Multifunctional Acrylates as Anti-ReversionAgents in Sulfur Cured Systems

*Speaker

5522 07/11

Steven K. Henning* and Scott A. Shapot

Presented at the Fall 168th Technical Meetingof the Rubber Division, American Chemical Society

Pittsburgh, PANovember 1-3, 2005

Cray Valley USA, LLC • Oaklands Corporate Center • 468 Thomas Jones Way, Suite 100 • Exton, PA 19341 877-US1-CRAY (877-871-2729) • Web: www.crayvalley.com

3

ABSTRACT

Multifunctional acrylates have found utility as effectiveanti-reversion agents in sulfur cured applications whereovercure or high temperature cure conditions exist. Ithas also been demonstrated that certain zinc salts ofmethacrylic acid reduce reversion. In this paper,commercially available multifunctional acrylates areevaluated in both conventional and semi-EV sulfurcured systems with the reversion resistance and curedphysical properties of the resulting natural rubber/carbon black vulcanizates compared to othercommercially produced agents.

It will be shown that the multifunctional acrylate (CrayValley® SR534) provides excellent reversionresistance. In addition, this anti-reversion agentdemonstrates the least change in physical propertieswhen overcure conditions are induced.

Data will be provided demonstrating the uniqueproperties of acrylates including reversion resistance,retention of physical properties and crosslink densityas a function of reversion, viscoelastic properties,flexural fatigue results, and the optimization of anti-reversion agent level in a model natural rubbercompound.

INTRODUCTION

Physical properties of vulcanized rubber systems areinfluenced by the curing process and chemical natureof the formed crosslinks. The bond strength of linkagescan help predict the performance of a given vulcanizatein specific applications. Peroxide cured systems areoften employed in static or compressive conditionswhere low set and heat resistance are required. Anetwork derived from polysulfidic linkages is preferredfor components subject to dynamic strain whereimproved flexural fatigue and tear properties aredesired. Bond dissociation energies are often used toquantify the strength of the crosslinks and differentiatethe systems.1,2 Carbon-carbon linkages possess highdissociation energies and resist failure to a limiting point,then fail catastrophically. Having lower dissociationenergies, polysulfidic bonds break more readily understrain, but due to their chemical nature also possessthe ability to reform and alleviate stresses.3,4

The dynamic nature of polysulfidic bonds characterizesthe utility of the system, but also facilitates reversion.Reversion is defined by a loss in physical propertiesassociated with degradation of network integrity. Thereversion process is thermally initiated and primarilyassociated with overcure or high temperature serviceconditions. Reversion involves reactions that lead tothe desulfuration of polysulfidic linkages and main-chainmodification (cis-to-trans isomerization) which resultsin weaker network structures.5,6 As desulfurationprogresses, the distribution shifts from polysulfidic tomono- and di-sulfidic crosslinks and eventuallycrosslink density is also lost. The result is a degradationof physical properties and a decrease in theperformance of the rubber article.7

Many strategies have been developed to prevent thereversion process. Reversion can be minimized in curesystems which promote mono- and di-sulfidiccrosslinks. Sulfur-donor systems or polysulfidepolymers can be utilized.8 It has recently beenproposed that certain zinc methacrylate salts, whenused as activators in accelerated sulfur vulcanization,promote crosslinks of lower sulfur rank and minimizereversion.9 These attempts to mediate reversion alsoalter the original structure of the network itself, typicallydecreasing dynamic and tear properties that are thedirect benefits of polysulfidic linkages.

Alternately, the inclusion of more stable crosslinks inaddition to the sulfur network has been proposed.10

Similarly, the mixing of peroxide and conventionalaccelerated sulfur cure systems has been attempted.11

A disadvantage of this compounding method is the needto reformulate in order to compensate for the new,structurally different crosslinks in order to maintain thedesired cured properties.

A more popular approach is to compensate for thedestruction of crosslinks by adding chemical additivesthat react when reversion processes are initiated. Insuch systems, the desulfuration mechanism is offset bythe formation of new, stable crosslinks. Ideally, theprocesses have similar rates such that crosslink densityis maintained. It has been shown that acrylate estersare effective anti-reversion agents that display thisbehavior.12 Earlier work has also demonstrated theutility of acrylates to improve the heat-aged properties

4

of rubber compounds, suggesting a similarmechanism.13

The present study evaluates the performance of amultifunctional acrylate ester product (Cray Valley®

SR534) specifically optimized to prevent reversion inrubber compounds. SR534 is compared to the imide-based anti-reversion agents 1,3-bis(citraconimidomethyl)benzene and N,N’-m-phenylenedimaleimide. The former compound hasbeen promoted as capable of crosslink compensationreactions,14 while the latter is commonly used as asecondary accelerator. Cure characteristics and anti-reversion properties will be compared. Tensile,viscoelastic, and flexural fatigue properties areevaluated. Special attention is placed on comparingthe efficiency with which the various additives maintainphysical properties as a function of reversion.

EXPERIMENTAL

Chemicals

The compounds evaluated in this study are given inTable I. The commercially available products wereused without further treatment or purification.

Table I. Anti-reversion agents.

Chemical Additive Code

multifunctional acrylate ester SR5341,3-bis (citraconimidomethyl)benzene CIMB

N,N'-m- phenylenedimaleimide PDM

Rubber Compounding

Using an internal mixer, a non-productive masterbatchwas prepared containing all ingredients except anti-reversion agents and curatives. Table II outlines theformulation used for the masterbatch and the specificcure packages. Conventional (CV), semi-efficient (SE-V), and efficient (EV) vulcanization systems wereevaluated. These are model formulations that have notbeen optimized for a specific application. Theproductive stage was mixed using a two-roll mill.Milling times were approximately 5 minutes. Productivestock was aged overnight prior to testing.

Physical Testing

The determination of vulcanization behavior wasperformed on a Tech Pro MDPT moving die rheometer(MDR) according to ASTM D 5289. Cure temperaturefor sample preparation is 160ºC unless otherwisenoted. Reversion was calculated as the difference intorque between the highest initial recorded value (MH)and that at 60 minutes (M60). The individual calculatedt90 times (optimal cure) were used for subsequent testsample preparation. In addition, reverted samples wereprepared by curing to 60 minutes. Stress-strain andtear data was acquired on a Thwing-Albert MaterialsTester following ASTM D 412 and D 624 (Die C).The MDR was also employed to evaluate bothunvulcanized and cured dynamic properties undersinusoidal shear conditions at frequencies andamplitudes defined later in the text. DeMattia flexfatigue was determined according to ASTM D 813under adiabatic conditions starting at room temperature.Complete testing results are provided in Appendices Iand II.

5

Table II. Formulations.

RESULTS AND DISCUSSION

Model Formulation Development

In order to compare the given anti-reversion agents, itis best to understand what formulation characteristicscontribute to the reversion process. Elastomers withlow degrees of unsaturation provide the least amountof reversion, while reversion is promoted in unsaturatedrubber containing high allylic hydrogen content. Figure1 correlates the extent of reversion as a function ofallylic hydrogen for a series of elastomers. Theformulation used was identical to Table II, substitutingvarious polymers (ethylene-propylene-dienemonomer, EPDM; emulsion styrene-butadiene rubber,SBR; cis-butadiene rubber, BR; cis-isoprene rubber,IR; natural rubber, NR).

Swelling Experiment

Samples were mixed in a Brabender Prep Mixer usingthe same formulation in Table II with the carbon blackomitted from the masterbatch. The unfilled compoundswere mixed for 4 minutes at 60 rpm and 60ºC. TheMDR was used to determine optimal cure times andvulcanizates were prepared. Swelling experimentswere conducted on cylindrical samples of knownweight and dimensions. Crosslink density wasdetermined using the experimental methodologyoutlined by Flory and Rhener1 and modified withequivalent terms derived from theories of rubberelasticity. The χ parameter used for the toluene/cis-poly(isoprene) system was 0.37.

Ingredient, phr CV S-EV EV

Non-Productive Natural Rubber (CV-60) 100.0 100.0 100.0Carbon Black (N 330) 50.0 50.0 50.0Zinc Oxide 5.0 5.0 5.0Stearic Acid 1.5 1.5 1.5

Productive Antioxidanta 1.0 1.0 1.0Anti-Reversion Agent 0, 2.0, 5.0 0, 2.0, 5.0 0, 2.0, 5.0Sulfur 2.5 1.2 0.2TBBSb 0.6 1.6 2.0TMTDc 2.0

a2,2,4-trimethyl-1,2-hydroquinolinebN-t-butylbenzothiazole-2-sulfenamidectetramethylthiuram disulfide

6

Figure 1. Reversion as a function of allylic hydrogen concentration.

Table III. Cure types and bond distributions.

The rheometer cure profiles for the listed cure packagesare provided in Figure 2. It is clear that as the sulfur-to-accelerator ratio decreases (CV to S-EV) reversionwas reduced. In an efficient vulcanization (EV), sulfurdonors are utilized and no reversion was evident.Kinetic and cure state data from these vulcanizationsis provided in Table IV.

EPDM SBR

BR

IR

NR

0

10

20

30

40

50

60

0 1 2 3 4 5 6 7 8

Allylic Hydrogen / Monomer Unit

MH

-M60

/MH

(%)

It has been suggested that while the amount of sulfurcontrols the overall reaction kinetics, the amount andtype of accelerator determines the overall length ofthe sulfidic linkages.1,2 The sulfur-to-accelerator ratiois used as a guideline for determining the type of curepackage in use and the distribution of sulfur crosslinkrank.3 Table III provides an outline of cure types, sulfur-to-accelerator ratios, and typical crosslink distributions.

CV S-EV EV

Sulfur-to-Accelerator Ratio 1 - 1.7 1.7 - 0.4 0.4 - 0.08

Polysulfidic % 60-80 30-50 5-25Mono-, Disulfidic % 20-40 50-70 75-95

Figure 2. Vulcanization profiles of conventional (CV), semi-efficient (SE-V) and efficient (EV)vulcanizations.

0

5

10

15

20

25

30

35

0 10 20 30 40 50 60

Time (minutes)

Torq

ue (d

Nm

)

CV

EV

S-EV

7

Table IV. Vulcanization data by cure type.

Having the highest allylic hydrogen concentration, cis-1,4 poly(isoprene) (synthetic and natural rubber)displayed the greatest loss in crosslink density. Inaddition, the data indicates that reversion was mostpronounced using a cure package which promotespolysulfidic crosslinks. Therefore, model formulationsbased on natural rubber using both CV and S-EV curesystems have been selected to best evaluate thebehavior and efficiency of anti-reversion agents.

Reversion Resistance

SR534 is a multifunctional acrylate ester compoundthat has been optimized for use in rubbercompounding. In a more traditional role as coagentsfor peroxide cure systems, acrylate and methacrylateesters are available in a wide range of structuralvariation. SR534 was developed specifically toaddress reversion resistance in sulfur vulcanized,unsaturated rubber compounds. Figure 3 comparesthe reversion resistance properties of SR534 to adiacrylate (1,4-butanediol diacrylate) andtrimethacrylate (trimethylolpropane trimethacrylate) at5 phr in the model formulation. SR534 is a blend ofselected multifunctional acrylates and outperforms thediacrylate. The data demonstrates that the methacrylatewas also not effective, a finding reported in earlierwork.12 The diacrylate and trimethacrylate appearedto limit cure (MH), while SR534 matched thecrosslinking efficiency of the control. Vulcanizationdata is outlined in Table V. A higher torque value atM60 compared to MH indicates marching modulusand therefore no reversion.

CV S-EV EV

t2 (minutes) 1.42 3.22 1.83t90 (minutes) 4.51 5.33 4.81delta torque (dNm) 29.60 28.35 26.46reversion (dNm) 14.85 8.41 0

Figure 3. Activity of SR534 versus diacrylate and trimethacrylate; 5 phr additive.

05

1015

202530

3540

4550

0 10 20 30 40 50 60

Time (minutes)

Torq

ue (d

Nm

) SR534

1,4-Butanediol diacrylate

control

TMPTMA

8

The use of acrylate ester products as coagents forperoxide-cured systems is a major application, andalso demonstrates their ability to participate in free-radical initiated addition and grafting reactions.1 Giventhe chemical environment of the model sulfur-curedformulation, a crosslinking mechanism relying onradical intermediates may be unlikely. Containingacrylate ester functionality, the structure of SR534 isfundamentally different than the imide-based products.However, the ability to participate in similar crosslinkingreactions has also been proposed.2,3,4 Figures 4 and 5provide a comparison of these agents in both the CVand SE-V formulations, at 5 phr of additive and a curetemperature of 160ºC. Table VI outlines thevulcanization data in these formulations at both 2 and5 phr of additive. It is shown that both SR534 andCIMB provide similar rheological profiles that aredistinguished by a lack of reversion. Both additivesnot only matched the MH of the control, but maintainedtorque with time. PDM increased MH and lostconsiderable torque during overcure.

Table V. Reversion data for various agents.

The imide-based commercially available anti-reversionagents PDM and CIMB were also evaluated in thesame formulation. The reported mode of activity forCIMB is to compensate for the loss of crosslinkdensity as a function of reversion by generatingadditional, non-sulfur crosslinks that are thermallystable.10,14 A Diels-Alder reaction has been proposed.PDM is a secondary accelerator also utilized forreversion resistance, but is more active and leads tothe addition of non-sulfur crosslinks during thevulcanization step through a similar reaction scheme.

Figure 4. Activity of SR534 compared to CIMB and PDM; CV formulation, 5 phr additive.

0

10

20

30

40

50

60

0 10 20 30 40 50 60

Time (minutes)

Torq

ue (d

Nm

)

SR534

PDM

control

CIMB

MH M60 Reversionphr (dNm) (dNm) (dNm)

control -- 31.05 16.20 14.85

SR534 5 30.60 33.25 0Diacrylate 5 24.54 22.30 2.24Trimethacrylate 5 25.29 16.60 8.69

9

Figure 5. Activity of SR534 compared to CIMB and PDM; S-EV formulation, 5 phr additive.

0

10

20

30

40

50

60

0 10 20 30 40 50 60

Time (minutes)

Torq

ue (d

Nm

)SR534, CIMB

PDM

control

As the most desirable aspects of sulfur-cured systemsare attributed to the polysulfidic nature of the crosslinks,the addition of non-sulfur crosslinks during the initialvulcanization step may be undesirable. Therefore,SR534 was developed considering that an ideal anti-reversion agent would provide minimal impact tooptimally cured (t90) properties, and primarily react tomaintain crosslink density during conditions that initiate

reversion. CIMB has also been promoted to exhibitthis type of behavior. Its form differs from PDM byaltering the attachment and sterics of the imide structure,favorably affecting the activity of the compound.Similarly, SR534 was developed by designing astructure that exhibits activity closely matching CIMB,without incorporating amine structures.

Table VI. Reversion data for various additives, 160ºººººC.

CV Formulation, 160oCMH M60 Reversion MH M60 Reversion

phr (dNm) (dNm) (dNm) phr (dNm) (dNm) (dNm)

control -- 31.05 16.20 14.85

SR534 2 27.57 23.70 3.87 5 30.60 33.25 0CIMB 2 29.60 26.20 3.40 5 32.99 32.89 0.10PDM 2 36.03 24.40 11.63 5 48.58 36.80 11.78

S-EV Formulation, 160oCMH M60 Reversion MH M60 Reversion

phr (dNm) (dNm) (dNm) phr (dNm) (dNm) (dNm)

control -- 29.61 21.20 8.41

SR534 2 30.04 26.10 3.94 5 29.74 30.00 0CIMB 2 29.44 29.00 0.44 5 28.54 30.68 0PDM 2 40.87 31.70 9.17 5 51.96 42.40 9.56

10

Table VI. Reversion data for various additives, 160ºººººC.

The same compounds were evaluated at elevated curetemperature as well. While extended cure time at160ºC provides a comparison of additive performanceunder overcure conditions, extending cure at 180ºCcan simulate reversion under high temperature serviceconditions. Table VI outlines the reversion resistance

The cure behavior of the CIMB and SR534 at elevatedtemperatures and 5 phr loading was characterized bya marching modulus (Figure 6). Such behavior

Reversion Reversion Reversion Reversionphr (dNm) phr (dNm) phr (dNm) phr (dNm)

control -- 14.30 control -- 12.24

SR534 2 2.33 5 0 SR534 2 2.89 5 0CIMB 2 0 5 0 CIMB 2 0 5 0PDM 2 9.79 5 8.61 PDM 2 9.71 5 7.60

CV Formulation, 180oC S-EV Formulation, 180oC

demonstrates the ability to provide adequate reversionprotection under high temperature conditions.

Both SR534 and CIMB were included in an expandedstudy to determine the optimal loading of anti-reversionagents in the model CV formulation. Reversion wasdetermined for loadings between 0.5 and 5.0 phr of

of the additives when cured to 60 minutes at 180ºC.Again, both SR534 and CIMB virtually eliminatedreversion under these more demanding conditions,while PDM allowed significant reductions in measuredtorque over time.

Figure 6. Activity of SR534 compared to CIMB and PDM; CV formulation, 5 phr additive, 180ºººººC.

0

10

20

30

40

50

60

0 10 20 30 40 50 60Time (minutes)

Torq

ue (d

Nm

)

SR534

PDM

control

CIMB

SR534 or CIMB. The data is presented in Figure 7.The reduction of reversion as a function of loading wassimilar for each compound. In the given formulation,reversion is effectively eliminated at 3 phr.

11

Figure 7. Optimization of additive loading in model formulation (CV).

0

2

4

6

8

10

12

0 1 2 3 4 5 6

phr

MH

-M60

(dN

m)

SR534CIMB

were significantly higher than the control. The modulusvalues of SR534 and CIMB were also elevated in thismodel formulation. However, both SR534 and CIMBdisplayed a significant improvement over PDM in termsof modulus contribution during vulcanization. Theloadings of additive used in this study were selected toprovide data points both above and below that whicheliminated reversion (3 phr, Figure 7). It is suggestedthat an optimized loading would typically be chosenbased on a balance between reversion resistance andmodulus increase. The optimal loading may be differentfor each additive studied.

Tensile Properties

The physical properties of rubber compoundsemploying SR534, CIMB, and PDM were comparedwith each sample cured to the individually calculatedt90 cure times. Data was normalized to the control (noanti-reversion agent) and is provided in the CVformulation at both 2 and 5 phr (Figures 8 and 9,respectively). The data highlights the departure fromthe physical properties of the control formulation. It isclear that PDM reacts preferentially during the initialvulcanization step, as both 100% and 300% modulus

Figure 8. Physical properties of compounds using 2 phr of additive, normalized to control (CV).

0

50

100

150

200

250

TensileStrength

Elongation %

100%Modulus

300%Modulus

TearStrength

Nor

mal

ized

Val

ue controlSR534CIMBPDM

12

Figure 9. Physical properties of compounds using 5 phr of additive, normalized to control (CV).

0

50

100

150

200

250

TensileStrength

Elongation %

100%Modulus

300%Modulus

TearStrength

Nor

mal

ized

Val

ue controlSR534CIMBPDM

Test samples of the same compounds were intentionallysubjected to conditions that would lead to reversion.Overcure conditions were simulated by curing past t90times to 60 minutes. Following the methodologyoutlined above, the data collected in the reverted statewas also normalized to the original control values (no

By using this equation, the efficiency with which theanti-reversion compounds maintain properties as afunction of heat history can be compared. The controlitself was also subjected to the same treatment, offeringa baseline for the relative amount of loss in properties.Figures 10 and 11 present the results in the CV andS-EV formulations at 5 phr additive. A negativedeviation represents a loss in properties with cure time.

additive, cured to t90). In order to provide perspectiveon how physical properties change as a function ofinduced reversion, the deviation in normalizedproperties is presented. Deviation is calculatedaccording to Equation 1:

Equation 1.

property @ 60 minute cure property @ t90 cure

control property @ t90 cure control property @ t90 cureDeviation = –

property @ 60 minute cure property @ t90 cure

control property @ t90 cure control property @ t90 cureDeviation = –

The data shows that compounds containing PDM lostthe most amount of modulus as a function of exposureto reversion conditions, followed by the control.SR534 and CIMB increased modulus. In bothformulations, SR534 had the least deviation in modulus,while CIMB gained considerably. Perhaps mostimportantly, the deviation in 300% modulus was theleast for SR534. Following the relative magnitude ofdeviation in modulus is the loss of elongation and tearproperties.

13

Figure 10. Deviation in physical properties as a function of induced reversion.CV formulation, 5 phr additive.

Figure 11. Deviation in physical properties as a function of induced reversion.S-EV formulation, 5 phr additive.

-100

-75

-50

-25

0

25

50

75

100

TensileStrength

Elongation %

100%Modulus

300%Modulus

TearStrength

Dev

iatio

n

controlSR534CIMBPDM

-75

-50

-25

0

25

50

75

100

TensileStrength

Elongation %

100%Modulus

300%Modulus

TearStrength

Dev

iatio

n

controlSR534CIMBPDM

By considering the overall deviation, the efficiency withwhich each additive maintains physical properties canbe compared. Taking the sum of the deviations foreach compound (absolute values are used as anydeparture from optimally cured properties would beundesirable) the relative performance can becompared. The data is presented in this manner for

both the CV and SE-V formulations in Figures 12 and13, respectively. The net deviation from the optimallycured properties is lowest for the multifunctionalacrylate ester product. SR534 was superior to CIMBin maintaining physical properties when subjected toovercure conditions.

14

Figure 12. Net deviation in physical properties in the CV formulation.

Figure 13. Net deviation in physical properties in the SE-V formulation.

0 25 50 75 100 125 150 175 200 225 250

SR534

CIMB

Net Deviation

5 phr2 phr

0 25 50 75 100 125 150 175 200 225 250

SR534

CIMB

Net Deviation

5 phr2 phr

Interestingly, net deviations increase at higher additiveloading (2 to 5 phr). These results are primarily drivenby the tendency of these compounds to increasemodulus when overcured. Again, optimization of

additive level balancing both modulus rise andreversion resistance was not considered in thisfundamental study, and is best addressed in a specificformulation.

15

Viscoelastic Properties

The viscoelastic properties of the CV compoundswere also compared. After curing to t90 in the MDR,the samples remained in the test cavity and weresubjected to various strain amplitudes at a frequencyof 1 Hz and a temperature of 100ºC. Actual valuesare provided in Appendix II. The normalized resultsare given in Figure 14 for compounds containing 5 phradditives. The addition of anti-reversion agents resultedin decreased tangent delta (1 Hz, 10% amplitude)

versus the control. A lower tangent delta value can becorrelated to improved compound hysteresis. Highstrain shear modulus (G’, 1 Hz, 50% amplitude) datais also provided. PDM increases G’ considerablycompared to the control, while SR534 and CIMBproduce values slightly below that of the control.Overall, higher loadings of anti-reversion agentproduced increased G’ values. The addition of SR534produces the lowest shear modulus. SR534 is the onlyliquid product, and may contribute to lower compoundshear viscosity (cured).

Figure 14. Viscoelastic properties of compounds using 5 phr of additive, normalized to control (CV).

Figure 15. Deviation in viscoelastic properties as a function of induced reversion. CV formulation, 5phr additive.

0

20

40

60

80

100

120

140

160

control SR534 CIMB PDM

Nor

mal

ized

Val

ue

tan dG'

-60

-40

-20

0

20

40

60

control SR534 CIMB PDM

Nor

mal

ized

Val

ue

tan dG'

16

02468

101214161820

0 50 100 150 200 250cycles (E+03)

crac

k w

idth

(mm

)

CVS-EVEV

1.92

1.52

1.62

Viscoelastic properties were subsequently evaluatedas a function of reversion. The samples were curedfor 60 minutes in the MDR and subjected to the strainsweep described previously. These results are providedin Figure 15 as a deviation from optimal properties(normalized) according to the methodology outlinedpreviously. SR534 maintains the lowest tangent deltavalues in the reverted state. The data suggests that notonly does SR534 maintain network integrity whensubjected to overcure conditions, but may also reducethe heat build-up associated with applicationsexperiencing severe dynamic flex, reducing the auto-acceleration quality of reversion associated with theseservice conditions. Both the control and PDMcompounds lost considerable shear modulus uponreversion (typically > 40%), while SR534 and CIMBlimited the reduction in G’. The loss in viscoelastic

properties was generally reduced at higher additiveloading.

Flex Fatigue

SR534, CIMB, and PDM compounds were alsoevaluated for flex fatigue resistance. The addition ofnon-sulfur crosslinks or crosslinks of lower sulfur rankwill produce negative effects in dynamic cut growthtesting. To illustrate the trend, Figure 16 presents theresults using the control compounds. The control seriescorrelates cut growth with sulfur rank distribution. Forthis series, 100% modulus values were similar. Theflex fatigue performance became progressively worseas the concentration of mono- and di-sulfidic crosslinksincreased at the expense of polysulfidic. Crack growthrate increases EV > S-EV > CV.

Figure 16. Flex fatigue results for CV, S-EV, and EV controls. 100% modulus values are indicated(MPa).

Figure 17 outlines the results of fatigue testing whenincorporating the anti-reversion agents in the CVcompound (5 phr) and curing to t90. SR534 performedequivalent to the CV control formulation, while CIMBhad a slightly elevated growth rate. The compound

using PDM showed highest cut growth. Interestingly,the performance of SR534 was superior despite havingan initial modulus higher than the control, establishingevidence that the inclusion of an anti-reversion agentcan provide benefits in dynamic flex fatigue resistance.

17

Figure 17. Flex fatigue results at 5 phr additive cured to t90. 100% modulus values are indicated(MPa).

02468

101214161820

0 50 100 150 200 250cycles (E+03)

crac

k w

idth

(mm

)SR534CIMBPDMcontrol

3.82

2.31

2.69

1.62

Test specimens of the same compounds were curedto 60 minutes and tested. The results are provided inFigure 18. In the reverted state, flex fatigue performancenow correlates with modulus. SR534 outperformed

CIMB, and these compounds increased moduluscompared to the results in Figure 17. The modulus ofthe control and PDM was lower, the outcome ofreversion.

Figure 18. Flex fatigue results at 5 phr additive cured to 60 minutes. 100% modulus values areindicated (MPa).

02468

101214161820

0 50 100 150 200 250cycles (E+03)

crac

k w

idth

(mm

)

SR534CIMBPDMcontrol

1.24

3.45 3.243.34

Anti-reversion agents SR534 and CIMB are not fullyreacted when the compounds are optimally cured, andmay help de-couple the dependence of modulus onflex fatigue performance by improving the curednetwork during testing. However, the vulcanizates thatwere intentionally overcured prior to flexing would not

have this mechanism available. Therefore, moduluswould again strongly influence the results. Flex testingfrom optimally cured samples may most accuratelyreflect actual conditions. By interpreting the resultsdisplayed in Figure 17 and 18, the benefits of includinganti-reversion agents can be realized.

18

Swelling Experiments

Equilibrium swelling experiments are utilized todetermine the crosslink density (Vx) of vulcanizates.An unfilled (no carbon black) version of the CV modelformulation was used to create compounds using bothSR534 and CIMB additives (5 phr). A controlcompound was also included. Optimal cure times werecalculated using a rheometer, and samples cured toboth t90 and 60 minutes were prepared. Crosslinkdensities were calculated for each vulcanizate. The

results are provided in Figure 19. The control losesconsiderable network integrity with reversion. SR534provides crosslink density values most similar to thecontrol at optimal cure, and maintains crosslink densitywhen subjected to overcure conditions. CIMB reactsduring vulcanization to increase crosslink densityprematurely, and loses considerable density as afunction of cure residence time. Swelling experimentsmay best characterize the advantages possible whenincorporating SR534 into a reversion-proneformulation.

0.0E+00

1.0E-05

2.0E-05

3.0E-05

4.0E-05

5.0E-05

6.0E-05

7.0E-05

8.0E-05

control SR534 CIMB

Vx (

mol

/cm3 ) t90

60 min

Figure 19. Crosslink density (Vx) of unfilled vulcanizates prepared at optimal cure times (t90) and 60minutes.

SUMMARY AND CONCLUSIONS

The ideal anti-reversion agent would provide reversionresistance in a manner that does not significantly alterthe physical properties of the compound. Of the severalstrategies that result in a network capable of maintainingcrosslink density under conditions that would normallylead to reversion, only those additives whichcompensate for the loss of network integrity at a similarrate through the addition of new, heat-stable linkagesmeet the above criteria. Alternately, products whichestablish non-polysulfidic crosslinks duringvulcanization negatively affect desirable properties suchas dynamic flex fatigue and tear, and lead to highercured modulus. The activity of PDM characterizes the

behavior of this type of anti-reversion additive.Compensation for the loss in performance necessitatesreformulation.

SR534 is a multifunctional acrylate ester product thatcontains no amine groups and has been developed tominimize the effects on optimally cured properties andprovide maximum protection under conditions whichinduce reversion. In the model formulation, SR534demonstrated reversion resistance performance similarto CIMB in overcure and high temperature conditions.Both products reduced the impact on physicalproperties and flex fatigue compared to PDM whencured to t90. However, SR534 performed bestmaintaining physical properties when subjected to

19

Appendix I. Physical property data.

Appendix II. Viscoelastic property data.

overcure conditions. Dynamic properties were alsoimproved using SR534. Swelling experiments showedthat crosslink density was maintained at elevatedtemperatures using SR534 while both the control andCIMB lost network integrity.

The data provided suggests that SR534 is capable offorming additional, heat-stable crosslinks underconditions that would normally result in reversion. Byeliminating reversion and best maintaining desirablephysical properties, SR534 is an effective amine-freealternative to available anti-reversion products.

control Cure time, minutes 4.51 60.00 5.33 60.00Tensile Strength, MPa 21.08 14.26 21.46 15.67Elongation, % 580 515 600 590100% Modulus, MPa 1.62 1.24 1.52 1.52300% Modulus, MPa 8.16 6.82 8.03 6.06Tear Strength (Die C), kN/m 84.00 33.25 101.50 30.63tan delta (1 Hz, 10%) 0.160 0.199G' (1 Hz, 50%) 865.17 397.50

phr 2 2 5 5 2 2 5 5

SR534 Cure time, minutes 7.14 60.00 7.00 60.00 5.45 60.00 6.01 60.00Tensile Strength, MPa 21.50 18.26 22.36 17.05 25.42 23.08 25.56 22.05Elongation, % 515 465 505 375 545 535 580 470100% Modulus, MPa 2.31 2.51 2.69 3.34 2.51 2.45 2.34 2.82300% Modulus, MPa 10.30 10.30 11.47 12.78 12.06 11.09 10.96 12.23Tear Strength (Die C), kN/m 75.24 33.25 72.564 38.5 100.625 41.125 92.75 42.875

CIMB Cure time, minutes 6.07 60.00 6.50 60.00 5.57 60.00 5.86 60.00Tensile Strength, MPa 24.39 21.46 24.84 20.74 24.39 23.80 24.94 21.98Elongation, % 635 460 645 430 630 480 640 435100% Modulus, MPa 2.07 3.07 2.31 3.45 2.03 2.93 2.24 3.45300% Modulus, MPa 8.89 12.61 8.96 13.44 9.44 13.06 9.37 14.02Tear Strength (Die C), kN/m 85.75 38.50 92.75 49.88 118.13 49.88 97.13 44.63

PDM Cure time, minutes 6.11 60.00 6.35 60.00 6.26 60.00 7.15 60.00Tensile Strength, MPa 23.12 17.29 22.87 19.19 25.05 22.01 25.05 21.36Elongation, % 490 490 450 435 480 500 435 450100% Modulus, MPa 2.89 2.10 3.82 3.24 3.45 2.51 4.62 3.41300% Modulus, MPa 12.33 8.89 14.47 12.30 14.71 11.44 17.16 13.26Tear Strength (Die C), kN/m 64.75 48.13 49.00 41.13 70.88 42.88 63.00 46.38

CV FORMULATION SE-V FORMULATION

control Cure time, minutes 4.51 60.00tan delta (1 Hz, 10%) 0.160 0.199G' (1 Hz, 50%) 865.17 397.50

phr 2 2 5 5

SR534 Cure time, minutes 7.14 60.00 7.00 60.00tan delta (1 Hz, 10%) 0.093 0.181 0.096 0.152G' (1 Hz, 50%) 789.74 546.60 820.91 694.76

CIMB Cure time, minutes 6.07 60.00 6.50 60.00tan delta (1 Hz, 10%) 0.098 0.190 0.102 0.165G' (1 Hz, 50%) 833.06 749.75 850.11 806.36

PDM Cure time, minutes 6.11 60.00 6.35 60.00tan delta (1 Hz, 10%) 0.091 0.193 0.096 0.176G' (1 Hz, 50%) 958.04 513.47 1204.48 733.84

CV FORMULATION

20

1 L. H. Palys and P. A. Callais, Rubber World229(3), 35 (December 2003).

2 P. R. Dluzneski, Rubber Chem. Technol. 74,451 (2001).

3 W. Cooper, J. Polym. Sci. 28, 195 (1958).

4 L. Bateman, ed., “The Chemistry and Physicsof Rubber-Like Substances,” MacLaren,London, 1963.

5 M. Mori, Rubber Chem. Technol. 76, 1259(2003).

6 C. H. Chen, J. L. Koenig, J. R. Shelton, andE. A Collins, Rubber Chem. Technol. 54, 734(1981).

7 U. Shankar, Rubber Chem. Technol. 25, 241(1952).

8 J. Hahn, P. Palloch and N. Thelen, and H. J.Weidenhaupt, Rubber Chem. Technol. 74, 28(2001).

9 S. Henning, Paper # 10, Spring TechnicalMeeting, Rubber Division, ACS, SanAntonio, TX, May 16-18, 2005.

10 T. Kleiner, J. Neilsen, H. J. Weidenhaupt, W.Jeske, and H. Buding, Rubber & PlasticsNews, January 14, 2002, p. 14.

11 J. B. Class, US #6,245,861 (2001).

12 E. J. Blok, M. L. Kralevich and J. E. Varner,Rubber Chem. Technol. 73, 114 (2000).

13 Y. Hirata, et al. JP #63-150339 (1988).

14 A. H. M. Schotman, P. J. C. van Haeren, A. J.M. Weber, F. G. H van Wijk, J. W Hofstraat,A. G. Talma, A. Steenbergen, and R. N.Datta, Rubber Chem. Technol. 69, 727(1996).

15 P. J. Flory and J. J. Rhener, J. Chem. Phys.11, 521 (1943).

16 M. R. Krejsa and J. L. Koenig, RubberChem. Technol. 66, 376 (1993).

17 A. Y. Coran in “Science and Technology ofRubber,” F. R. Eirich, ed, Academic Press,New York, 1978.

18 C. H. Chen, E. A. Collins, J. R. Shelton, andJ. L. Koenig, Rubber Chem. Technol. 55,1221 (1982).

19 S. Henning, and R. Costin, Paper E, SpringTechnical Meeting, Rubber Division, ACS,San Antonio, TX, May 16-18, 2005.

20 C. J. Albisett, N. G. Fisher, M. J. Hogsed, andR. M. Joyce, J. Am. Chem. Soc. 78, 2637(1956).

21 W. K. Johnson, J. Org. Chem. 24, 864(1959).

22 T. Unukai and T. Kojima, J. Org. Chem. 31,1121 (1966).

The information in this bulletin is believed to be accurate, but all recommendations are made without warranty since the conditions of use are beyond Cray Valley Company'scontrol. The listed properties are illustrative only, and not product specifications. Cray Valley Company disclaims any liability in connection with the use of the information,

and does not warrant against infringement by reason of the use of its products in combination with other material or in any process.