Polymer 55 (2014) 6426e6434

Contents lists available at ScienceDirect

Polymer
journal homepage: www.elsevier.com/locate/polymer

Vulcanization kinetics of nano-silica filled styrene butadiene rubber

Seyed Mostaffa Hosseini, Mehdi Razzaghi-Kashani*
Polymer Engineering Department, Faculty of Chemical Engineering, Tarbiat Modares University, P. O. Box: 14115-114, Tehran 1411713116,
I.R. Iran

a r t i c l e i n f o a b s t r a c t

Article history: It was shown that the physical filler-polymer and fillerefiller interactions, apart from the filler surface
Received 5 August 2014 chemistry, has a substantial role in controlling the vulcanization kinetics of styrene butadiene rubber
Received in revised form filled with nano-silica in a sulfur vulcanization system. Kinetic studies by the oscillating disc rheometer,
29 September 2014
differential scanning calorimeter, and swelling tests revealed that the vulcanization rate goes through a
Accepted 30 September 2014
Available online 15 October 2014
maximum as loading of silica increases, but conversion in crosslinking continuously decreases as the
amount of silica increases. The effect of silica loadings on the vulcanization reactions was linked to the
immobilization of rubber chains around particles as well as in a polymer-mediated filler network, which
Keywords:
Rubber
were differentiated by the nonlinear viscoelastic behavior of rubber vulcanizates. By surface modification
Vulcanization kinetics of nano-silica, the accelerating/decelerating effects of nano-silica on the vulcanization reactions were
Nano-silica altered corresponding to the non-linear viscoelastic behavior of the vulcanizates. Therefore, a mecha-
nism was proposed which correlates vulcanization kinetics of rubber to the dynamics of chains influ-
enced by the reinforcing fillers.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction Kinetics of vulcanization for unfilled rubbers had been inten-

sively investigated and well established for decades [5]. It is also
Elastomers need to be reinforced with significant quantities well known that the presence of reinforcing fillers can influence
(about 25 vol.%) of reinforcing fillers such as carbon black and/or vulcanization kinetics and crosslink density of rubber vulcanizates
silica to achieve the mechanical properties needed for high per- [4,6e8]. However, impacts of fillers on the vulcanization kinetics of
formance applications such as tires and other engineering prod- rubber have mostly been studied by focusing on the filler surface
ucts. Silanized silica has gained wide attention in the Green-Tire chemistry. For instance, it has been concluded that the accelerating
technology because it provides lower rolling resistance at equal effects of carbon black is due to a large variety of oxygen containing
wear resistance and even better wet grip for the tire tread vulca- functional groups, such as carboxyl, hydroxyl, phenol, lactones,
nizates than carbon blacks [1]. Specific mechanical and dynamic quinones, ketones, aldehydes, and hydroperoxides [9,10]. Whereas,
properties are expected from such complex vulcanizates, and the silica surface is occupied by acidic hydroxyl, siloxane and silanol
chemical crosslink network of rubber has indispensable contribu- groups, which adsorb basic accelerators by the surface hydrogen
tions in these properties. Contributions of the rubber network to bonds, deactivate them, and slow down the rate and degree of
the elastic modulus of rubber vulcanizates depend on the crosslink vulcanization [11e14]. Therefore, modification of silica surface by
density as well as the nature of rubber matrix [2]. There is an op- coupling agents, such as silanes, has been considered in order to
timum degree of crosslinking in the rubber vulcanizates for prac- reduce silanol groups of silica and their adverse effects on the
tical applications, which must be high enough to prevent viscous vulcanization reactions. Ansarifar et al. [11] argued that introduc-
flow but low enough to avoid brittle failure in the vulcanizates [3]. tion and progressive increase in the silica loading (10e50 phr)
Characteristics of the chemical crosslinks in the rubber vulcanizates significantly prolong the vulcanization time by decreasing the
are mainly defined by the mechanism and kinetics of the chemical vulcanization rate. Choi [12] showed that in the compounds filled
reactions in the vulcanization process [4]. with both silane-modified silica and carbon black, vulcanization
time extends by increase in the silica content. Mujtaba et al. [15] by
double quantum nuclear magnetic resonance (NMR) measure-
ments, showed that the average crosslink density in the silica-filled
* Corresponding author. SBR vulcanizates decreases systematically with increasing content
E-mail address: [email protected] (M. Razzaghi-Kashani). of nano-silica particles at high loadings (about 45% for filler

http://dx.doi.org/10.1016/j.polymer.2014.09.073
0032-3861/© 2014 Elsevier Ltd. All rights reserved.
 S.M. Hosseini, M. Razzaghi-Kashani / Polymer 55 (2014) 6426e6434 6427

contents of about 80 phr), compared to the unfilled sample with the (NOES) in an alcohol/water solution in the laboratory. Excess
otherwise identical composition. They assumed that the vulcani- amount of silane was used to achieve the highest degree of sila-
zation agents were partly deactivated by adsorption on the acidic nization of silica. Un-reacted silane was separated by centrifuging,
surfaces of silica, and therefore silica decreased the efficiency of and the weight percent of reacted silane in the modified silica was
vulcanization process. They did not evaluate this behavior at low evaluated by a burning test. Value of 11.3 wt% was confirmed for
loadings of silica. Coupsil8113, and the value of 9.9 wt% was obtained for the silane
In spite of many reports that show organoclays accelerate the used in NOES (Coversil). Recipes of the rubber compounds are given
vulcanization of rubber due to the participation of amine type in Table 1. Compounds starting with Ultrasil are the ones containing
modifiers in the sulfur vulcanization reactions [16e20], our previ- the un-modified Ultrasil VN3. The other compounds starting with
ous work [21] was the first to show that formation of an organoclay Coupsil or Coversil are the ones containing Ultrasil VN3 modified by
network decelerates the rate of vulcanization in butyl rubber due to TESPT or NOES, respectively. The following number shows the part
the limitation in diffusion and accessibility of curing agents to the per hundred rubber (phr) of the neat silica in the compound. The
vulcanization sites, a physical effect which retards the vulcaniza- compound with no silica is labeled as SBR.
tion of butyl rubber with sulfur. Later on, Wu et al. confirmed this
phenomenon in the vulcanization of natural rubber containing 2.2. Sample preparation
graphene, and they also correlated this phenomenon to the re-
strictions in physical diffusion of the vulcanization agents in the After drying silica at 110
C for two days, compounds were
highly filled rubber compounds [22]. They also attributed the prepared by a two-step mixing method. In the first step a master
accelerating effect of graphene at low loadings to the surface batch of rubber and silica was prepared using an internal mixer
chemistry of this filler. Also in the field of thermosets, few re- (Brabender-W50ETH) for 14 min. Then the SBR-silica master batch
searchers have shown that molecular restrictions by means of the was compounded with other ingredients on a two roll-mill (Bra-
filler networks has a destructive effect on the vulcanization kinetics bender-PM2000) with the friction ratio 1:15 for another 14 min.
as well as the physical properties of these material such as glass Samples were formed and vulcanized by compression molding at
transition temperature [23e25]. 160
C.
Bohm and Nguyen [26] and Bohm et al. [27] showed that in the
highly filled rubber compounds most of the filler network is formed
during molding and particularly in the first few minutes of vulca- 2.3. Characterization methods
nization as a result of the filler flocculation at high temperatures. It
was shown that surface modifiers strongly decrease the size of Vulcanization of the compounds was analyzed by an oscillating
aggregates [28] as well as the rate and intensity of the silica disc rheometer (ODR) (Gotech-GT-7070-S2) at 160
C. Also, differ-
network formation [27]. ential scanning calorimeter (DSC) (Netzsch-200 F3Maia) with a
Majority of studies on the kinetics of rubber vulcanization are heating rate of 5
C/min was used to study the specific heat of
performed on compounds filled with high volume fractions of sil- vulcanization reaction in which chemical crosslinking occurs.
ica, and they consider only the chemical effects of this filler on the Nonlinear Viscoelastic behaviors of the vulcanizates were ob-
vulcanization kinetics. However, it seems that chemical character- tained by sweep of strain at the fixed frequency of 1.67 Hz, using a
istics of the fillers are not the only affecting parameters, and a rubber processing analyzer (RPA) (Alpha Technology-RPA2000) at
systematic study of filler-polymer and fillerefiller interactions and 60
C.
their role in the vulcanization of rubber is required. Swelling of the vulcanized samples was performed in toluene at
In the present work, the effects of silica loading, interactions, the room temperature for 7 days. After equilibrium was reached,
and network formation on the kinetics and efficiency of sulfur the solvent was initially wiped off from the sample surface. Then,
vulcanization of styrene butadiene rubber were investigated sys- the sample was weighed in the swollen state, dried at 80
C under
tematically. Silane coupling agents were applied to alter the extent vacuum, and weighed again in the fully dried state. Crosslink
of slow-dynamic chains and nonlinear viscoelastic behavior of the
compounds in order to study its impacts on the vulcanization re- Table 1
actions. It was discussed that immobilization of rubber chains, due Recipe for the Reference and silica filled compounds (in phr).
to the presence of silica, has the major role in altering the kinetics
Compounds Master batch Curing package
and conversion degree of sulfur vulcanization reactions of SBR,
which has not been considered by other researchers. Rubber Silica Modifier 6PPDa Zinc oxide TBBSb Sulfur

SBR 100 0 0 1 0.5 3 2.5

2. Experimental ULTRASIL2 100 2 0 1 0.5 3 2.5
ULTRASIL5 100 5 0 1 0.5 3 2.5
ULTRASIL10 100 10 0 1 0.5 3 2.5
2.1. Materials ULTRASIL20 100 20 0 1 0.5 3 2.5
ULTRASIL40 100 40 0 1 0.5 3 2.5
Styrene butadiene rubber (SBR-1502) with styrene content of ULTRASIL60 100 60 0 1 0.5 3 2.5
23.5%, specific gravity of 0.945, Moony viscosity (ML4) of 84 at COUPSIL2 100 2 0.254 1 0.5 3 2.5
COUPSIL5 100 5 0.637 1 0.5 3 2.5
100
C, and number average molecular weight of 176,000 g/mol COUPSIL10 100 10 1.274 1 0.5 3 2.5
was obtained from the Bandar-Imam Petrochemical Company COUPSIL20 100 20 2.547 1 0.5 3 2.5
(BIPC), Iran. Hydrophilic precipitated amorphous silica (Ultrasil COUPSIL40 100 40 5.096 1 0.5 3 2.5
VN3) was obtained from the Evonik-Degussa AG, Germany with a COUPSIL60 100 60 7.744 1 0.5 3 2.5
COVERSIL5 100 5 0.546 1 0.5 3 2.5
specific surface area of 180 m2/g (measured by N2 adsorption). Also
COVERSIL10 100 10 1.091 1 0.5 3 2.5
two types of pretreated silica were used in this study. One was COVERSIL20 100 20 2.184 1 0.5 3 2.5
Coupsil8113, a surface-modified version of Ultrasil VN3 with COVERSIL40 100 40 4.367 1 0.5 3 2.5
11.3 wt% bis (3-triethoxysilylpropyl) tetra sulfide (TESPT) supplied COVERSIL60 100 60 6.555 1 0.5 3 2.5
by the Evonik-Degussa AG, Germany. The other one was prepared a
6PPD; N-(1,3-Dimethylbutyl)-N0 -phenyl-p-phenylenediamine.
b
by silane modification of Ultrasil VN3 by N-octyl three ethoxy silane TBBS; N-tert-Butyl-2-benzothiazolesulfenamide.
6428 S.M. Hosseini, M. Razzaghi-Kashani / Polymer 55 (2014) 6426e6434

density was determined by the equilibrium swelling measure- further loadings of silica [22]. To study this trend more accurately,
ments based on the FloryeRehner equation [29]: vulcanization kinetics needs to be investigated. For this purpose,
h i h 1 ni kinetic parameters for samples were obtained using rheometry
 lnð1  nr Þ þ nr þ cn2r ¼ V0 n n3 
r
(1) data [32]. The degree of conversion (a) in the vulcanization reaction
2 is defined as:
where nr is the volume fraction of polymer in the swollen mass, V0 is
Mt  M0
the molar volume of the solvent (106.2 cm3 for toluene), n is the a¼ (3)
Mmax  M0
number of active network chain segments per unit volume
(crosslinking density), c is the Flory-Huggins polymeresolvent
where M0 is the initial torque at the onset of vulcanization, Mt is the
interaction term, with the value of 0.44 for toluene-SBR [30]. The
torque value corresponding to a time during vulcanization, and
value of nr was obtained according to the method used by Bala et al.
Mmax is the torque value at the end of reaction. The above mathe-
[31]:
matical expression is based on the fact that crosslink density is
u2 =r2 proportional to the stiffness of rubber. The rate of vulcanization can
nr ¼ (2) be defined as (da/dt) and plotted against time, as shown in
u2 =r2 þ ðu1  u2 Þ=r1
Fig. 2aec for all the samples.
where u1 and u2 are the weights of the swollen and de-swollen As explained by other researchers, vulcanization kinetics of a
samples, and r1 and r2 are the densities of the solvent and the neat rubber, such as SBR, consists of two-stages: the ascending part
polymer, respectively. which is a chemically controlled stage and the descending part as a
stage controlled by diffusion. Vulcanization rate increases in the
first stage since a radical-initiated reaction is promoted by the re-
3. Results
action product. In the second stage, vulcanization rate decreases
due to the increase in viscosity and reduction of molecular mobility
3.1. Vulcanization kinetics by rheometer
[23,24,33]. Ding and Leonov [34] found that occurrence of a peak in
the vulcanization rate is possibly due to gelation (gel point),
In order to study the influence of unmodified and modified silica
resulting from formation of an infinite chemical network [35]. As
on the vulcanization kinetics of SBR, vulcanization curves of sam-
shown in Fig. 2aec, by increasing silica loading, the gel point shifts
ples obtained from ODR were analyzed, and results are summarized
to lower times, and magnitude of the maximum vulcanization rate
in Table 2. Minimum torque (ML), maximum torque (MH) and their
(obtained from fitting data with autocatalytic model) initially in-
difference (DM) increase as the amount of silica increases in all
creases, going through a peak, and then decreases, as it is plotted
samples. There is a notable rise in the values of ML and MH above
against silica loading in Fig. 3. The filler loading at which this peak
10 phr, where the rubber-mediated silica network is presumably
occurs will be called “critical loading”, as shown by a vertical line.
formed in the compounds containing both unmodified and modi-
Below the critical loading (stage I), the maximum vulcanization rate
fied silica. Large rate of increase in the torques with loadings of
increases, but beyond this loading (stage II), the maximum vulca-
10 phr or above can be attributed to the slow-dynamic chains
nization rate reduces as more silica is added to the compounds. As it
bridging highly attractive silica aggregates, which are more in
will be discussed in more details latter, presence of a critical loading
samples containing unmodified silica than modified ones, as shown
and occurrence of the peak in the maximum vulcanization rate will
in Fig. 1. Silane modification of silica reduces the fillerefiller
be attributed to the formation of a physical infinite network of filler,
interaction and networking to a great extent.
mediated by semi-glassy and immobilize rubber chains. The critical
As seen in Table 2, scorch time (tS2) monotonically decreases as
loading, as low as 5 phr for unmodified silica or Ultrasil, has been
all types of silica are loaded. However, the time to reach full
hardly noticed by other researchers in the practical applications of
vulcanization (tC100) goes through a minimum before increasing
silica in rubber compounds. Therefore, decelerating effect of silica
again by silica loading. Using Dt ¼ tC100tS2 as a measure of the
at high loadings has been solely attributed to the surface chemistry
vulcanization time and rate, it seems that in all cases, silica initially
of silica as acidic hydroxyl groups adsorb the basic accelerators. As
accelerates vulcanization rate, but it has a decelerating effect with
it will be discussed latter, acceleration of vulcanization process by

Table 2
Vulcanization characteristics of compounds obtained from rheometry.

Compounds ML MH (N m) DM (MHeML) tS2 (min) tC100 (min) Dt (min)

(N m) (N m)

SBR 0.482 2.743 2.261 11.033 15.467 4.434

ULTRASIL2 0.528 2.903 2.375 7.433 10.750 3.317
ULTRASIL5 0.545 3.556 3.011 4.583 7.200 2.617
ULTRASIL10 0.570 4.098 3.528 4.067 6.883 2.816
ULTRASIL20 0.956 4.487 3.531 3.833 8.200 4.367
ULTRASIL40 2.567 6.104 3.537 3.483 10.183 6.700
ULTRASIL60 6.584 11.160 4.576 2.833 40.317 37.484
COUPSIL2 0.481 2.745 2.264 8.950 13.350 4.400
COUPSIL5 0.515 2.953 2.438 8.017 11.717 3.700
COUPSIL10 0.598 3.849 3.251 6.083 9.467 3.384
COUPSIL20 0.771 5.134 4.363 5.467 10.050 4.583
COUPSIL40 1.368 6.943 5.575 4.483 11.033 6.550
COUPSIL60 2.608 8.983 6.375 3.550 13.500 9.950
COVERSIL5 0.539 2.921 2.382 10.517 14.550 4.033
COVERSIL10 0.607 3.201 2.594 10.033 14.150 4.117
COVERSIL20 0.759 3.740 2.981 9.383 13.933 4.550
COVERSIL40 1.397 5.929 4.533 7.267 12.500 5.233
Fig. 1. Effect of silica loading on the minimum torque (ML). The lines are the guides to
COVERSIL60 2.920 8.376 5.455 5.317 28.967 23.650
the eyes.
 S.M. Hosseini, M. Razzaghi-Kashani / Polymer 55 (2014) 6426e6434 6429

the filled compounds as chemical and diffusion controlled stages

[22]. However, we now hypothesize that immobilization of rubber
chains, influenced by the filler-polymer and the fillerefiller in-
teractions, has a substantial role in the kinetics of rubber vulcani-
zation. It is also worth noting from Fig. 3 that silica modification has
postponed the critical loading from about 5 phr for the un-modified
silica (solid line) to about 10 phr for the modified ones (dotted line),
and the vulcanization rate is less dependent on the silica loading in
both stages.
An autocatalytic kinetic model normally predicts vulcanization
reactions in the rubber compounds [36]. In these reactions, the
conversion rate is related not only to the un-reacted materials, but
also to the reaction products. Therefore, the maximum rate of
vulcanization occurs at a conversion other than zero (Equation (4))
[16]:

da
¼ kðTÞam ð1  aÞn (4)
dt

where k(T) is the rate constant, m is the reaction order of autocat-

alytic reaction, and n is the reaction order of non-autocatalytic re-
action. This model was fitted to the vulcanization rate plotted
against conversion, as shown in Fig. 4aec. The values of k(T), m and
n for all vulcanizates were calculated using a non-linear regression
fit and listed in Table 3. As shown in this table, the experimental
data can be well fitted with Equation (4), regardless of the modi-
fication or concentration of filler. It can be seen that by incorpo-
ration of silica particles, k(T) initially increases followed by a
decrease as more silica is added. This signifies that the vulcaniza-
tion rate is promoted by silica at low loadings and subsequently
suppressed by it. In agreement with Fig. 2, the gel point shifts to the
lower conversions as concentration of silica increases.

3.2. Vulcanization kinetics by DSC

Differential Scanning Calorimetery (DSC) is another widely used

technique to study the vulcanization kinetics. Through monitoring
changes in the heat flow, the extent of chemical crosslinking is
measured with no influence from physical crosslinks appears in

Fig. 2. Rate of vulcanization as a function of time obtained from rheometry for: a)

ULTRASIL, b)COUPSIL and c)COVERSIL compounds. The lines are the guides to the eyes.

increasing filler loadings below the critical point will be attributed

to the larger amount of cross-link agents per volume of free rubber
available for vulcanization in the filled compounds than in the neat
compound. The amount of immobilized rubber, in the form of shell
and occluded rubber, deprived from the vulcanization agents will
be influenced by filler-polymer interactions. As seen in Fig. 3,
modification of silica reduces this accelerating effect in the stage I
due to the reduced filler-polymer interaction and the extent of shell
and occluded rubber.
Suppression of vulcanization rate above the critical point was
initially explained in our previous work as formation of the filler
network and entrapment of the rubber which limits diffusion of the
vulcanizing agents and reduces the availability of the reaction sites
Fig. 3. Effect of silica loading and various surface modifications on maximum vulca-
of rubber matrix to these agents [21]. Wu et al. used the same nization rate obtained from fitting data with autocatalytic model. The lines are the
argument to propose a two-stage mechanism for vulcanization of guides to the eyes.
6430 S.M. Hosseini, M. Razzaghi-Kashani / Polymer 55 (2014) 6426e6434

temperature (Tend) have been postponed to higher temperatures,

which can be correlated to the molecular mobility restrictions exerted
by silica as its amount increases. Immobilization of rubber chains by
reinforcing fillers has been equated to the effects of lower tempera-
tures [38], therefore more energy is needed to overcome the mobility
restrictions in order to advance the crosslinking reactions. As a result,
not only the rate, but also the conversion degree of vulcanization
reactions decreases as more rubber immobilizes in the filler network
of the highly filled compounds. As can be seen in Fig. 5, the exothermal
peak of vulcanization reactions is split into two peaks for the com-
pounds containing 60 phr silica. Wu et al. [22] interpreted this as two-
stage reactions, i.e. chemical reaction controlled stage and diffusion
controlled stage. However, we interpret this as the two-stage vulca-
nization for highly filled rubber compounds, i.e. a free rubber vulca-
nization stage at lower temperatures and an immobilized rubber
vulcanization stage at higher temperatures.

3.3. Nonlinear viscoelastic behavior by RPA

In order to correlate vulcanization kinetics of filled SBR to the

formation of a polymer-mediated filler network and understand
the effect of surface modification of silica on this correlation,
dynamic-mechanical analysis in a strain-sweep mode was
employed using RPA. Fig. 6 shows reduction in the storage modulus
(G0 ), known as the Payne effect [39], for Coupsil compounds loaded
with different amounts of modified silica. The critical loading point
for the vulcanization of Coupsil compounds was distinguished to be
around 10 phr. As shown in Fig. 6, compounds below the critical
loading, e.g. the reference and Coupsil5, show negligible non-linear
behavior originating from the destruction of the polymer-mediated
filler network. However, for the compounds above the critical
point, e.g. Coupsil20 and Coupsil60, destruction of an already
existing network is obvious. Therefore, the critical point of loading
silica at which vulcanization behavior of rubber transforms co-
incides with the percolation limit of loading silica at which an
infinite network of filler, mediated by the entrapped rubber, forms.
Fig. 7 shows similar results for the compounds containing 60 phr of
unmodified or modified silica. This figure shows reduced Payne
effect for the compounds containing modified silica, resulting from
less polymer-mediated filler network. In both figures, G00 and G0∞
are the limiting values of the storage modulus at small and large
strain amplitudes, respectively. The value of DG0 ¼ G00 G0∞ can be
used as a measure of intensity of the filler network as it breaks
down by the dynamic-mechanical energy [15,39]. As shown in

Table 3
Fig. 4. Vulcanization rate as a function of conversion for: a)ULTRASIL, b)COUPSIL and c)
Kinetic parameters for compounds, obtained by fitting the vulcanization experi-
COVERSIL. The lines are the result of fitting data with Equation (4).
mental data with Equation (4)

Compound k (sec1)  100 m n r2

results from rheometer. The non-isothermal heat flow curves of
SBR 1.372 1.181 0.531 0.929
vulcanization for selected compounds are shown in Fig. 5. Char-
ULTRASIL2 3.559 1.398 0.754 0.981
acteristics of these curves are listed in Table 4. ULTRASIL5 6.224 1.561 0.787 0.990
The most important point in Fig. 5 and Table 4 is a large decrease in ULTRASIL10 4.694 1.355 0.796 0.987
the enthalpy of vulcanization per gram rubber which is an indication ULTRASIL20 4.578 1.403 1.063 0.991
of less chemical reactions as loading of all types of silica increases. In ULTRASIL40 1.281 0.689 0.843 0.974
ULTRASIL60 0.851 0.524 1.695 0.987
other words, silica has immobilized and eliminated part of the rubber COUPSIL2 2.759 1.384 0.755 0.994
from being crosslinked by sulfur, which results in a lower reaction COUPSIL5 3.979 1.586 0.779 0.978
conversion. Surface modification of silica, in the Coupsil and Coversil COUPSIL10 5.351 1.535 0.845 0.994
compounds, increases the enthalpy of vulcanization due to reduction COUPSIL20 5.286 1.431 1.052 0.991
COUPSIL40 4.934 1.401 1.035 0.992
of the fillerefiller interactions [37]. Therefore, the amount of immo-
COUPSIL60 4.230 1.215 1.089 0.990
bilized rubber, which does not participate in the vulcanization re- COVERSIL5 2.463 1.345 0.637 0.991
actions, reduces. Also, increasing the amount of silica shifted the onset COVERSIL10 2.733 1.411 0.647 0.985
temperature (Tonset) to lower temperatures. It suggests that the COVERSIL20 2.681 1.314 0.742 0.983
presence of silica particles caused the vulcanization process to start COVERSIL40 2.460 1.140 0.777 0.993
COVERSIL60 2.393 1.087 1.200 0.981
sooner. However, the exothermal temperature (Tpeak) and the end
 S.M. Hosseini, M. Razzaghi-Kashani / Polymer 55 (2014) 6426e6434 6431

Fig. 6. Payne effect representation for COUPSIL compounds. The lines are the guides to
the eyes.

Fig. 7. Payne effect representation for highly filled compounds. The lines are the
guides to the eyes.

Table 5, DG0 increases as the loading of Coupsil increases, and it is

larger for the Ultrasil60 containing unmodified silica than the two
other compounds containing modified silica. This is due to the
strong fillerefiller interaction in the Ultrasil60, which alleviates by
the surface modifications. As shown in Table 5, DG0 has the order:
Ultrasil60 > Coversil60 > Coupsil60 [40]. Also, it was already
concluded from the vulcanization kinetics studies that the rate and
Fig. 5. Heat flow curves of nanocomposites at a heating rate of 5
C/min for: a)
conversion of the reactions has the order:
ULTRASIL, b)COUPSIL and c)COVERSIL. The cures are shifted vertically for clarity. Coupsil60 > Coversil60 > Ultrasil60. Therefore, one may conclude
that the rate and conversion of the vulcanization reaction has a
reverse correlation to the entrapment and immobilization of rub-
ber in the silica network.

Table 4
Vulcanization characteristics of compounds obtained from DSC. 3.4. Crosslink density of vulcanizates by equilibrium swelling
Tonset (
C) Tpeak (
C) Tend (
C) DHRubber (J/g)
The results of estimating crosslink density from equilibrium
SBR 171.9 176.8 186.0 18.843
swelling measurements and the FloryeRehner equation are shown
ULTRASIL2 168.3 177.4 188.4 15.943
ULTRASIL20 160.7 183.3 220.7 11.715
in Fig. 8.
ULTRASIL60 138.0 160.1, 205.3 226.6 10.957
COUPSIL5 171.1 175.2 191.6 15.600
COUPSIL20 165.7 180.0 192.9 14.120 Table 5
COUPSIL60 149.4 166.4, 199.6 214.7 13.630 Payne effect amplitude for unfilled and filled compounds.
COVERSIL5 160.9 169.9 182.2 14.598
SBR COUPSIL5 COUPSIL20 COUPSIL60 COVERSIL60 ULTRASIL60
COVERSIL20 159.9 174.7 189.4 13.920
COVERSIL60 151.3 173.9, 200.6 225.5 13.100 DG0 (kPa) 101.22 121.58 610.0 2782.2 4356.5 9896.7
6432 S.M. Hosseini, M. Razzaghi-Kashani / Polymer 55 (2014) 6426e6434

energy required for the reaction to happen. Therefore, the proba-

bility of the reaction to occur is represented by a rate constant, k(T)
in the kinetic equations. It has been shown that the rate constant in
small molecule reactions is a sole function of the absolute tem-
perature, according to the Arrhenius equation [42]. However, it was
observed in Table 3 that the rate constant for the rubber vulcani-
zation reaction also varies with the silica loading at a fixed tem-
perature. Therefore, for the filled rubber compounds, the kinetic
energy of the reacting chains depends not only on the reaction
temperature, but also on the conformation of the chains in the vi-
cinity of the reinforcing particles. It has been shown that interac-
tion of the active filler particles with rubber chains drastically
changes the conformational behavior of the chains [43e45]. These
changes are due to the adsorption of some chains on the filler
surface, loss of entropy on some more entangled chains [46,47], as
well as the excluded-volume effect of hard filler particles [2]. As a
Fig. 8. Effect of silica loading on the crosslink density of compounds. The lines are the result, there is a gradient of glass transition temperature (Tg) of
guides to the eyes. chains from the surface of filler toward the bulk of polymer [40,46].
Also, rubber immobilizes in confinements generated by the filler-
As shown in this figure, by addition of silica crosslink density efiller interactions, leading to the formation of a rubber-mediated
increases. It is also shown that the order of crosslink density is: network of filler at high concentrations of filler in the compounds
Coupsil > Ultrasil > Coversil. The FloryeRehner equation gives the [2,40].
sum of physical and chemical crosslinks, so that by increasing the At low concentration of the reinforcing fillers, only shell and
silica loading, the physical crosslinks strongly increase. However, occluded rubber are affected by the filler-polymer interaction and
DSC results showed that by increasing silica loading, especially depleted from the vulcanizing agents [48,49]. Therefore, we hy-
when the filler network forms, conversion in the vulcanization pothesize that the amount of vulcanizing agents per volume of the
reactions decreases. In order to separate the chemical crosslinks free rubber in the low-filled compounds is higher than that in the
from the physical ones, the modified FloryeRehner theory pre- unfilled reference compound, due to the elimination of shell and
sented by Kraus for particulate filled vulcanizates was used [41]. occluded rubber from vulcanization. This will result in faster onset
Degree of restriction exerted by the reinforcing fillers has been and higher rate of vulcanization with respect to the neat rubber
given by the equation: with no filler. This phenomenon can be correlated to the strength of
! filler-polymer interaction rather than the chemistry of filler surface
yr0 ff which was explained by Wu et al. [22]. As a result, even for silica
¼1m (5)
yr 1  ff with apparently detrimental surface chemistry for rubber vulca-
nization [11e14], this accelerating effect at lower loadings is
where nr0 and nr are the volume fraction of polymer in the fully observed. The accelerating effect largely reduces by the silane-
swollen unfilled and filled samples, respectively. ff is the volume modification of silica and coverage of the acidic hydroxyl groups
fraction of the filler and the slop m is a direct measure of the on the surface, as otherwise would be expected if chemistry of the
reinforcing capacity of the filler in the matrix. The reinforcing effect filler surface was involved.
of filler can be obtained by plotting nr0/nr against ff/(1ff). Ac- At high concentration of reinforcing fillers with strong filler-
cording to the Kraus theory, the curve with a larger negative slope efiller interactions, flocculation of filler occurs at high vulcaniza-
indicates a higher reinforcing effect. The slope of Coupsil, Ultrasil, tion temperatures, and the polymer-mediated filler network is
and Coversil filled vulcanizates have been found to formed which entraps and immobilizes a large portion of rubber
be 1.730, 1.638 and 0.942, respectively. It can be concluded [26,27,50]. This entrapped and immobilized rubber, although con-
that the rubberefiller interaction and the degree of reinforcement tains already dispersed vulcanizing agents, does not have enough
increases in the order Coupsil > Ultrasil > Coversil. The higher
reinforcement capacity of the Coupsil samples than the Ultrasil
samples is probably due to the chemical links between silica and
rubber generated by TESPT. The lower reinforcement capacity of
Coversil than Ultrasil vulcanizates is probably due to the lower
fillererubber interaction in the Coversil [37]. Based on the Kraus
modification, the chemical crosslink density for the Coupsil sam-
ples were calculated and shown in Fig. 9. As shown in this figure,
with increasing silica loading chemical crosslink density initially
increases up to about 20 phr loading, but it decreases afterward.
Reduction in the chemical crosslink with increasing silica at high
loadings was also confirmed by Mujtaba et al. [15] in Ultrasil 7000
GR filled rubber compounds.

4. Discussion on the physical effects of silica on vulcanization

kinetics of rubber
Fig. 9. The crosslink density obtained from FloryeRehner and Kraus modification and
According to the collision theory in chemical reactions, any enthalpy of vulcanization per gram rubber obtained from DSC for the Coupsil vulca-
reactant must have enough kinetic energy to surpass the activation nizates. The lines are the guides to the eyes.
 S.M. Hosseini, M. Razzaghi-Kashani / Polymer 55 (2014) 6426e6434 6433

kinetic energy to participate in the vulcanization reaction. Naka- in the polymer matrix which affects vulcanization of rubber, as
jima and Nishi used an atomic force microscope (AFM) to show that reported by other researchers. However, it was shown that there
the bound rubber had a two-layer structure consisting of a glassy are physical phenomena originating from the filler-polymer and
layer of about a few 2e3 nm in thickness and an uncured rubber fillerefiller interactions which impact the kinetics and conversion
layer of about several 6e7 nm in thickness [51]. Also Lipatov stated of the vulcanization reactions in rubber. These two types of in-
that “near the filler surface, the rate of vulcanization is much lower teractions had opposite effects on the vulcanization reactions,
than in the bulk” [43]. leading to a critical point in filler loadings. At low loadings prior to
Modification of silica surface by silane decreases silica-SBR in- the critical point, silica even accelerated the vulcanization reaction
teractions [37], and allows for more diffusion of vulcanizaing agents with respect to the neat compound, which cannot be explained
into the shell and occluded rubber, as they are normally added after based on the chemistry of silica surface. However, at high loadings
incorporation of filler during mixing. The result is a decreased above the critical point, silica suppressed the vulcanization reaction
accelerating effect of silica at low loadings, as show in Fig. 3. Silane compared to the vulcanization of the neat rubber compound. The
modification of silica has also a drastic effect on decreasing filler- accelerating effect was explained based on the deprivation of shell
efiller interactions, improvement of filler dispersion, and subse- and occluded rubber from crosslinking agents, resulting from the
quent reduction of the rate and intensity of filler flocculation and filler-polymer interactions. On the other hand, the suppressing ef-
networking at high loadings [27,28,37]. Therefore, surface modifi- fect of silica was correlated to the loss of entropy and immobili-
cation of silica reduces the immobilized rubber entrapped in the zation of rubber chains entrapped in the polymer-mediated filler
filler network and increases the kinetic energy and the probability of network, resulting from the fillerefiller interactions. The latter in-
vulcanization reaction to occur at a certain temperature. As a result, teractions greatly reduce the required kinetic energy of chains to
the rate of vulcanization at high modified silica loadings will remain participate in the vulcanization reaction. Modification of silica by
higher and less dependent on the concentration of the filler in silane could alleviate the filler-polymer and fillerefiller interactions
comparison with the un-modified silica, as shown in Fig. 3. Also, the and reduce both accelerating and decelerating effects of silica on
second peak in the DSC graphs of Fig. 5 for highly filled compounds the vulcanization reactions. The critical point in the loading of silica
can be explained as the vulcanization of immobilized chains was correlated to the loadings at which the non-linear viscoelastic
entrapped in the filler network which occurs at higher temperatures. behavior, resulting from formation of the filler network, appears in
In summary, we propose a mechanism for the effect of rein- the highly filled rubber vulcanizates.
forcing fillers on the rubber vulcanization so that the accelerating
effect of filler at low loadings is defined based on the filler-polymer References
interactions and diffusional limitations for the vulcanizing agents
in the shell and occluded rubber. Whereas, the decelerating effect [1] Mihara S. Reactive processing of silica-reinforced tire rubber: new insight into
of the reinforcing fillers at high loadings is explained based on the the time and temperature dependence of silica rubber interaction. Ph.D
Thesis. University of Twente; 2009.
fillerefiller interactions and the reduced dynamic mobility and [2] Frohlich J, Niedermeier W, Luginsland HD. Compos Part A 2005;36:449e60.
kinetic energy of rubber chains, which reduces the probability of [3] Akiba M, Hashim AS. Prog Polym Sci 1997;22:475e521.
vulcanization reaction. The proposed mechanism of the filler ef- [4] Blow C, Loo C. Polymer 1975;16:205e8.
[5] Mark JE, Erman B, Eirich FR. Science and technology of rubber. 3rd ed. New
fects on the chain conformation and subsequent vulcanization York: Academic Press; 2005.
process is schematically depicted in Fig. 10. [6] Coran AY. J Appl Polym Sci 2003;87:24e30.
[7] Dijkhuis KA, Noordermeer JW, Dierkes WK. Eur Polym J 2009;45:3302e12.
[8] Saatchi MM, Shojaei A. Polym Int 2012;61:664e72.
5. Conclusions [9] Semsarzadeh MA, Bakhshandeh GR, Ghasemzadeh-Barvarz M. Iran Polym J
2005;14:573e8.
Effect of silica on the vulcanization kinetics and crosslink den- [10] Li ZH, Zhang J, Chen SJ. Express Polym Lett 2008;2:695e704.
[11] Ansarifar MA, Chugh JP, Haghighat S. Iran Polym J 2000;9:81e7.
sity of styrene butadiene rubber was investigated by means of [12] Choi SS. Polym Int 2001;50:524e30.
rheometer, differential scanning calorimeter, and equilibrium [13] Choi SS, Park BH, Song H. Polym Adv Technol 2004;15:122e7.
swelling methods. It was discussed that it is not necessarily the [14] Ramier J, Chazeau L, Gauthier C. Rubber Chem Technol 2007;80:183e93.
[15] Mujtaba A, Keller M, Ilisch S, Radusch HJ, Thurn-Albrecht T, Saalwachter K,
chemical nature of the filler surface or diffusion of crosslink agents et al. Macromolecules 2012;45:6504e15.
[16] Lopez-Manchado MA, Arroyo M, Herrero B, Biagiotti J. J Appl Polym Sci
2003;89:1e15.
[17] Varghese S, Karger-Kocsis J, Gatos KG. Polymer 2003;44:3977e83.
[18] Arroyo M, Lopez-Manchado MA, Herrero B. Polymer 2003;44:2447e53.
[19] Varghese S, Karger-Kocsis J. J Appl Polym Sci 2004;91:813e9.
[20] Kader MA, Nah C. Polymer 2004;45:2237e47.
[21] Sepehri A, Razzaghi-Kashani M, Ghoreishy MHR. J Appl Polym Sci 2012;125:
E204e13.
[22] Wu J, Xing W, Huang G, Li H, Tang M, Wu S, et al. Polymer 2013;54:3314e23.
[23] Sanctuary R, Baller J, Zielinski B, Becker N, Kruger JK, Philipp M, et al. J Phys
Condens Matter 2009;21:035118.
[24] Baller J, Becker N, Ziehmer M, Thomassey M, Zielinski B, Muller U, et al.
Polymer 2009;50:3211e9.
[25] Roman F, Calventus Y, Colomer P, Hutchinson JM. Thermo Chim Acta 2012;54:
76e85.
[26] Bohm GA, Nguyen M. J Appl Polym Sci 1995;55:1041e50.
[27] Bohm GA, Tomaszewski W, Cole W, Hogan T. Polymer 2010;51:2057e68.
[28] Qu L, Yu G, Wang L, Li C, Zhao Q, Li J. J Appl Polym Sci 2012;126:116e26.
[29] Flory PJ. Principles of polymer chemistry. Ithaca, New York: Cornell University
Press; 1953. p. 576.
[30] He MJ, Zhang HD, Chen WX, Dong XX. Polymer physics. 3rd ed. Shanghai PRC:
Fudan University Press; 2007. p. 51.
[31] Bala P, Samantaray BK, Srivastava SK, Nando GB. J Appl Polym Sci 2004;92:
3583.
[32] Ding R, Leonov AI. J Appl Polym Sci 1996;61:455e63.
Fig. 10. Schematic representation of the chemical network structures of [33] Kamal MR, Sourour S. Polym Eng Sci 1973;13:59e64.
nanocomposites. [34] Ding R, Leonov AI. Rubber Chem Technol 1999;72:361e83.
6434 S.M. Hosseini, M. Razzaghi-Kashani / Polymer 55 (2014) 6426e6434

[35] Halley PJ, George GA. Chemorheology of polymers. New York: Cambridge [45] Montes H, Lequeux F, Berriot J. Macromolecules 2003;36:8107e18.
University Press; 2009. p. 180. [46] Berriot J, Montes H, Lequeux F, Long D, Sotta P. Macromolecules 2002;35:
[36] Piloyan GO, Ryabchikov ID, Novikova OS. Nature 1966;212:1229. 9756e62.
[37] Stockelhuber KW, Svistkov AS, Pelevin AG, Heinrich G. Macromolecules [47] Havet G, Isayev AI. Rheol Acta 2003;42:47e55.
2011;44:4366e81. [48] Fukahori Y. J Appl Polym Sci 2005;95:60e7.
[38] Wang X, Robertson CG. Phys Rev E 2005;72:031406. [49] Fukahori Y. Mechanism of the carbon black reinforcement of rubbers. In:
[39] Payne AR. J Appl Polym Sci 1962;6:57e63. Bhowmick AK, editor. Current topics in elastomers research. Boca Raton, FL:
[40] Fritzsche J, Kluppel M. J Phys Condens Matter 2011;23:035104. Taylor & Francis; 2008. p. 518e39.
[41] Kraus G. J Appl Polym Sci 1963;7:861e71. [50] Richter S, Saphiannikova M, Stockelhuber KW, Heinrich G. Macromol Symp
[42] Christov SG. Collision theory and statistical theory of chemical reaction. Berlin 2010;291-292:193e201.
Heidelberg: Springer-Verlag; 1980. [51] Nakajima K, Nishi T. In: Bhowmick AK, editor. Current topics in elastomers
[43] Lipatov YS. Polymer reinforcement. Chem Tec Publishing; 1995. p. 198. research recent developments in rubber research using atomic force micro-
[44] Berriot J, Lequeux F, Monnerie L, Montes H, Long D, Sotta P. J Non-Cryst Solids scopy. Boca Raton: CRC Press; 2008. p. 579.
2002;307e310:719e24.