Paper #72

OPTIMIZATION OF MIXING CONDITIONS FOR SILICA-REINFORCED

NATURAL RUBBER TIRE TREAD COMPOUNDS

By W. Kaewsakul1,2, K. Sahakaro1,2, W. K. Dierkes2, J. W. M. Noordermeer2*

1
Centre of Excellence in Natural Rubber Technology (CoE-NR), Department of Rubber

Technology and Polymer Science, Faculty of Science and Technology,

Prince of Songkla University, Pattani 94000, Thailand

2
Department of Elastomer Technology and Engineering, University of Twente, P.O. Box 217,

7500 AE Enschede, the Netherlands

Presented at the Fall 180th Technical Meeting of the Rubber Division of

the American Chemical Society, Inc.

Cleveland, OH

October 11-13, 2011

ISSN: 1547-1977

* Speaker

1
 ABSTRACT

The dump temperature and mixing interval between rubber, silica and silane coupling

agent for silica-filled natural rubber (NR) tire tread compounds using bis-triethoxysilylpropyl

tetrasulfide (TESPT) as silane were optimized. The dump temperature turns out to be the key

parameter governing the properties of the silica-filled NR compounds. The increase in

viscosity of the compounds by changing the dump temperature from 100-150oC indicates that

inevitably some cross- linking of NR occurs by sulfur contained in TESPT, simultaneous with

the silanization reaction between silica and silane. However, the viscosity decreases again

when dump temperatures above 150oC are applied, indicating a dominant occurrence of

degradation of the NR-molecules. The results are in good agreement with bound rubber

contents. The overall properties indicate that a dump temperature in the range of 135-150oC

and a silica-silane-rubber mixing interval of 10 minutes are the most appropriate mixing

conditions for silica-filled NR compounds with TESPT as coupling agent.

Keywords: silica; natural rubber; silane; reinforcement; mixing

2
 INTRODUCTION

Natural rubber (NR) is a renewable material which combines excellent mechanical and

dynamic properties, and its use covers a wide variety of applications, mostly in the form of

filler-reinforced vulcanizates. The commonly used reinforcing fillers in rubber compounds

are carbon black and silica. Replacement of carbon black by silica is growing, especially in

the tire industry since the introduction of the “Green Tire Technology” by Michelin in 1992.1

Silica significantly reduces rolling resistance and improves wet traction of tire tread

compounds compared to carbon black.2 However, a major drawback of using silica are

difficulties in processing which come from the polarity difference between silica and rubber.

The polar nature of silica negatively affects its compatibility with non-polar

elastomers. The silica surface is acidic with a number of silanol groups. Hence, it forms

strong hydrogen bonds or reacts with either basic or polar materials such as accelerators

resulting in the adsorption of curatives by silica. In addition, strong hydrogen-bonds between

the silica particles themselves result in tight silica aggregates and agglomerates causing poor

dispersion of silica in rubber compounds.3 A good dispersion of filler in a rubber compound is

very important for the mechanical and dynamic properties.3,4 The dispersion of silica-filled

rubber compounds is generally worse compared to that of carbon black-filled ones. Moreover,

when a large amount of silica is added, the viscosity substantially increases making the rubber

compounds more difficult to process, and causing wear of the processing equipment. These

difficulties can be overcome by use of silane coupling agents, which are able to combine

chemically with silica and rubber during the mixing and vulcanization stages, respectively.5-7

As a consequence, mixing of silica-filled compounds is basically different from that of carbon

black.

Mixing of rubber and silica with a silane coupling agent involves the reaction between

silanol groups on the silica and ethoxy or hydroxyl groups of the silane, i.e. the silanization

3
reaction. However, the efficient use of the silane as coupling agents in the silica-filled rubber

compounds is rather limited, caused by side-reactions. Various mixing-processing conditions

need to be optimized for silanized silica-filled compounds.8 The temperature window for

mixing silica compounds is rather limited by a too low silanization rate versus the risk of

scorch. High temperatures improve the silanization rate due to the temperature dependence of

the reaction5 and reduction of sterical hindrance of the silylpropyl groups of the coupling

agent by increased thermal mobility.9 Based on a study with SBR/BR compounds,10 the dump

temperature should be in the range of 145oC to 155oC to achieve good silanization and to

avoid pre-crosslinking. Furthermore, a mixing time of at least 10 minutes is needed for the

silanization reaction during the first mixing step.

Most of the work in this area was focused on synthetic rubbers, in particular SBR/BR

based-compounds for passenger tires. There are only a few studies dealing with silica-filled

NR. The optimization of mixing conditions for silanized silica-filled NR compounds was

studied by Wolff.11 The author concluded that an increase of mixing temperature and time for

silica modification with silane enhances the compatibility between silica and NR through a

chemical coupling bond of TESPT with the rubber. The overall properties are dependent on

the extent of this interaction. However, it was stated that the in-situ modification reaction of

silica with TESPT should not be considered as an optimum reaction, since the recipe was

composed of only 20 phr of silica in combination with 40 phr of carbon black.

The present work aims at optimizing the mixing conditions, i.e. mixing temperature

and silica-silane-rubber mixing interval, which may have an influence on mixing quality and

silanization efficiency of silica-filled natural rubber compounds. Mooney viscosity, cure

characteristics, filler-filler and rubber-filler interactions, as well as mechanical and dynamic

properties of the rubber compounds and vulcanizates are evaluated.

4
 EXPERIMENTAL

MATERIALS

The natural rubber used was Ribbed Smoked Sheet 3 (RSS3), locally produced in

Pattani (Thailand). The solution SBR (Buna VSL 5025-0 HM, Lanxess GmbH, Germany)

contains 25% styrene and 75% butadiene, of which 50% in the vinyl configuration. The

compounding ingredients were highly dispersible silica with cetyl trimethylammonium

bromide (CTAB) specific surface area12 171 m2/g (Ultrasil 7005, Evonik GmbH, Germany),

bis-triethoxysilylpropyltetrasulfide (TESPT) with sulfur content of approximately 22 wt.-%

(Zhenjiang Wholemark Fine Chemicals, China), treated distillate aromatic extract or TDAE

oil (Vivatec 500, Hansen & Rosenthal, Germany), zinc oxide (Global Chemical, Thailand),

stearic acid (Imperial Chemical, Thailand), polymerized 2,2,4-trimethyl-1,2-dihydroquinoline

or TMQ, diphenylguanidine or DPG and N-cyclohexyl-2-benzothiazolesulfenamide or CBS

(all from Flexsys, Belgium), and sulfur (Siam Chemical, Thailand).

COMPOUND PREPARATION

The compound recipe used for this study is shown in Table I. The amounts of TESPT

and DPG applied in this recipe were based on the CTAB specific surface area of the silica

according to Equation (1) and (2) as suggested by Guy et al.13:

Amount of TESPT (phr) = 0.00053 x Q x CTAB (1)

Amount of DPG (phr) = 0.00012 x Q x CTAB (2)

where Q is the silica content (phr) and CTAB is specific surface area of the silica used (m2/g).

The mixing was performed using an internal mixer with a mixing chamber of 500 cm3

(Chareon Tut Co., Ltd., Thailand). The mixer was operated at a fill factor of 70 % and a rotor

speed of 60 rpm. Initial temperature settings of the mixer were adjusted in the range of 50-

140oC (Table II). NR was initially masticated for 2 mins. Then, half of the silica and silane

were added and mixed for half the duration of the silica-silane-rubber mixing interval, i.e.:

5
2.5, 5, 7.5 and 10 mins, prior to adding the second half of silica and silane together with

TDAE oil. The mixing was then continued to the full intervals, i.e.: 5, 10, 15 and 20 mins.

Subsequently, the other ingredients: ZnO, Stearic acid, TMQ and DPG, except CBS and

sulfur, were added and mixed for 3 mins. The compounds were then dumped, sheeted out on

a two-roll mill, and kept overnight prior to incorporation of CBS and sulfur on a two-roll mill.

Mixer temperature settings and silica-silane-rubber mixing intervals prior to adding

other chemicals were varied following the experimental design as shown in Table II. Gum or

unfilled compounds were included in this study in order to determine the reinforcement

parameter (αF) from the cure characteristics; see below.

MOONEY VISCOSITY AND CURE CHARACTERISTICS

The compounds were tested for their Mooney viscosities by using a Mooney

viscometer (ViscTech+, TechPro, USA) at 100oC with large rotor according to ASTM D1646.

The value is represented as ML(1+4)100oC.

The cure characteristics were determined using a Moving Die Processability Tester or

MDPt (TechPro, USA). The increase in torque (S’) at a temperature of 150oC, a frequency of

0.833 Hz and 2.79 % strain was measured for 30 minutes. The optimum vulcanization time

(t90) was determined and used for press-curing of the samples.

CHARACTERIZATION OF FILLER-FILLER AND FILLER-RUBBER INTERACTIONS

Reinforcement parameter. − The reinforcement parameter (αF) was calculated by

using the data from the cure characteristics according to Equation (3)14:

′ − S min
S max ′ ⎛m ⎞
− 1 = α F ⎜⎜ F ⎟⎟ (3)
S ′ max − S ′ min
0 0
⎝ mP ⎠

where S’max – S’min = torque difference of the silica filled rubber; S’0max – S’0min = torque

difference of the corresponding gum or unfilled compound; mF/mP = filler loading, where mF

and mP correspond to the mass fractions of filler and polymer, respectively. The

6
reinforcement parameter, αF turns out to be a filler specific feature, independent of the cure

system, and closely related to the morphology of the filler in the compound.14

Flocculation rate constant. − Silica flocculation can be observed by following the

changes of the storage modulus of compounds at low strain amplitude during thermal

annealing.15 In order to calculate the flocculation kinetics of the silica, the storage modulus at

0.56 % strain amplitude can be monitored during heating at 100oC for 12 mins by using the

Moving Die Processability Tester, according to Mihara et al.16 The degree of flocculation (x)

can be expressed by the ratio of the Payne effect at time t and at infinite time:

G′(t ) − G′(i )
x= (4)
G′( f ) − G′(i )

where G′(t) is the storage modulus at 0.56% strain amplitude after heating time t; G′(i) is the

storage modulus after preheating for 1 minute; G′(f) is the storage modulus at infinite time,

taken as 12 minutes in order to reduce measuring time.

As mentioned by Bohm and Nguyen,15 silica flocculation can be experimentally

described as a first order reaction for which the kinetics can be described by a simple model

equation. Mihara et al.16 take this information and apply the following equation to calculate

the flocculation rate constant (ka).

ln(1 − x1 ) − ln(1 − x2 )
ka = [s-1] (5)
t2 − t1

where ka is the rate constant; t is the heating time; x is the degree of flocculation; 1 and 2 refer

to different heating times t.

Payne effect. − Storage shear moduli (G′) of compounds cured to their t95 at 150oC

were measured under shear deformation using a Rubber Process Analyzer (RPA, Alpha

Technologies, USA). A strain sweep test was done in the range of 0.56 - 100 % strain at 0.50

Hz and 100oC

7
 Bound rubber content. − Bound rubber measurements were carried out both with and

without ammonia treatment. The ammonia treatment was done in order to cleave the physical

linkages between rubber and silica, so that the true chemically bound rubber could be

determined.17 The procedure adopted for the bound rubber measurement was as follows:

- 0.2 g of sample, i.e. uncured masterbatch (without curatives), was put into a metal

cage and immersed in 20 ml of toluene for 72 h at room temperature. Toluene was renewed

every 24 h.

- The sample was removed from the toluene, dried at 105oC for 24 h.

- The sample was immersed again in 20 ml of toluene for 72 h at room temperature in

either a normal or an ammonia atmosphere in order to cleave the physical linkages. Toluene

was renewed every 24 h.

- The sample was dried at 105oC for 24 h and weighed.

The bound rubber content was calculated according to Equation (6)17:

W fg − W f
Bound rubber content (%) = (6)
Wp

where Wfg is the weight of silica with the bound rubber attached; Wf is the weight of silica in

the specimen; Wp is the weight of polymer in the specimen.

VULCANIZATION AND MEASUREMENT OF TENSILE AND TEAR PROPERTIES

The compounds were press vulcanized at 150oC to t90. The vulcanized sheets having a

thickness of about 2 mm were die-cut to dumbbell shaped specimens for tensile tests and to

angle shape specimens (die type C) for tear tests. The tests were performed at a crosshead

speed of 500 mm/mins according to ASTM D412 and ASTM D624, respectively, using a

Hounsfield Tensile Tester. The mean values of tensile and tear properties taken from five

specimens are reported.

8
 DETERMINATION OF DYNAMIC MECHANICAL PROPERTIES

The vulcanizates cured to their t95 at 150oC in the RPA were tested for dynamic

mechanical properties using a frequency sweep test. The frequencies were varied in a range of

0.05 - 33 Hz at a fixed strain and temperature of 3.49 % and 60oC, respectively.

CURE CHARACTERISTICS OF NR AND SBR COMPOUNDS IN THE PRESENCE OF

TESPT AND DPG (WITHOUT CBS AND SULFUR)

In order to explain the difference in change of Mooney viscosities with dump

temperature of silica-filled NR compounds in comparison with previous experience with SBR

compounds,10 two additional compounds were prepared. These compounds were composed

of rubber with only TESPT and DPG at amounts of 5.0 and 1.5 phr, respectively. The

compounds were prepared using a two-roll mill; rubber was milled for 6 mins, then DPG was

added and mixed for 2 mins, and TESPT was subsequently incorporated and mixed for 2

mins. The compounds were sheeted out to a thickness of about 0.5 cm, and kept overnight

prior to testing.

The cure characteristics of these compounds were analyzed by using the RPA with a

frequency of 0.833 Hz and strain 2.79 % for 12 mins. The testing temperature was varied

from 120-180oC with an increment of 10oC for each separate test.

RESULTS AND DISCUSSION

PROCESSING PROPERTIES

Mixing temperatures and torques. − The initial mixer temperatures settings and silica-

silane-rubber mixing intervals prior to the addition of other chemicals were varied so that the

influence of different dump (end) temperatures on the silica-filled NR compounds could be

evaluated. The dump temperatures as function of initial temperature settings and silica-silane-

rubber mixing intervals of the silica-filled NR compounds, in comparison with those of gum

compounds, are shown in Figure 1 and 2, respectively. By increasing the initial mixer

9
temperature setting (Figure 1), the dump temperatures of both filled and gum compounds

increase linearly. The filled compound shows a significantly higher dump temperature than

that of the gum compound, attributed to the influence of hydrodynamic effects leading to

higher shear forces generated during mixing. By choosing the initial mixer temperature

settings in the range of 50 – 140oC, dump temperatures in the range of 90 – 170oC could be

obtained. Considering the different silica-silane-rubber mixing intervals of filled and gum

compounds (Figure 2), longer periods clearly give higher dump temperatures for both

compounds, and the filled compounds, have again higher dump temperatures than the gum

compounds.

Figure 3 shows typical torque curves during mixing of silica-filled NR compounds

prepared by adjusting the initial temperature setting in order to reach different dump

temperatures, as they are indicated in the figure. The relation between dump temperatures and

initial mixer temperature settings are given in Figure 4. The torque curves show quite

different behaviors at different mixing times. Higher mixer temperature settings correspond to

lower mixing torques in the range of 0 – 600 s mixing time, as depicted also in Figure 4: the

torque average at 300 s substantially decreases with increasing mixer temperature setting.

This is commonly the result of lower mixture viscosity due to the higher temperature.

However, for mixing times larger than 600 s, a crossover of the curves can be observed in

Figure 3. Figure 4 shows that the mixing torque at 720 s increases with increasing mixer

temperature setting to reach a maximum for a temperature setting of 110oC. At the end of

mixing, i.e. 890 s, the torque also displays such a maximum. There are 3 possible reactions

occurring during mixing: 1) silanization, 2) premature crosslinking due to sulfur in TESPT,

and 3) rubber degradation. At dump temperatures lower than 150oC, silanization and some

premature crosslinking are dominant so that the mixing torque increases at the later stage of

mixing. However, at very high initial mixer temperature settings and consequent high

10
compound dump temperatures, degradation of the natural rubber becomes dominant resulting

in a decrease of torque.

Mooney viscosity of the compounds. − The Mooney viscosities of the silica-filled NR

compounds prepared with the different mixer temperatures and silica-silane-rubber mixing

intervals in comparison with those of the gum compounds are shown in Figure 5(a) and 5(b),

respectively. The filled compounds show much higher viscosities than the gum compounds,

attributed to the hydrodynamic effect, filler-filler and filler-rubber interactions, and possibly

some premature crosslinking of the rubber caused by sulfur in the TESPT in presence of DPG

accelerator. For the silica-filled compounds, the Mooney viscosities initially increase with

higher dump temperature, upon which the values gradually drop off when the dump

temperature is over 150oC. This result corresponds with the change in mixing torque average

(Figure 4). Furthermore, the silica-filled compounds mixed with longer silica-silane-rubber

mixing intervals, i.e. longer overall mixing times give higher Mooney viscosities, given an

initial mixer temperature setting, as shown in Figure 5(b). On the other hand, the Mooney

viscosities of the gum compounds show composite trends, to decrease with higher mixer

temperature settings and with prolonged silica-silane-rubber mixing intervals. This is while a

longer mixing time between silica, silane and NR results in a higher dump temperature, as

displayed in Figure 2, giving an increase in degree of silanization of the silica and possibly

some premature crosslinking of the rubber. Whereas the gum compound does not involve

silica-silane-rubber reactions, the higher mixing temperatures can only lead to softening and

destruction of the NR molecules. For these reasons, it is important to reconfirm that was the

mixer temperature setting has more of an effect than the time allocated for silanization of the

silica-filled NR system, as already reported by Wolff’s.11

Cure characteristic of the compounds. − The cure characteristics of the silica-filled

NR compounds: scorch time (tc10), cure time (t90), cure rate index (CRI) as well as rheometer

11
minimum torque (S’min), are shown in Figure 6. With higher dump temperature the scorch

time and cure rate index of the silica-filled compounds slightly increase, while the optimum

cure time shows no significant change. The longer scorch time enhances processing safety

while the higher cure rate leads to a similar optimum cure time, i.e. better scorch safety with

the same curing cycle. Moreover, the rheometer minimum torques of the compounds reach a

maximum value at a dump temperature in the range of 135 – 150 oC, which corresponds with

a similar maximum in the mixing torque average at the end of compounding, i.e. 890 s of

mixing time (Figure 4), and in the Mooney viscosity (Figure 5).

SILICA DISPERSION

Reinforcement parameter of the compounds. − Basically, a lower reinforcement

parameter indicates a better dispersion of a filler in a polymer.14 In Figure 7, the

reinforcement parameters of the silica-filled NR compounds are shown as function of dump

temperature. The compounds prepared with higher dump temperature show a lower

reinforcement parameter, indicating a better dispersion. However, a minimum αF is observed

in the dump temperature range of 135 – 170 oC, which agrees with the highest mixing torque

average, Mooney viscosity, and minimum cure torque as displayed in Figures 4, 5 and 6(b),

respectively.

Filler flocculation of the compounds. − Dispersion stability of the filler particles or

filler aggregates in the rubber matrix influences the morphology of the filled rubber

compounds. Reinforcing fillers, i.e. carbon black and silica, are able to flocculate caused by

the poor compatibility with the rubbers and their self-association, leading to a change in

morphology and consequently deterioration in the physical properties of the vulcanizates.15 It

has been documented before that silica causes a greater degree of flocculation than carbon

black, while silica is more polar and has a stronger self-association tendency through

hydrogen bonding when compared to carbon black.18 It is worth to mention that the

12
flocculation process can occur during compound storage as well as directly prior to the onset

of vulcanization in the absence of shear.15 The kinetics of the flocculation process can be

monitored by determination of the storage modulus during thermal annealing of a compound

under a condition that represents vulcanization. The flocculation rate constant (ka) of the

silica-filled NR compounds as a function of dump temperature is shown in Figure 8. This

value indicates how fast the silica flocculation process develops just before crosslinking, so

that the parameter can be considered to represent silica dispersion stability as well as the

degree of dispersion in silica-filled rubber compounds. Generally, a lower ka value means a

smaller tendency towards agglomeration in the rubber matrix. In other words, a lesser amount

of silica flocculation corresponds to a better dispersion stability and finer morphology,

presuming that silica is well dispersed during mixing. As expected, mixing till elevated dump

temperature and at longer mixing times enhances the dispersion stability of silica in the NR

compounds, indicated by a decrease of the flocculation rate constant. This is in agreement

with the degree of dispersion indicated by the reinforcement parameter (Figure 7).

BOUND RUBBER CONTENT OF THE COMPOUNDS

The bound rubber contents of the silica-filled compounds were evaluated under two

conditions, i.e. with and without ammonia treatments. The ammonia treatment was used to

cleave the physical linkages formed by physical adsorption, so that only the amount of

chemically bound rubber, i.e. strong filler-rubber interactions and covalent bonds, were

determined. Figure 9 shows the bound rubber contents as a function of dump temperature of

the compounds measured without and with ammonia treatment. It is evident that the bound

rubber content measured with ammonia treatment is lower than that measured without

ammonia treatment. With increasing dump temperature and larger silica-silane-rubber mixing

intervals, both total and chemically bound rubber contents in the compounds significantly

increase, but level off when the dump temperature exceeds 135oC. Considering the physical

13
linkages, obtained by subtracting the chemically bound rubber from the total bound rubber

content (Figure 9b), the physically bound rubber of the compounds prepared at dump

temperatures over 150oC starts to decrease, whereas the chemically bound rubber content

remains unchanged. The decrease of physically bound rubber content might be attributed to

the lower molecular weight of the NR-molecules after mixing. Basically, natural rubber

molecules can easily be broken down by excessively high temperatures and shearing action,

because it has reactive double bonds in every repeating unit along the rubber chain. Hence,

under excessive mixing conditions, the NR chains are broken down leading to lower physical

linkages in the compound. This phenomenon corresponds well with the decrease of Mooney

viscosities at very high dump temperatures, as previously shown in Figure 5. Interestingly, the

chemically bound rubber does not show any such dependence.

DYNAMIC PROPERTIES OF THE COMPOUNDS AND VULCANIZATES

The Payne effect, i.e. the storage modulus difference between small (0.56 %) and large

(100 %) strains, analyzed with a first and second sweep, and tan δ at 60oC of the silica-filled

NR vulcanizates are shown as a function of dump temperature in Figure 10. By increasing the

dump temperature and mixing time, the Payne effect as a measure of filler-filler interaction

decreases, even in the second sweep test (Figure 10a). However, for dump temperatures

above 135oC, there is no further change of the Payne effect. Regarding the tan δ at 60oC of

the vulcanizates (Figure 10b), which is commonly used as an indication of rolling resistance

of tires for which a lower value indicates a lower rolling resistance or less fuel consumption,2

the compounds prepared with higher dump temperature and longer silica-silane-rubber mixing

interval have lower tan δ at 60oC. The reduction of tan δ at 60oC levels off for dump

temperatures above 135oC.

14
 MECHANICAL PROPERTIES OF THE VULCANIZATES

Figure 11 shows the tensile strength, elongation at break, reinforcement index (i.e. the

ratio of the moduli at 300% and 100% elongations or M300/M100) and tear resistance of the

silica-filled NR vulcanizates prepared with varying dump temperatures and silica-silane-

rubber mixing intervals. With varying dump temperatures and mixing times the elongation at

break of the NR vulcanizates remains more or less the same, while the tensile strength slightly

decreases with increasing dump temperature but shows no dependence on mixing times. The

decrease of tensile strength, reinforcement index and tear resistance when the dump

temperature exceeds 150oC is caused by the inevitable NR degradation under high

temperature and shearing force. The degradation of NR with and without antioxidant after

mixing in an internal mixer at 160oC and 180oC for 10 mins. was clearly demonstrated by

Narathichat et al.19 Considering the reinforcement index, an increasing dump temperature and

silica-silane-rubber mixing interval lead to an optimum at 135oC; thereafter the values

gradually decrease when the dump temperature exceeds 150oC. A change of tear resistance of

the vulcanizates with varying dump temperatures and mixing times displays the same sort of

dependence as that of the reinforcement index.

COMPETITIVE REACTIONS BETWEEN SILANIZATION, PREMATURE

CROSSLINKING AND DEGRADATION IN TESPT-SILICA-FILLED NR COMPOUNDS

With respect to the change of the NR compound viscosities with dump temperature, as

reflected in the final mixing torques, Mooney viscosities and rheometer minimum torques in

Figure 4, 5 and 6(b) respectively, each separate property reveals the same trend. With

increasing mixing temperatures, the compound viscosities initially increase, then reach a

maximum for a dump temperature range of 135-150oC. For dump temperatures above 150oC,

the compound viscosity drops. This phenomenon is different from the work reported by

Wolff,11 in which a steady decrease of viscosity with increasing dump temperature from 110

15
to 170oC for silica-filled NR compounds was observed. However, it should be noted that only

20 phr of silica was used in combination with 40 phr of carbon black in that work. In the case

of silica-TESPT filled SBR/BR compounds, Reuvekamp et al.10 reported a rather constant

viscosity with increasing dump temperatures from 120-150oC, but an increased Mooney

viscosity when dump temperatures above 150oC were obtained, caused by premature scorch

of the rubber by sulfur contained in the TESPT. It is clear that the NR and SBR/BR

compounds display remarkably opposite trends of compound viscosities with varying mixing

conditions.

The mixing torque can be used as an indication for what is happening. During mixing

times of 120-720 s in Figure 3 which represent the first and second incorporations of silica,

silane, and processing oil, the decrease in the mixing torques with increasing temperature,

especially after the first incorporation, is due to several factors. These include softening of

the rubber at high temperature, and the breakdown of NR chains and silica structures by the

mixing temperature due to mechanical energy input, in combination with the hydrophobation

effect of the TESPT silica surface modification. However, interestingly, a crossover of mixing

torques is noticed at the final stage of mixing. It demonstrates that, besides the silanization

reaction between silica and silane, a crosslinking or scorch reaction of NR with sulfur released

from TESPT is simultaneously taking place, i.e. a bond between silica-silane-rubber and

rubber-rubber is created in the compound and consequently causes a rise in the compound

viscosity. In order to prove this, the cure characteristic of NR and TESPT was investigated in

comparison with that of SBR as shown in Figure 12. The test was performed with compounds

without CBS and elemental sulfur. It is clearly seen that NR readily reacts with TESPT at a

temperature as low as 120oC and the reaction rate increases rapidly at higher temperatures, as

seen from a change of the cure curves. SBR shows a different behavior since it starts to react

with TESPT at a higher temperature, i.e. 150oC. The rheometer torque differences, commonly

16
used to indicate the extent of crosslinking of both compounds are shown in Figure 13. The

value for the NR+TESPT compound substantially increases from a test temperature of 120oC

up to 150oC, and then gradually decreases at higher temperatures: a sort of reversion. On the

other hand, the torque difference for the SBR+TESPT compound remains constant till a

temperature of 140oC, and then increases rapidly when the cure temperature exceeds 150oC.

These results indicate that the viscosity of silica-TESPT filled rubber compounds significantly

depends on a premature crosslinking reaction which takes place during the mixing process

due to available sulfur in TESPT, in addition to the silanization reaction and the thermo-

mechanical shearing action. In this context it is worth to mention that TESPT was first

developed as a sulfur-cure accelerator with sulfur-donor properties, before it became

commonly accepted as a coupling agent.20,21 Therefore, the increase of the silica-filled NR

compound viscosities for dump temperatures in the range of 100-145oC (Figure 5) is to a

large extent due to the silica-silane-NR coupling reaction that occurs during the mixing

process, and this reaction reaches its maximum at a dump temperature around 135-150oC.

This premature crosslinking reaction dominates over the breakdown of rubber chains and

silica structures.

As seen from Figure 5, the viscosities of the silanized silica-filled NR compounds tend

to decrease again at dump temperatures above 150oC. This may be attributed to an optimum

degree of silica-rubber and rubber-rubber interactions/reactions induced by the TESPT.

Furthermore, degradation of rubber chains and/or polysulfide linkages becomes prominent.

This observation can be also confirmed by the change of bound rubber content in the

compounds (Figure 9). The chemically bound rubber reaches a plateau at a dump temperature

of approximately 135oC, indicating that the silica-silane-rubber and rubber-rubber

interactions/reactions remain intact at higher temperature and shear force. Meanwhile, the

physically bound rubber content in the compounds prepared at dump temperatures over 150oC

17
decreases somewhat, suggesting a reduction of chain entanglements or the physical network

of rubber molecules due to a chain shortening by the degradation reaction.

The mixing of silica-silane and NR can not avoid this premature crosslinking when

TESPT is used as coupling agent, as the NR starts to react with sulfur present in TESPT at a

temperature as low as 120oC (Figure 13). The increased Mooney viscosity is normally an

undesired processing property since it leads to processing difficulties in extrusion and

calendering. Surprisingly, this phenomenon does not have an adverse effect on other

properties of the compounds and vulcanizates. As illustrated in Figure 6, even better cure

characteristics were obtained for compounds prepared with higher dump temperature, as

indicated by the faster cure rate and longer scorch time. The improvement in cure behavior

could be the result of better silanization so that all silica surfaces are well covered with silane

resulting in less acidity and polarity, which in turn prevents a reduction of cure retardation

and accelerator adsorption. Furthermore, when DPG is used in the recipe to enhance the

silanization reaction22 as well as to act as a secondary accelerator, the higher level of

silanization could also mean that a smaller portion of CBS is adsorbed on the still uncovered

silica surface, so that this accelerator is more efficiently used for the curing. The improvement

of the mechanical properties with increasing dump temperature to the optimum point, i.e. in

the range of 135-150oC, especially in terms of reinforcement index (M300/M100) and tear

resistance (Figure 11) is accounted for by a better silica dispersion in the NR as indicated by

reinforcement parameter (αF), flocculation rate constant (ka) and Payne effect of the

compounds in Figures 7, 8 and 10, respectively. This also corresponds with the higher bound

rubber content which indicates a better filler-rubber interaction. The lightly crosslinked

network formed at the later stage of mixing restricts the movement of dispersed silica

particles. As a consequence, the re-agglomeration of silica is reduced, as depicted by a

decrease of the flocculation rate constant (Figure 8). The results of this present work show a

18
good agreement with the work reported by Lin et al.,23 in which the compounds comprising of

TESPT showed a reduction of silica flocculation when they were mixed at higher dump

temperature. Based on the present results, re-agglomeration of silica/TESPT-filled NR

compounds can be effectively suppressed when sufficiently high mixing temperature is used.

This is not only due to better silica dispersion, higher degree of hydrophobation, greater

extent of polymer-filler interaction, but also due to the lightly crosslinked network as a result

of a small amount of crosslinking during mixing.

The tan δ at 60oC, a measure of rolling resistance of tires based on silica-filled NR, is

shown to improve when the compound dump temperatures are increased and the optimum

temperature is also in a range of 135-150oC, in agreement with all other properties.

Based on our investigations on the reinforcement parameter, flocculation rate constant,

bound rubber content, Payne effect, loss tangent at 60oC (indication for rolling resistance),

and mechanical properties, the optimal mixing conditions for silica-filled NR with TESPT as

a coupling agent are a dump temperature in the range of 135-150oC, and 10 mins mixing

interval of silica-silane-rubber after mastication and before the addition of the other

ingredients: ZnO, Stearic acid, TMQ and DPG. The decrease of reinforcement parameter,

flocculation rate constant and Payne effect combined with the increase of bound rubber

content can be taken as a proof for proper silanization. However, it is worth to emphasize that

the change of Mooney viscosity with dump temperature is different for the NR and SBR cases

when TESPT is used. The Mooney viscosity of silica-filled NR compound increases due to

sulfur in the TESPT to the maximum and thereafter decreases at higher temperatures as NR

degradation starts to take place.

CONCLUSIONS

Varying initial mixer temperature settings from 50-140oC and mixing intervals from 5

to 20 minutes for the silica-filled NR compounds lead to the dump (end) temperatures in a

19
range of 90-170oC. The properties of both compounds and vulcanizates show a strong

dependence on dump temperature since the silanization reaction and silica dispersion are key

parameters in this system. The optimal mixing conditions for the silica-filled NR compound

with TESPT as a coupling agent are a dump temperature in the range of 135-150oC, and 10

mins mixing interval of silica-silane-rubber prior to addition of the other ingredients: ZnO,

Stearic acid, TMQ and DPG. Mixing performed till dump temperatures above 150oC leads to

a decrease in vulcanizate properties as a result of NR degradation. During mixing of silica-

filled NR compounds with TESPT as coupling agent, it is evident that in addition to the

silanization reaction, premature crosslinking reactions take place since NR can start to react

with sulfur coming from TESPT molecules at a temperature as low as 120oC. However, this

phenomenon does not have an adverse effect on the vulcanizate properties.

ACKNOWLEDGEMENTS

The authors acknowledge the financial support from the Netherlands Natural Rubber

Foundation (Rubber Stichting), and would like to thank N. Gevers from Apollo Vredestein

Cie. and S.S. Sarkawi for fruitful discussions.

REFERENCES
1
R. Rauline (to Compagnie Generale des Etablissements Michelin-Michelin & Cie),

U.S.Patent 5227425, July 13, 1993 ; E.P. 0501227A1, Sept.2, 1992.

2
H.-D. Luginsland, W. Niedermeier, Rubber World. 228, 34 (2003).
3
A.I. Medalia, RUBBER CHEM. TECHNOL. 47, 411 (1974).
4
M.-J. Wang, RUBBER CHEM. TECHNOL. 72, 430 (1999).
5
U. Görl, A. Hunsche, A. Müller, H.G. Koban, RUBBER CHEM. TECHNOL. 70, 608

(1997).
6
H.-D. Luginsland, J. Frohlich, A. Wehmeier, RUBBER CHEM. TECHNOL. 75, 563 (2002).

20
7
J.W. ten Brinke, P.J. van Swaaij, L.A.E.M. Reuvekamp, J.W.M. Noordermeer, Kautsch.

Gummi Kunstst. 55, 244 (2002).

8
W.K. Dierkes, J.W.M. Noordermeer, K. Kelting, A. Limper, Rubber World. 229, 6 (2004).
9
J.E. Mark, B. Erman, F.R. Eirich, The Science and Technology of Rubber, 3rd edition,

Elsevier Academic Press, Burlington 2005, p. 424.

10
L.A.E.M. Reuvekamp, J.W. ten Brinke, P.J. van Swaaij, J.W.M. Noordermeer, Kautsch.

Gummi Kunstst. 55, 41 (2002).

11
S. Wolff, RUBBER CHEM. TECHNOL. 55, 967 (1982).
12
O. Stenzel, H.-D. Luginsland, S. Uhrlandt, A. Wehmeier (to Degussa AG). U.S. Patent

7608234, Oct. 27, 2009.

13
L. Guy, S. Daudey, P. Cochet, Y. Bomal, Kautsch. Gummi Kunstst. 62, 383 (2009).
14
S. Wolff, Kautsch. Gummi Kunstst. 23, 7 (1970).
15
G.G.A. Bohm, M.N. Nguyen, J. Appl. Polym. Sci. 55, 1041 (1995).
16
S. Mihara, R.N. Datta, J.W.M. Noordermeer, RUBBER CHEM. TECHNOL. 82, 524

(2009).
17
S. Wolff, M.J. Wang, E.H. Tan, RUBBER CHEM. TECHNOL. 66, 163 (1993).
18
M.-J. Wang, S. Wolff, J.-B. Donnet, RUBBER CHEM. TECHNOL. 64, 714 (1991).
19
M. Narathichat, K. Sahakaro, C. Nakason, J. Appl. Polym. Sci. 115, 1702 (2010).
20
F. Thum and S. Wolff, Kautsch. Gummi Kunstst. 28, 733 (1975).
21
S. Wolff, Kautsch. Gummi Kunstst. 32, 760 (1979).
22
C. Penot (to Michelin Recherche et Technique S.A.), U.S. Patent 6951897, Oct. 4, 2005.
23
C.J. Lin, W.L. Hergenrother, E. Alexanian, G.G.A. Bohm, RUBBER CHEM. TECHNOL.

75, 865 (2002).

21
 TABLE I

COMPOUND FORMULATION

Ingredients Amount (phr)

NR (RSS3) 100.0
Silica (Ultrasil 7005) 55.0
Silane (TESPT) 5.0
Process oil (TDAE) 8.0
Zinc oxide 3.0
Stearic acid 1.0
TMQ 1.0
DPG 1.1
CBS 1.5
Sulfur 1.5

22
 TABLE II

EXPERIMENTAL DESIGN FOR OPTIMIZING MIXER TEMPERATURE AND MIXING

INTERVAL FOR SILICA-FILLED NR COMPOUNDS

Mixer temperature Mixing interval (mins.)

setting (oC) 5 10 15 20
50 x x x x
60 x x x x
70 x x x x
80 - x - -
90 - x - -
100 - x - -
110 - x - -
120 - x - -
130 - x - -
140 - x - -

x is assigned to the conditions for which both gum and filled compounds were prepared.

23
 200

180

160 Filled compound

Dump temperature ( C)
o 140

120

100

80 Gum compound

60

40

20

0
40 50 60 70 80 90 100 110 120 130 140 150
o
Initial mixer temperature setting ( C)

FIG. 1. — Dump temperature as a function of initial mixer temperature setting for gum and

silica-filled NR compounds, with 10 mins. mixing interval for silica+silane and rubber.

24
 140

120

Dump temperature ( C)
o
100

80

60

40
Gum, 50 Gum, 60 Gum, 70
20
Filled, 50 Filled, 60 Filled, 70

0
0 5 10 15 20 25
Silica-silane-rubber mixing interval (mins.)

FIG. 2. — Dump temperature as a function of silica-silane-rubber mixing interval for gum

and silica-filled NR compounds. The initial mixer temperature settings were varied from 50 to

70oC, as indicated.

25
FIG. 3. — Mixing torques for silica-filled NR compounds prepared with various mixer

temperature settings, as indicated for each line. The silica-silane-rubber total mixing interval

was 10 mins.

26
 180 260

240
160
220
140 200
Mixing torque average (N.m)

Dump temperature ( C)
180

o
120
160
100 140

80 120

100
60
80
300 s
40 60
720 s
40
20 820 s
20

0 0
40 60 80 100 120 140
o
Initial mixer temperature setting ( C)

FIG. 4. — Correlation between the mixer temperature setting, mixer torque and dump

temperature of silica-filled NR compounds at 300, 720 and 890 seconds when a 10 mins.

silica-silane-rubber mixing interval was used.

27
 90 80

80 70
70oC
70 60oC
60
50oC
60

ML(1+4)100 C
ML(1+4)100 C

50 Filled compound

o
o

Filled compound
50
40 Gum compound
40
Gum compound 30 50oC
30 60oC
20 70oC
20
10
10 (b)
(a)
0
0 0 5 10 15 20 25
80 95 110 125 140 155 170 185
o
Dump temperature ( C) Silica-silane-rubber mixing interval
(mins.)

FIG. 5. — Effect of (a) dump temperature and (b) silica-silane-rubber mixing interval on

Mooney viscosity of gum and silica-filled NR compounds; parameter: initial temperature

settings of mixer.

28
 10 26 1.2

Scorch, tc10 and optimum cure time, t90 (mins. )

Rheometer minimum torque, S'm in (dN.m)

24
9
22 1.0
8
CRI 20

Cure rate index, CRI

7 18
0.8
6 16
14
5 t90 0.6
12
4 10
tc10 0.4
3 8
6
2
4 0.2
1
(a) 2 (b)
0 0 0.0
90 110 130 150 170 190 90 110 130 150 170 190
o o
Dump temperature ( C) Dump temperature ( C)

FIG. 6. — Effect of dump temperature on (a) scorch time, optimum cure time, cure rate index

and (b) rheometer minimum torque of silica-filled NR compounds.

29
 14

12

Reinforcement parameter, α F
10

0
80 90 100 110 120 130 140 150 160 170 180
o
Dump temperature ( C)

FIG. 7. — Effect of dump temperature on reinforcement parameter of silica-filled NR

compounds prepared with various silica-silane-rubber mixing intervals.

30
 5.0E-03

Flocculation rate constant, ka (s )

-1
4.0E-03

3.0E-03

2.0E-03

1.0E-03

0.0E+00
80 90 100 110 120 130 140 150 160 170 180
o
Dump temperature ( C)

FIG. 8. — Effect of dump temperature on flocculation rate constant of silica-filled NR

compounds prepared with various silica-silane-rubber mixing intervals.

31
 80 25
Untreated with NH3

Physically bound rubber content (%)

70

Bound rubber content (%)

20
60

50 15

40

10
30
Treated with NH3

20
5
10
(a) (b)
0 0
85 100 115 130 145 160 175 85 100 115 130 145 160 175
o o
Dump temperature ( C) Dump temperature ( C)

FIG. 9. — Influence of dump temperature on bound rubber contents of silica-filled NR

compounds: (a) total (untreated with NH3) and chemically bound rubber (treated with NH3);

and (b) physically bound rubber content.

32
 2.5 1.8E-01

1.6E-01
o
1 sweep
2.0
1.4E-01
[G'(0.56) - G'(100)] (MPa)
1.2E-01

tan δ at 60 C
1.5

o
1.0E-01

8.0E-02
1.0
6.0E-02

o
2 sweep 4.0E-02
0.5

2.0E-02
(a) (b)

0.0 0.0E+00
80 100 120 140 160 180 80 100 120 140 160 180
o o
Dump temperature ( C) Dump temperature ( C)

FIG. 10. — Influence of dump temperature on dynamic properties: (a) Payne effect; and (b)

tan δ at 60oC (frequency 10 Hz) of silica-filled NR vulcanizates prepared with various silica-

silane-rubber mixing intervals.

33
 35 700

Elongation at break (%)

Tensile strength (MPa)
30 600
25 500
20 400
15 300
10 200
5 (a) 100
(b)
0
0
80 100 120 140 160 180 80 100 120 140 160 180
Dump temperature (oC) o
Dump temperature ( C)
6 180

Tear resistance (N/mm)

160
Reinforcement index

5
140
4 120
100
3
80
2 60
40
1 20
(c) (d)
0 0
80 100 120 140 160 180 80 100 120 140 160 180
o
Dump temperature ( C) o
Dump temperature ( C)

FIG. 11. — Effect of dump temperature and silica-silane-rubber mixing interval on

mechanical properties of silica-filled NR vulcanizates: (a) tensile strength; (b) elongation at

break; (c) reinforcement index; and (d) tear resistance.

34
 0.40 1.2
120 oC
130 oC
0.35
1.0 140 oC
o
150 C
0.30 o
160 C
o
0.8 170 C

Torque (dN.m)
Torque (dN.m)

0.25 o
180 C

0.20 0.6

0.15 120 oC
130 oC 0.4
140 oC
0.10 150 oC
160 oC 0.2
0.05 170 oC
(a) 180 oC (b)
0.00 0.0
0 2 4 6 8 10 12 0 2 4 6 8 10 12
Time (mins.) Time (mins.)

FIG. 12. — Cure characteristics of (a) NR and (b) SBR compounds, in the presence of TESPT

(5.0 phr) and DPG (1.5 phr) at different cure temperatures.

35
 0.25 1.0

Torque difference of SBR + TESPT (dN.m)

Torque difference of NR + TESPT (dN.m)
0.20 0.8

0.15 0.6

0.10 0.4

0.05 0.2

0.00 0.0
110 120 130 140 150 160 170 180 190
o
Temperature ( C)

FIG. 13. — Rheometer torque difference of NR and SBR compounds as a function of cure

temperature in the presence of TESPT (5.0 phr) and DPG (1.5 phr).

36