mentioned thus far, zinc

dibutyldithiocarbamate, ZDBDC, and tellurium S S A Bu Zn Bu tetrakis(diethyldithiocarbamate), TDEDC, act as N S S N Bu Bu accelerators and are able to incorporate sulfur to Et Et form sulfurating species (Figure 11). N S Et S S N B Et S Te S Et 1.4 Kinetics N S S Et S N To attempt to understand the curing process Et Et Figure 11. A: ZDBDC B: TDEDC of EPDM rubber, which determines the final properties given to the rubber after cure, a kinetic approach is necessary. A kinetic

equation that will accurately describe how a rubber will cure is possible. Due to the complicated nature of how each ingredient, polymer orientation and structure interacts and contributes to the cure and degradation rates, a broad definition of kinetic behavior is preferred. Since this study utilizes isothermal conditions, first order kinetics can be applied 13. That being said, there are only three species that need to be considered: the initial polymer, the cross‐linked rubber, and the degradation product (Equation 1).

Equation 1. General kinetic model. A: initial polymer

B: cross‐linked rubber C: degradation product

௞భ ௞మ ܣ ՜ܤ՜ܥ

By turning Equation 1 into a rate equation with interest in the creation of B forms

Equation 2.

Equation 2. The rate equation for B ݀ሾܤሿ ൌ݇ ሾܣሿ ݁ି௞భ௧ െ݇ ሾܤሿ ݐ ଵ ଴ ଶ݀

With Equation 2 being integrated, Equation 3 is formed.

Equation 3. Kinetic equation used to describe curing behavior

݇ଵ ି௞మ௧ ି௞భ௧ െ݁ ݁כ ሿ଴ܣሿ௧ ൌ ሾܤሾ ݇ଵ െ݇ଶ

In Equation 3, k1 and k2 can be solved to describe how and why curing is behaving the way it is as well as determining how much degradation is taking place.

1.5 Instrumentation

For this study, a moving die rheometer and a rubber process analyzer were used to better understand what is happening during the curing process. They Figure 12. A diagram of how the operate by applying pressure to an uncured moving die rheometer and rubber process analyzer works rubber sample between two heated dies. The bottom die is motorized and moves with a specific arc angle and frequency while the top die measures the torque, or rotational force, that the rubber sample transmits from the bottom die (Figure 12). As the rubber sample heats up, the curing process is

initiated and cross‐links begin to form. As cross‐links form the rubber sample becomes more rigid and the torque increases. The data typically analyzed from these instruments are torque vs. time plots.

From the generated torque vs. time plots the maximum torque (MH), minimum torque (ML), curing rates, and degradation rates can all be gathered. As seen in Figure

13, there are three zones of interest: the induction zone, the curing zone, and the overcure zone. In the induction zone, the polymer softens and its viscosity is reduced, which is seen by a drop in torque. Also in this zone is where cross‐link precursors are

made. In the curing zone is where the cross‐links are formed and this area possesses the maximum rate of cross‐linking.

In the overcure zone, the thermal aging behavior of the cured rubber sample can be observed. There are three behaviors that can be seen: marching modulus, plateau modulus, and reversion/degradation modulus. In EPDM, a marching modulus can be seen when the sample is inadequately cured, which can be remedied by increasing the cure temperature. Reversion happens due to degradation and side reactions that occur in the rubber system. A plateau modulus is the desired outcome due to curing being

complete and an absence of degradation 14.

25 1 A B 2 20

15 3 (in*lbs)

10 C Torque 5

0 0246810 Time

Figure 13. A torque vs. time plot. Area A: induction zone Area B: cure zone Area C: overcure zone Line 1: marching Line 2: plateau Line 3: reversion/degradation

2 Experimental

2.1 Materials

EPDM polymers 1, 2, 5, and 6 were purchased from Exxon Mobil. EPDM polymers 3 and 4 were purchased from Dow Chemical. EPDM polymer 7 was purchased from Kumho Polychem. EPDM polymer 8 was purchased from Lion Copolymer Geismar.

Carbon Black was purchased from Colombian Chemicals Company. Paraffinic oil was purchased from Holly Refining and Marketing. Zinc oxide was purchased from

Horsehead Corporation. 90% Stearic Acid was purchased from Acme Hardesty. Di(p‐ octylphenyl)amine, zinc dibutyldithiocarbamate (ZDBDC), tellurium diethyldithiocarbamate (TDEDC), and Dipentamethylene thiuram tetrasulfide (DPTT) were purchased from R.T. Vanderbilt Company. Sulfur was purchased from S.F. Sulfur

Corporation. 2‐Mercaptobenzothiazole (MBT) and tetramethyl thiuram disulfide

(TMTD) were purchased from Harwick Standard Distribution Corporation.

2.2 Instrumentation

A Monsato Rheometer (MDR 2000E) and Alpha Technology Rheometer (RPA

2000) were used to collect the rheology data. These data were collected at intervals of

10‐20 minutes at temperatures ranging from 170‐200°C. All NMR data were obtained utilizing a Bruker AVANCE 300 MHz NMR spectrometer using toluene‐d8 as the solvent.

All DSC data were acquired using a TA Instruments Auto Q20 DSC; all samples were run isothermally at 200°C for 20 minutes. All TGA data were acquired using a TA instruments Q50 TGA; all samples were run isothermally at 200°C for 30 minutes.

2.3 Rubber Mixing

All rubber formulation variations were based on the standard formulation listed

in Table 1. The ingredients were added in concentrations of parts per hundred with respect to the polymer (PHR). A two‐pass process was used to make the test sheet utilized for sampling. The two‐pass process consisted of adding the non‐curing

ingredients in a Banbury mixer to form a master batch. This master batch was then removed once it reached 125°C. If the master batch did not contain filler, it was then added to a Bolling 10x20 mill and the curatives were added, otherwise the master batch

was reintroduced to the Banbury mixer at which point the curatives were added. The test slab was considered finished once a temperature of 100°C was reached.

Table 1. The standard minimal EPDM formulation

Ingredient PHR Polymer 100 Carbon Black 45 Paraffinic oil 10 Zinc Oxide 5 Stearic Acid 90% 1 Di(p‐octylphenyl)amine 2 Sulfur 1 MBT 2 TMTD 0.8 ZDBDC 0.8 TDEDC 0.8

2.4 Rheology Testing

ASTM D2084‐11 ‐ Standard Test Method for Rubber Property—Vulcanization

Using Oscillating Disk Cure Meter was used in collecting the rheological data. The

Moving Die Rheometer (MDR) and Rubber Process Analyzer (RPA) were calibrated daily with a standardized rubber sample with a weight 4.5 times its specific gravity in

accordance to the ASTM method. Testing consisted of isothermal curing for 10‐20 minutes.

2.5 Curve Fitting

The curve fitting software within Microsoft Excel was utilized for data analysis.

Tangential estimates, forward derivatives, and a Newtonian method of optimization were employed with a precision of 1x10‐6, a 5% tolerance, and a convergence of 1x10‐4.

3 Results and Discussion

3.1 Ziegler‐Natta vs. Metallocene Catalysts

The initial focus of the project was to determine how changing the ethylidene norbornene (ENB) level in the rubber formulation would affect the kinetics of the curing reaction. The formulations of these sets of rubbers, as seen in Table 2 and Table 3,

were simplified in order to limit the number of variables.

Table 2. All the formulations in this set include the same ingredients except for

the polymer used.

Ingredient PHR Polymer 100 Carbon Black 56.26 Paraffinic oil 12.5 Zinc Oxide 6.25 Stearic Acid 90% 1.25 Di(p‐octylphenyl)amine 2.5 Sulfur 1.25 MBT 2.5 TMTD 1 ZDBDC 1 TDEDC 1

Table 3. Polymers used in each formulation and their ENB content.

Formulation Polymer ENB content Catalyst VL1719Z221 EPDM‐1 4.50% Metallocene VL1719Z221A EPDM‐2 9.20% Ziegler‐Natta VL1719Z221B EPDM‐3 4.90% Metallocene VL1719Z221C EPDM‐4 1.80% Metallocene VL1719Z221D EPDM‐5 7.50% Metallocene VL1719Z221E EPDM‐6 3.80% Ziegler‐Natta VL1719Z221F EPDM‐7 7.90% Ziegler‐Natta VL1719Z221G EPDM‐8 9.40% Ziegler‐Natta

These rubbers were cured at 170, 180, 190, or 200°C by use of the MDR. As can be seen in Figure 14, comparing the difference of the maximum torque and the minimum torque

(MH‐ML) vs ENB content gave uncorrelated data when viewed as a whole.

MH‐ML vs. ENB Content Between Differing Catalysts Systems 20

16 (in*lbs)

Metallocene ML

‐ 14 Ziegler‐Natta MH 12

10 0246810 ENB Percentage

Figure 14. Comparison of the metallocene and Ziegler‐Natta rubbers using MH‐ML vs. ENB content.

The data indicate that there might be two different classes of materials, which is

not unexpected due to the polymers being prepared using either a metallocene or

Ziegler‐Natta catalyst system. Metallocene catalysts lead to polymers with a narrow polydispersity while the Ziegler‐Natta systems result in a much broader distribution of chain sizes. Since the Ziegler‐Natta polymers gave more sporadic results while the metallocene polymers gave a more linear trend, further work was carried out with the polymers prepared with the metallocene catalysts.

3.2 ENB Incorporation

Due to the nature of the catalyst used, the ENB can be integrated into the

polymers in two different pathways, through the ring double bond or via the external double bond. The different incorporations of ENB can have an effect on the properties of curing due to spatial interactions. To determine if there was any variation in the ENB incorporation 1H NMR spectroscopy was employed. The 1H NMR spectrum of ENB from

Liu et al. 15 shows that the ethylidene proton present at 5.2 and 5.4 ppm are from the

E/Z isomers and the internal ring alkene protons are at 6.1 ppm. Three of the four metallocene polymers were analyzed and it was found that there was an absence of the internal ring alkene protons (Figure 15). Based on this information the metallocene catalyst only incorporated the ENB by the internal double bond and not by the ethylidene double bond. Since there is no variation on how the ENB was incorporated

into the polymer, it can be concluded that the metallocene catalyst system used to create these polymers was discriminate with regards ENB orientation.

Figure 15. Alkene region of the 300 MHz 1H NMR spectra of A: EPDM‐1, B: EPDM‐3, C: EPDM‐5.

3.3 Filler‐less Formulations

With carbon black present in the formulations, an increase in degradation can be seen when compared to filler‐less formulations (Figure 16). This increase in degradation is due to shear heating that is introduced into the system due to the added viscosity from carbon black. New formulations were created that excluded carbon black and since di(p‐octylphenyl)amine does nothing to contribute to the curing process, it was also removed (Table 4). The new formulations were then compared to their previous formulation counterparts. The formulations without filler exhibited a 50.7% decrease

in MH, however, a 23.6% reduction in degradation at higher temperatures was observed when compared to the first formulation sets (Figure 16). The decreased MH can be attributed to the lack of reinforcement to the polymer network in the absence of carbon black.

Table 4. The filler‐less formulations

Ingredients PHR Formulation Polymer Polymer 100 VL1710Z1 EPDM‐4 Paraffinic oil 5 VL1710Z3 EPDM‐3 Zinc Oxide 5 VL1710Z5 EPDM‐1 Stearic Acid 90% 1 VL1710Z6 EPDM‐5 Sulfur 1 MBT 2 TMTD 0.8 ZDBDC 0.8 TDEDC 0.8

EPDM‐1 Formulations With and Without Carbon Black at 200°C

16 14 12 10 VL1710Z221

(in*lbs) (with filler) 8 6 VL1710Z5 Torque 4 (without filler) 2 0 0246810 Time (min)

Figure 16. Comparison of MDR (or RPA) data for identical formulations with and without carbon black.

3.4 Curative Effects

In order to study the effects of the various curatives on the curing process formulations based on VL1710Z1 (1.8% ENB) were developed as listed in Table 5.

These five new formulations were VL1710Z1 except that with each one, a different curative was removed. These formulations were then cured at 170, 180, 190, and

200°C, which gave an indication of the effect of each curative on the overall process.

Table 5. Basic formulations for curative studies.

VL1710Z1 VL1710ZA1 VL1710ZB1 VL1710ZC1 VL1710ZD1 VL1710ZE1 Ingredients (Standard) (PHR) (PHR) (PHR) (PHR) (PHR) (PHR) EPDM‐4 100 100 100 100 100 100 Paraffinic oil 5 5 5 5 5 5

Zinc Oxide 5 5 5 5 5 5

Stearic Acid 1 1 1 1 1 1 90% Sulfur 1 1 1 1 1 0 MBT 2 0 2 2 2 2 TMTD 0.8 0.8 0 0.8 0.8 0.8 ZDBDC 0.8 0.8 0.8 0 0.8 0.8

TDEDC 0.8 0.8 0.8 0.8 0 0.8

At low temperatures VL1710Z1 exhibited an unusual curing profile in the form of

a shoulder (Figure 17). This change in the rate of torque increase can be attributed a

change in reaction rates in the cross‐linking process. These different rates may be attributed to one of two reasons: alternative reactions or inhomogeneous mixing. If inhomogeneous mixing is the cause of the rate fluctuations, then it should be absent in other experiments. As depicted in Figure 17, this behavior is observed not only at 170°C, but also at 180°C. The effect is not as visible at higher temperatures. This rate change is

absent at higher ENB concentrations, which suggests that kinetic factors are causing this

phenomenon. Also observed in Figure 17 is that higher temperatures seem to lead to

lower MH values which will be discussed later in the next section. Due to this

phenomenon’s visiblity in this EPDM system and not others is why VL1710Z1 was used to explore how each curative affects the curing process as a whole.

VL1710Z1 at Various Temperatures 12

170 (in*lbs) 6 180

Torque 190 4 200

0 024681012 Time (min)

Figure 17. The curing of VL1710Z1 (1.8% ENB) at different temperatures.

With a baseline now established the curative levels were adjusted, one at a time, in an effort to determine if a particular curative contributes to this observed rate change. These tests were done at 170°C so that rate effects could be easily visualized.

The first curative experiment that was performed was the removal of MBT (VL1710ZA1), which increased the induction time and the rate change was no longer observed (Figure

18). The change in the induction time and lack of a fluctuation in rate suggests that MBT

plays a significant role in the beginning of the curing process and a minimal role later on.

Since the curing process did occur, this indicates that there are compounds within the formulation that can perform a similar role as MBT, but are not as efficient.

VL1710Z1 vs. VL1710ZA1 at 170°C 12

8 (in*lbs) 6 VL1710ZA1

Torque VL1710Z1 4

0 024681012 Time (min)

Figure 18. RPA traces of VL1710Z1 and VL1710ZA1 (no MBT) at 170°C

The removal of TMTD instead of MBT (VL1710BZB1) showed a different effect; the induction time returned to normal, but the later rate was reduced (Figure 19). In addition to these variations the rate fluctuation was also absent. With this, it can be said that TMTD acts later in the curing process. It seems that VL1710ZB1 will eventually

have the same amount of torque as VL1710Z1 which means, just like MBT, other compounds can carry out the same functions as TMTD, just not to the same extent or with the same rate of reaction.

VL1710Z1 vs VL1710ZB1 at 170°C 12

8 (in*lbs) 6 VL1710ZB1

Torque VL1710Z1 4

0 024681012 Time (min)

Figure 19. RPA traces of VL1710Z1 and VL1710ZB1 (no TMTD) at 170°C

With the removal of ZDBDC (VL1710ZC1), it was noticed that, unlike with removal of the other two curatives discussed above, the rate fluctuation was still observed, albeit to a lesser degree (Figure 20). In addition, the initial and final rates

seem to follow the standard formulation quite well. With all of this said, it seems that

ZDBDC does not contribute to the curing process as much as the previous curatives.

Without ZDBDC, curing actually seems to be more efficient as there is less of a change

between the initial and ending curing rate with minimal impact on the maximum torque when comparing to the standard (Figure 20).

VL1710Z1 vs. VL1710ZC1 at 170°C 12

8 (in*lbs) 6 VL1710Z1

Torque VL1710ZC1 4

0 024681012 Time (min)

Figure 20. RPA traces of VL1710Z1 and VL1710ZC1 (no ZDBDC) at 170°C

When TDEDC (VL1710ZD1) was removed the induction time increased slightly, when compared to the standard (Figure 21). Although not as vital as MBT, TDEDC seems to work in the early stages of curing. Looking at the slope of VL1710ZD1 it appears as if there is not a rate fluctuation, but examination of the derivative plot of this

cure system shows that the fluctuation is still present (Figure 22). The initial rate, based on the slope, looks to be about the same, there is just more of an induction time.

VL1710Z1 vs. VL1710ZD1 at 170°C 12

8 (in*lbs) 6 VL1710ZD1

Torque VL1710Z1 4

0 024681012 Time (min)

Figure 21. RPA traces of VL1710Z1 and VL1710ZD1 (no TDEDC) at 170°C

Figure 22. Derivative plots of VL1710ZA1‐VL1710ZD1 at 170°C. VL1710ZA1 (no MBT), VL1710ZB1 (no TMTD), VL1710ZC1 (no ZDBDC), and VL1710ZD1 (no TDEDC).

The last experiment that was performed was a study of how the absence of sulfur affects the curing of the rubber (VL1710ZE1). Without sulfur, virtually no vulcanization occurred (Figure 23). Only at high temperatures does any type of curing become noticeable (Figure 24). When compared to the standard formulation,

VL1710ZE1 experiences a 93% decrease in MH at 170°C and an 86% decrease at 200°C.

The only curing that does take place is due to sulfur donation from ingredients such as

TMTD, but as observed this donation is minimal at best.

VL1710Z1 vs. VL1710ZE1 at 170°C 12

8 (in*lbs) 6 VL1710ZE1 4 Torque VL1710Z1

0 024681012 Time (min)

Figure 23. RPA plot of VL1710Z1 and VL1710ZE1 (no sulfur) at 170°C.

VL1710ZE1 at Various Temperatures 2.5

1.5 170 (in*lbs)

180 1

Torque 190 200 0.5

0 024681012 Time (min)

Figure 24. VL1710ZE1 (no sulfur) at different temperatures.

3.5 Degradation Effects and Factors

In an effort to understand and describe how EPDM cures, consideration of the degradation reactions and how they contribute to the overall process must be taken into account. As elluded to earlier, there is a relationship between temperature and degradation. Determining how the cure reaction is related to the degradation reactions will help determine the optimal conditions under which the rubber can be manufactured. The formulations that were studied did not contain any filler and utilized

EPDM prepared with metallocene catalysts (Table 6).

Table 6. Basic formulations for degradation studies.

VL1710Z1 VL1710Z3 VL1710Z5 VL1710Z6 VL1710ZI2 Ingredients (PHR) (PHR) (PHR) (PHR) (EV)(PHR)

100 (1.8% 100 (4.9% 100 (4.5% 100 (7.5% 100 (7.5% Polymer ENB) ENB) ENB) ENB) ENB) Paraffinic oil 5 5 5 5 5 Zinc Oxide 5 5 5 5 5 Stearic Acid 1 1 1 1 1 90% Sulfur 1 1 1 1 0 MBT 2 2 2 2 0 TMTD 0.8 0.8 0.8 0.8 2.5 ZDBDC 0.8 0.8 0.8 0.8 0 TDEDC 0.8 0.8 0.8 0.8 0.8 DPTT 0 0 0 0 2

As a formulation is cured using either the MDR or RPA, there is a point after the maximum is reached, where cross‐links may still be forming, however, the rate has

slowed considerably and the competing process of degradation begins to command the

observed torque vs. time plots. It is likely that degradation also occurs during the curing process, but the formation of cross‐links is considerably faster, thus, dominating.

Degradation after the MH can be attributed to cross‐links degrading, the polymer

backbone degrading, or a combination of the two. Sulfur pickup and desulfuration are reciprocal reactions and since there are still reactive species present, resulfuration is also a distinct possibility for cross‐link degradation.

To explore whether resulfuration was the cause of the degradation an efficient vulcanization was carried out using VL1710ZI2 (Figure 25) at a temperature of 200°C. An efficient vulcanization is a type of vulcanization that does not utilize elemental sulfur and instead relies on sulfur donation directly from the curatives. This, in turn, limits the maximum amount of sulfur that can be contained in the preliminary cross‐link. Because there was still degradation observed with this system, even in the absence of large initial cross‐links, suggests that resulfuration does not occur and that the degradation might be caused by bond rupture in the polymer backbone.

EV Cure at 200°C for 20 Minutes 7

4 (in*lbs)

Torque 2

0 0 5 10 15 20 Time (min)

Figure 25. Curing of VL1710ZI2 at 200°C for 20 minutes.

To test if the polymer backbone was breaking down, TGA and DSC analyzes were performed on raw polymer, uncured and cured VL1710Z6 (7.5% ENB) samples. On both instruments a 200°C isothermal run was done and the data are presented in Figure 26 andFigure 27. Based on these results, there is no observable polymer degradation, which might indicate that degradation is not a result of temperature alone, but coupled to the mechanical stresses induced by the MDR or RPA analysis. In addition, a sample of

EPDM‐5, which was the polymer used in VL1710Z6, was then heated in an oven at 200°C for 1 hour. This sample was no longer soluble in NMR solvents due to thermally induced cross‐link formation, thus ATR‐FTIR spectroscopy was utilized in an attempt to analyze the polymer sample. Due to ENB only consisting of 7.5% of the polymer by weight at

maximum, quantification of the double bond signal as confirmation of degradation was unobtainable.

Weight vs. Time 120

100

60 Raw Polymer Percent

40 Uncured Cured Weight 20

0 0 5 10 15 20 25 30 ‐20 Time (min)

Figure 26. TGA traces of (blue) raw polymer (EPDM‐5), (green) cured and (red) uncured samples of VL1710Z6. Samples were run under an air atmosphere.

Heat Flow vs. Time 10 0 0123456789101112131415161718192021 ‐10 ‐20 (mW) Cured ‐30

Flow Uncured

‐40 Raw Polymer Heat ‐50 ‐60 ‐70 Time (Min)

Figure 27. DCS traces of raw polymer (EPDM‐5) (blue), cured (green) and uncured (red) samples of VL1710Z6

Without a proper method to analyze how degradation was taking place the focus was shifted to comparing the rate of degradation to the rate of cure and how temperature is factored into their relationship. To understand general trends, an initial qualitative approach was taken. For all of the standard formulations and temperatures,

the slope of the cure, at the steepest part of the curve, and the slope after MH were utilized (Figure 28). The point where the trendlines of cure and degradation intersect is the theoretical MH if the rate change at MH was instantaneous. From this data, it is readily apparent that degradation does happen in the curing process.

Cure and Degradation rates of VL1710Z6 at 180°C

25 Whole Plot 20 Derivative 15 (in*lbs)

Degradation 10 Cure

Torque 5 Linear (Degradation) 0 0246810Linear (Cure) ‐5 Time (min)

Figure 28. An example of how the intersect points were determined. The slope of the curing process was taken near the highest rate and the slope of the degradation was taken from where the MH was reached until the end.

All of the intersect points were then compared and the theoretical MH values

showed some interesting trends based on ENB levels (Figure 29). The MH itself does

not follow a particular trend with ENB content, which is due to different polymer

manufacturers being used. However, based on the slopes, it appears that as the temperature is increased, the higher ENB formulations are less prone to the degradation process.

Torque vs. Temperature of the Theoretical MH 10 VL1710Z1 ENB ‐ 1.8%

9.5 y = ‐0.0544x + 18.698

9 VL1710Z3 ENB ‐ 4.9% (in*lbs) 8.5 y = ‐0.0345x + 15.588 8 Torque VL1710Z5 ENB ‐ 4.5% 7.5 y = ‐0.0416x + 15.978 7 VL1710Z6 ENB ‐ 7.5% 160 170 180 190 200 210 Temperature (°C) y = ‐0.0275x + 14.264

Figure 29. Torque vs. temperature of the intersect points of the various formulations at different temperatures.

Since the cure and degradation rates appear to be highly temperature dependent, a relationship between reaction rate and temperature needs to be established.

Accordingly, Arrhenius plots were made to examine if the activation energy confirmed this dependency, as well as to probe why degradation happens more with higher temperatures when compared to curing (Figure 30‐Figure 32). The cure rate was

determined by taking the slope of the steepest part of the cure curve and the degradation rate was determined by taking the slope of the curve after MH was achieved. These slopes ended up becoming the rate constants due to the linear

relationships used to obtain the slopes.

Arrhenius Plot of Curing 3.5

2.5

2 VL1710Z1 (ENB 1.8%) (k)

ln 1.5 VL1710Z3 (ENB 4.9%)

1 VL1710Z5 (ENB 4.5%) VL1710Z6 (ENB 7.5%) 0.5

0 0.0021 0.00215 0.0022 0.00225 0.0023 T (1/K)

Figure 30. The Arrhenius plots for the curing process of VL1710Z1, VL1710Z3, VL1710Z5, and VL1710Z6.

Arrhenius Plot of Degradation 0 ‐0.50.0021 0.00215 0.0022 0.00225 0.0023 ‐1 ‐1.5 ‐2 VL1710Z1 (ENB (1.8%) ‐2.5 VL1710Z3 (ENB 4.9%) ln(k) ‐3 VL1710Z5 (ENB 4.5%) ‐3.5 VL1710Z6 (ENB 7.5%) ‐4 ‐4.5 ‐5 T (1/K)

Figure 31. Arrhenius plots for the degradation process of VL1710Z1, VL1710Z3, VL1710Z5, and VL1710Z6.

Ratio of Cure to Degradation 8 7 6 5 VL1710Z1 (ENB 1.8%) 4 (kc/kd)

VL1710Z3 (ENB 4.9%)

ln 3 VL1710Z5 (ENB 4.5%) 2 VL1710Z6 (ENB 7.5%) 1 0 0.0021 0.00215 0.0022 0.00225 0.0023 T (1/K)

Figure 32. Arrhenius plots for the ratio of curing vs. degradation of VL1710Z1, VL1710Z3, VL1710Z5, and VL1710Z6.

Based on Figure 32, as the temperature was increased, the rate of degradation increased more than the rate of cure increased. An example of this can be seen in

Figure 17 where the MH of a 200°C cure is lower than that of a 170°C cure. Based on

Figure 32, at the temperature of 211°C there is a convergence to where no matter which polymer is used, the same cure to degradation ratio is observed. Since the

Arrhenius plot data showed different slopes for the different formulations, the relationship between ENB content and activation energy was explored for each process

(Figure 33).

Activation Energy vs. ENB Content 200

150 (kJ/mol)

100

Barrier Cure

50 Degradation

Activation 0 02468 ENB (%)

Figure 33. Activation energy required for both processes for varying ENB content

As seen in Figure 33, with an increase in ENB content, the activation energy for degradation is higher, which contributes to higher ENB content polymers being more resistant to degradation. A higher percentage of ENB yields more cross‐links, adding to the rigidity of the rubber, thus making the degradation process less likely. Another consequence of higher ENB levels is the presence of lower activation energies for the curing process. The decrease in Ea, with an increase in ENB level, could be caused by an increase in reactive site availability and/or mobility issues in lower diene polymers.

With higher diene polymers it takes less energy investment to find reactive species to react at cross‐link sites, whereas with the lower diene polymers the opposite is true.

3.6 Curve Fitting to Predict Reaction Rates

Some significant trends were observed with the initial kinetics study, however, since the reaction rate constants determined were only approximations, a more statistical method was needed. Curve fitting software, utilizing Equation 3, was employed. In almost every case the rate constant for degradation was unobtainable

due to a marching modulus for the lower temperatures. This marching modulus prevented obtaining a proper MH from the data and not showing clear degradation process after that point.

Figure 34. Fitted curve data for VL1710Z1 at various temperatures. Blue: RPA traces; Red: theoretical data obtained from curve fitting software; Green: the absolute difference between the theoretical and experimental curves.

It can be observed in Figure 34 that at lower temperatures the fitted curve does not follow the experimental data accurately, but as the temperature is increased it fits better. This is due to the rate fluctuation that is present at lower temperatures. The equation that is utilized to fit the curve does not account for this phenomenon as it assumes a steady‐state approximation for the formation of cross‐links. With the delay in the formation of cross‐links at the fluctuation point, the approximation fails to fit the experimental data.

Figure 35. Fitted curve data for VL1710Z3 at various temperatures. Blue: RPA traces; Red: theoretical data obtained from curve fitting software; Green: the absolute difference between the theoretical and experimental curves.

VL1710Z3 had better curve fits than VL1710Z1 as seen from comparing the absolute difference curves between Figure 34 andFigure 35. The reason for this can be

attributed to VL1710Z3 not exhibiting a rate fluctuation like that observed with

VL1710Z1.

Figure 36. Fitted curve data for VL1710Z5 at various temperatures. Blue: RPA traces; Red: theoretical data obtained from curve fitting software; Green: the absolute difference between the theoretical and experimental curves.

VL1710Z5 showed the best fit out of all of the formulations, even at lower temperatures. The absolute difference in curves, in Figure 36, is the lowest when

compared to the other fitted curve sets.

Figure 37. Fitted curve data for VL1710Z6 at various temperatures. Blue: RPA traces; Red: theoretical data obtained from curve fitting software; Green: the absolute difference between the theoretical and experimental curves.

VL1710Z6, as seen in Figure 37, also exhibits good fits with experimental data. Since there appeared to be an additional degradation reaction occurring at higher temperatures in some plots, that portion of the curve was excluded. This additional degradation may be due to an auto‐catalytic reaction that happens well after the MH has been reached and since the auto‐catalytic reaction is not part of the normal degradation, it can be excluded from the data. By using k1 and k2, as solved by the fitting software, Arrhenius plots were prepared from these data (Figure 38 Figure 39).

Since the fitted data at 170°C and 180°C resulted in negative k2 values, those values

were excluded in Figure 39.

Arrhenius Plot of Curing 2.5

1.5 VL1710Z1 (1.8% ENB)

ln(k) 1 VL1710Z3 (4.9% ENB) VL1710Z5 (4.5% ENB) 0.5 VL1710Z6 (7.5% ENB) 0 0.0021 0.00215 0.0022 0.00225 0.0023 T (1/K)

Figure 38. Arrhenius plot of curing for all four formulations.

Arrhenius Plot of Degradation 0 0.0021 0.00212 0.00214 0.00216 0.00218 ‐1 ‐2 ‐3 VL1710Z1 (1.8% ENB) ‐4 VL1710Z3 (4.9% ENB) ln(k) ‐5 VL1710Z5 (4.5% ENB) ‐6 VL1710Z6 (7.5% ENB) ‐7 ‐8 T (1/K)

Figure 39. Arrhenius plot of degradation for all four formulations.

When comparing Figure 30 andFigure 31 to Figure 38 andFigure 39, the trends set forth in the former are not as neatly followed in the latter. The fitted line attempts to fit the whole experimental curve, but a standard kinetic model cannot be used to describe the

entirety of the data set. However, focusing on only points of interest, as done prior to using kinetic modeling, leaves out some sections of the curve that could provide a more realistic representation of what is being shown.

Activation Energy vs. ENB Content 700 90 80 (kJ/mol)

600 70 (kJ/mol) 500 60 400 50 Curing

Degradation 40

300 of Degradation

of 30 200 Cure 20 Barrier 100 Barrier

10 0 0 Energy

Energy 02468 ENB (%)

Figure 40. Activation energy required for both processes.

When examining the trend of the cure data from Figure 40, it follows a similar pathway as seen earlier from Figure 33. Degradation, on the other hand, seems to have an opposite slope. A possibility could be that only two rate constants of degradation were

able to be used when constructing the Arrhenius plot, which could also have contributed to the outlier caused by VL1710Z3.

4 Conclusion

To understand what contributes to the curing of rubber, each ingredient, side reactions that occur, even polymer structure are important details to consider. As a result, much about how exactly curing happens is poorly understood. Although direct analysis of the degradation pathway is currently impossible, two different methods were used to investigate how degradation effects curing. The findings suggest that variables such as ENB content and temperature influence this degradation. It was found that curing metallocene polymers with higher ENB content allowed for resistance to degradation. However, at a temperature over 211°C, there is no benefit to utilizing high contents as all of the polymers showed similar degradation to cure profiles.

5 References

1. Kirk, R. E.; Othmer, D. F. Concise Encyclopedia of Chemical Technology, 3rd ed.; Wiley‐ Interscience: New York, 1985; p 393.

2. Morawetz, H. History of Rubber Research. Rubber Chemistry and Technology 2000, 73 (3), 405‐426.

3. ExxonMobile Chemical. Vistalon™ ethylene propylene diene (EPDM) rubber. http://www.exxonmobilchemical.com/Chem‐English/brands/vistalon‐ethylene‐propylene‐ diene‐epdm‐rubber.aspx?ln=productsservices (accessed July 2, 2014).

4. Duin, M. v. Chemistry of EPDM Cross‐linking. KGK, Kautsch. Gummi Kunstst. 2002, 55, 150‐ 156.

5. Cossee, P. Ziegler‐Natta Catalysts I. Mechanism of Polymerization of α‐Olefins with Ziegler‐ Natta Catalysts. Journal of Catalysts 1964, 3, 80‐88.

6. Alt, H. G.; Koppel, A. Effect of the Nature of Metallocene Complexes of Group IV Metals on Their Performance in Catalytic Ethylene and Propylene Polymerization. Chem. Rev. 2000, 100, 1205‐1221.

7. Kaminsky, W. Highly Active Metallocene Catalysts for Olefin Polymerization. J. Chem. Soc. Dalton Trans 1998, No. 9, 1413‐1418.

8. Sperling, L. H. Introduction to Physical Polymer Science, 4th ed.; John Wiley and Sons: Hoboken, 2005.

9. Ghosh, P.; Katare, S.; Patkar, P.; Caruthers, J. M.; Venkatasubramanian, V. Sulfur Vulcanization of Natural Rubber for Benzothiazole Accelerated Formulations: From Reaction Mechanisms to a Rational Kinetic Model. Rubber Chemistry and Technology 2003, 76 (3), 592‐693.

10 Milani, G.; Milani, F. Comprehensive Numerical Model for the Interpretation of Cross‐linking . with Peroxides and Sulfur: Chemical Mechanics and Optimal Vulcanization of Real Items. Rubber Chemistry and Technology 2012, 85 (4), 590‐628.

11 Dullaert, K.; Doremaele, G. V.; Duin, M. V.; Dikland, H. Quanitative Assessment of the . Branching Architecture of EPDM With High Content of 5‐vinyl‐2‐norbornene as Third Monomer. Rubber Chemistry and Technology 2013, 86 (4), 503‐520.

12 Mitra, S.; Jorgensen, M.; Pedersen, W. B.; Almdal, K.; Banerjee, D. Structural determination . of Ethylene‐Propylene‐Diene Rubber (EPDM) Containing High Degree of Controlled Long‐ Chain Branching. Journal of Applied Polymer Science 2009, 113, 2962‐2972.

13 Ding, R.; Leonov, A. I.; Coran, A. Y. A study of the Vulcanization Kinetics of an Accelerated‐ . Sulfur SBR Compound. Rubber Chemistry and Technology 1996, 69 (1), 81‐91.

14 Future Foundation. Oscillating Disk RheometerTechnical Datasheet, 2003. Future . Foundation. http://www.ffinstruments.com/odrpdf/Oscillating%20Disc%20Rheometer%20Technical%20 DataSheet.pdf (accessed July 15, 2014).

15 Wang, L.; Wang, X.; Yang, M.; Wang, Y.; Li, L.; Liu, B.; Kim, I. Addition Polymerization of 5‐ . Ethylidene‐2‐norbornene by Cationic Palladium Cationic Complex and Subsequent Derivatization. Macromolecular Research 2011, 19 (10), 1071‐1076.

16 Levine, I. N. Physical Chemistry, 6th ed.; McGraw‐Hill Science: Boston, 2008. .