Influence of Additives

Polymer blends Mechanical properties Dynamic mechanical thermal
analysis Partition of additives Natural rubber and butadiene rubber
NR/BR blends (50/50) were prepared
by using four different sequences to
incorporate the additives. After curing under constant conditions the
vulcanizates demonstrate sequence-dependent mechanical
properties. Based on dynamic mechanical thermal analysis and
scanning electron microscopy a
correlation between the mechanical
performance and the blend morphology can be established. The effects are caused by the selective
partition of the additives in each
rubber phase.
Einfluss der Mischreihenfolge
auf die Eigenschaften von NR/
BR Verschnitten
Polymerverschnitte mechanische
Eigenschaften dynamisch-mechanische thermische Analyse Verteilung von Additiven NR/BR-Verschnitte
NR/BR-Blends (50/50) denen die
Additive nach vier unterschiedlichen
Mischreihenfolgen zugemischt wurden, weisen Unterschiede in ihren
mechanischen Eigenschaften auf,
die von der Reihenfolge der Zugabe
gepraÈgt sind. Mit Hilfe von dynamisch-mechanischer Thermoanalyse sowie Rasterelektronenmikroskopie konnte gezeigt werden, dass
die Unterschiede in den mechanischen Eigenschaften von den entstandenen Morphologien abhaÈngen.
Diese sind von der selektiven Verteilung der Additive in den jeweiligen Kautschukphasen gegeben.
Influence of Additives
Incorporation Sequence
on NR/BR Blend Properties
D. Freitas de Castro, A. F. Martins, J. C. M. Suarez,
R. C. Reis Nunes, L. L. Y. Visconte, Rio de Janeiro (Brasil)
Rubber formulations are developed to
meet specific requirements. Thus, each
one of the various components taking
part in a particular formulation has an important role as to give its contribution to
the final properties [1]. However, even
with the wide choice of ingredients available, the whole set of characteristics may
still not be reached and many times a
combination of two or more rubbers
has to be used in the preparation of compounds for special properties. So, generally speaking, elastomeric materials are
blended for properties improvement, better processing or lower cost [1 ± 7]. Natural rubber (NR) is the adequate choice
when good tensile and tear strengths
are demanded, since these characteristics can be developed due to the capability this rubber has to crystallize under
stress [8 ± 13]. Polybutadiene rubber
(BR) is characterized by its superior abrasion resistance, so that blends of NR and
BR that combine the excellent processing and physical properties of the former
with the superior abrasion resistance of
the latter are largely used in the industry
in the production of tyre treads and conveyor belts [1 ± 6, 14]. However, the rubber performance is determined not only
by the right choice of the formulation
components but also by such design factors as constituents patterns and fabrication process parameters. The final properties will be strongly influenced by the
additive distribution in each elastomeric
phase and by the interfacial compatibility
between the components. As the rubbers mixtures are usually multiphase systems, according to the used compounding mode, different distributions of the additives in each rubber phase can be
KGK Kautschuk Gummi Kunststoffe 56. Jahrgang, Nr. 1-2/2003
achieved, depending on the degree of affinity that each additive has towards each
rubber. In general, competitive vulcanization occurs due to different rates of vulcanization and/or rates of diffusion of the
additives (including sulfur, accelerator
and other compounds) in each of the
elastomeric phases [15 ± 17].
In the present work, which is part of a
general investigation on compounding
modes in the preparation of rubber materials, the influence of four preparation
modes on the properties of 1:1 (w/w)
NR/BR mixture is investigated. The behavior of each compound mixture was evaluated by tensile and tear tests, dynamic
mechanical thermolanalysis and scanning electron microscopy.
The formulation used to prepare the mixtures is shown in Table 1. Curing parameters as determined on an oscillating
disk rheometer, model TI-100 from TecnologõÂa Industrial, operating at 160 8C in
a 38 arc, are given in Table 2 [18-a]. The
mixtures were prepared in a Berstorff
two rolls mill, operating at 50 8C. Vulcanization was carried out at 160 8C on an
electrically heated press.
The mixtures were compounded according to the following four procedures:
M1: the additives except the accelerator were introduced into NR. The BR was
added to this previous mixture. The accelerator was added after the homogenization of the new mixture;
M2: the additives except the accelerator were introduced into BR. The NR was
added to this previous mixture. The ac-
Influence of Additives Incorporation Sequence on NR/BR Blend Properties
Table 1. Formulation used to prepare the
Natural rubber (NR) a
Polybutadiene rubber (BR)
Zinc oxide
Stearic acid
Aminox c
Mooney viscosity, ML(1 ‡ 4) at 100 8C ˆ 102.6;
Mooney viscosity, ML(1 ‡ 4) at 100 8C ˆ 41.7;
Reaction product between diphenylamine and
acetone, obtained at low temperature;
celerator was added after the homogenization of the new mixture;
M3: the two rubbers were previously
mixed before the addition of the additives;
M4: each rubber was initially compounded with half amount of each additive, except the accelerator that was the
last ingredient to be incorporated.
The effects of compounding mode on
the polymer behavior were evaluated by
mechanical (tensile, tear and hardness)
tests, dynamic mechanical thermolanalysis as well as scanning electron microscopy (SEM).
Tensile and tear tests were carried out
at room temperature in a model 1101 Instron universal testing machine with a
cross-head speed of 500 mm/min, according, respectively, to ASTM D 412
[18-b] and D 624 [18-c]. The tensile specimens were die cut from vulcanized rubber plates. For each compounding mode
at least 5 specimens were tested and the
stress and strain at break determined.
Hardness measurements were performed according to ASTM D 2240 [18d] on a Shore A hardness tester.
Dynamic mechanical thermolanalysis
was performed on a Rheometric Scientific DMTA analyzer, model MK III, under
the following conditions: frequency
1 Hz, heating rate 2 8C/min, single cantilever bending mode; temperature ranging
from ÿ 1308 to 20 8C.
A JEOL scanning electron microscope
model JSM 5800LV was used to analyze
the fracture aspects of the rubber compounds. The study of the failure mechanisms was carried out by direct observation of the sample topography of samples
Table 2. Curing characteristics of the rubber compounds
Optimum cure time ± t90 (min)
Minimum torque ± ML (dN.m)
Maximum torque ± MH (dN.m)
Scorch time ± ts2 (min)
Cure rate index ± CRI
cryofractured after being immersed for at
least 10 min in liquid nitrogen. All samples
were sputter-coated with gold in a vacuum chambre before examination.
Results and Discussion
Rheometric and mechanical
Table 2 shows rheometric parameters for
the compositions with the isolated rubbers and the four rubber blends. Concerning the isolated rubbers compositions, it can be seen that NR has higher
reactivity in comparison to BR, given by
the shorter value of t90 and the higher value of CRI. Among the rubber blends, a
few differences in rheometric parameters
can be observed as a function of the preparative mode.
Table 3 presents rate constants for the
rubber compounds, estimated according
to the first order kinetics. When two or
more rubbers are blended, one of the
problems one must be concerned with
is the difference in vulcanization rate
each rubber presents towards the particular vulcanizing system used. If the rates
are very different care must be emphasized as vulcanization of the blend as a
whole may induce over or undercuring
of the rubber phases.
Figure 1 presents the rheografic curves
for the mixtures under investigation. It can
be seen that NR vulcanizes much faster
than BR so the values of t90 measured
for the blends, lying in between those
for the isolated rubbers, may be leading
to an overcured NR phase and/or an undercured BR phase.
NR is known to have a strong tendency
to undergo reversion as the result of thermo oxidative degradation brought about
by long heating times [19]. Upon thermo
oxidative degradation, with normally occurs via scission or depolymerization of
the polymer molecules, NR gradually becomes softer and sticky and as a consequence, the maximum torque, which is
the parameter related to the number of
crosslinks, decreases. This is shown in
Figure 1 for the curve corresponding to
this rubber. Thus long vulcanization times
Fig. 1. Rheographic
curves for NR/BR
KGK Kautschuk Gummi Kunststoffe 56. Jahrgang, Nr. 1-2/2003
Influence of Additives Incorporation Sequence on NR/BR Blend Properties
Fig. 2. Mechanical properties of NR/BR mixtures
are not advisable if the good characteristics of NR are to be preserved. On the
other hand, if t90 is not long enough,
the BR phase will not be fully vulcanized,
so that many of the features developed in
the blends will be dependent of an adequate balance of these two factors.
Among the blends investigated M2 and
M3 are the ones with t90 more similar to
NR compound (Table 2).
Concerning maximum torque, the
compositions NR or BR present similar
values while the blends, except for M4,
have higher values, which are not very different from each other. Figure 1 also
shows that analogously to NR the rheometric curves for M1, M2 and M3, go
through a maximum and then start to decrease, indicating that a reversion process is taking place, characteristic of
NR degradation.
The close behavior of M2 and M3 towards vulcanization, given by the similarity of their rheometric parameters, suggests that although being prepared ac-
Table 3. Rate constant of rubber compounds
k (10 k,minÿ1)
Fig. 3. Tear strength and shore A hardness values for the rubber
cording to distinct preparative modes,
these two blends have similar susceptibility to undergo vulcanization, which allows
us to presume that the additives might be
distributed very much in the same way
within these two mixes. This pressuposes
a preferential migration of the additives
towards one rubber phase.
In Figures 2 and 3 the results of mechanical properties for the isolated rubber
compounds as well as for the four blends
are presented.
As seen in Figure 2, the mechanical
performance for NR compound is much
superior than for BR one. As for the
blends, intermediate values were found,
which were dependent on the preparation mode. In blend M2, vulcanization
of the BR phase was stimulated since
the additives were all incorporated into
this phase. A comparatively low value
of t90 and a relative protection of the other
phase against excessive cure resulting in
a good mechanical response was
achieved. In opposition to this, the incorporation of the additives into the NR
phase in M1 allowed an excessive cure
of this phase giving rise to a significant
degradation of the rubber. At the same
time, BR may have not been sufficiently
vulcanized and the net result was the
showed by this blend.
With regard to tear strength, Figure 3
shows that this property is not as dependent on the preparation mode as the tensile strength. The blend M2 is again
KGK Kautschuk Gummi Kunststoffe 56. Jahrgang, Nr. 1-2/2003
superior compared to the others. The
same trend is found for the strain at
break. The figure shows that this property
is much higher for NR than for BR. As for
the blends, as expected, intermediate values were found, with M2 having the best
performance which may be the result of
the short time used to vulcanize this mixture, then contributing to a lower level of
NR degradation.
Figure 3 also presents the values of
shore hardness. The results are very
close to each other and do not vary signicatively with the preparation mode.
It has been already stressed the superior properties of NR composition in comparison to BR. One of the reasons to
blend BR to other elastomers is the difficulty this rubber presents concerning
processability. By blending these two
rubbers, BR and NR, the properties obtained were inferior to those for NR but,
in comparison to BR, there have been improvements in performance as well as in
processability. The lower price of NR is
also a positive factor in favour of these
Scanning electron microscopy
Micrographs of the fracture surfaces of
cryofractured samples at low magnification scanning electron microscopy
(SEM) are presented in Figure 4. The materials show fracture surfaces with different topographic aspects. The characteristic features related to each compound-
Influence of Additives Incorporation Sequence on NR/BR Blend Properties
Fig. 4. SEM photomicrographs of NR/BR mixtures
ing mode indicate that the fracture behavior of these blends can be affected by
the processing conditions. High magnification microfractographs are shown in
Figure 5.
The blend M1 presents characteristic
elements of a mixed fracture mechanism,
with flat regions associated to localized
plastic strain areas and tearing patterns
(Figure 4a). At higher magnification, the
flat regions show the existence of parallel
cracks (Figure 5a), characterizing a material with lower plasticity.
The fracture surfaces of M2 and M3
samples present similar microscopic features (Figures 4b and 4c), with low surface roughness and the presence of striations, characterizing a more ductile fracture mechanism. The M2 mixture presents an increase in the striations amount
(Figures 5b and 5c) showing a better
plastic deformation capability than M3.
Concerning M4, a partially ductile fracture with tearing regions near the edges
and a coarse roughness region in the
central area (Figure 4d) is seen. At higher
magnification it is observed facets in the
central region (Figure 5d), indicating the
occurrence of a discontinuous fracture
These SEM features are in agreement
with the mechanical test results and support the measured tensile strength values.
Dynamic mechanical thermolanalysis
Fig. 5. SEM photomicrographs of NR/BR mixtures
Dynamic mechanical thermolanalysis
was also carried out on these materials.
Figures 6a and 6b shows the variation
of tan d with temperature. The measured
glass transition temperature, Tg, for BR
and NR compositions (Figure 6a) are
ÿ 89 8C and ÿ 45 8C, respectively. For
the blends, each one of them shows
two peaks in the curve of tan d versus
temperature (Figure 6b). The temperature
at which these peaks appear are closely
related to the glass transition temperature
of BR and NR, indicating that these
blends are incompatible. However, the
temperature at which the peaks show
up depends on the blend preparation
Blend M1 was found to present the
worst result of tensile strength which
was credited to the high value of t90.
This long vulcanization time would cause
KGK Kautschuk Gummi Kunststoffe 56. Jahrgang, Nr. 1-2/2003
Influence of Additives Incorporation Sequence on NR/BR Blend Properties
they would preferentially migrate to BR,
they still are present in the NR phase in
amounts that are higher in this blend
than in any of the others.
The highest values of Tg for the NR
phase were measured for M2 and M3,
suggesting that the mixing sequence
used to prepare these blends caused
less deterioration of this phase, in agreement with the better mechanical properties shown by these two blends.
The observation of DMTA curves also
gives information on material homogeneity. From Figure 6b, the width of the tan d
peak for the NR phase in M1 is the largest, which means that this blend is
the least homogeneous.
In the region corresponding to BR, it is
observed that Tg for M2 and M4 is very
close as well as the values of tan d (Figure 6b), which suggests that by adding
the whole amount of additives or half
that much, the degree of vulcanization
is not affected. It is known that polybutadiene, due to its structure, can form
¹spontaneous crosslinksª during vulcanization so that even small amounts of sulfur would be enough to provide appropriate levels of crosslinking, differently from
other diene rubbers [19]. In comparison
to the other two blends, the glass transition temperatures for M2 and M4 are
higher, which may be related to a larger
amount of crosslinks formed in these
blends because of the higher concentration of curatives in the BR phase. The
peaks are narrower in these blends, suggesting a higher degree of homogeneity
as compared to M1 or M3.
Fig. 6. Tan d as a function of temperature of the rubber compounds
degradation of NR phase, resulting in
scission of the macromolecular chains,
and affecting Tg. This can be seen in Figure 6b as the value of Tg for the NR phase
in this blend is shifted to a lower temperature. It is known from literature [7] that
overcure also affects tan d since the higher the degree of vulcanization, the lower
will be the magnitude of tan d, which is
also seen in Figure 6b for M1. These results show that when the additives are
all incorporated into NR, even though
KGK Kautschuk Gummi Kunststoffe 56. Jahrgang, Nr. 1-2/2003
In the preparation of rubber blends, the
addition sequence with which the components of the formulation are mixed
does have influence on the behavior of
the final material. This allows the adjustment of certain properties without changing the formulation. These differences in
properties result from different morphologies which are developed, caused mainly
by differences in curative contents present in each rubber, this affecting the
rate of crosslinks formation.
In the case of the NR/BR blends investigated, better mechanical performance
is achieved when vulcanization of BR is
favoured while, at the same time, NR is
preserved from an excessive cure.
Influence of Additives Incorporation Sequence on NR/BR Blend Properties
The authors are indebted to Petroflex InduÂstria e ComeÂrcio S.A. for the polybutadiene sample and to
Conselho Nacional de Desenvolvimento CientõÂfico
e TecnoloÂgico (CNPq) for financial support.
[1] A.S. Siqueira Filho and M. C. BoÂ: ¹Controle de
Qualidade na InduÂstria de Artefatos de Borrachaª, Manuais CNI (1985).
[2] B . Klei and J.L. Koenig, Rubber Chem. Technol
70 (1997) 231.
[3] R. Joseph, K.E. George, D.J. Francis and K.T.
Thomas, Int. J. Polym. Mater. 12 (1987) 53.
[4] R. Joseph, K.E. George and D.J. Francis, Int. J.
Polym. Mater. 12 (1988) 111.
[5] R. Joseph, K.E. George, D.J. Francis and K.T.
Thomas, Int. J. Polym. Mater. 12 (1987) 29.
[6] D.J. Hurston and M. Song, J. of Appl. Polym.
Sci. 76 (2000) 1791.
[7] M.G. Huson, W.J. McGill and P.J. Swart, J.
Polym. Sci. Polym. Lett. Ed. 22 (1984) 143.
[8] A.N. Gent and L.Q. Zhang, J. Polym. Sci.: Part
B: Polymer Physics 39 (2001) 811.
[9] V.G. Costa and R.C. Nunes, Eur. Polym. J. 30
(1994) 1025.
[10] L.A. Ultracki: ¹Polymer Alloys and Blends: Thermodynamic and Rheologyª, Hanser Publishes,
New York (1990).
[11] A.D. Thorn and R.A. Robinson: ¹Compound Designª in: Rubber Products Manufacturing Technology, Marcel Dekker Inc., New York (1994) Ch
[12] J.A. Brydson: ¹Natural Rubberª in: Rubber Materials and Their Compounds, Elsevier Science
Publishers LTD, New York (1988) Ch 4, 69.
[13] P. Boochathum and S. Chiewnawin, Eur. Polym.
J. 37 (2001) 429.
[14] C.L. Leutewiler, Borracha Atual 32 (2001) 34.
[15] B.L. Lee, Polym. Eng. Sci. 21 (1981) 294.
[16] S.A. Groves and A.J. Tinker, J. of Natural Rubber Res. 11 (1996) 125.
[17] P. Boochathum and W. Prajudtake, Eur. Polym.
J. 37 (2001) 417.
[18] American Society for Testing and Materials, section 9, vol. 9.01, Philadelphia, 1986. a. ASTM D
2084-81, ¹Standard test method for rubber
property ± Vulcanization characteristics using
oscillating disk cure meterª. b. ASTM D 41287, ¹Standard test method for rubber properties
in tensionª. c. ASTM D 624-86, ¹Standard test
method for rubber property ± Tear resistanceª.
d. ASTM D 2240-86, ¹Standard test method
for rubber property ± Durometer hardnessª.
[19] G.F. Bloomfield ¹Raw Polymeric Materialsª in:
Rubber Technology and Manufacture, C. M.
Blow, C. Hepburn, London (1982) Ch 4, 77.
The authors
Daniele Freitas de Castro, Ms. C. student at the Instituto de MacromoleÂculas Prof. Eloisa Mano, Universidade Federal do Rio de Janeiro, Agnes FrancËa Martins, Ph. D. student at the Instituto de Macromole¬'culas Prof. Eloisa Mano, Universidade Federal do Rio
de Janeiro, JoaÄo Carlos M. Suarez, Professor at Departamento de Engenharia MecaÃnica e de Materiais,
Instituto Militar de Engenharia, Regina CeÂlia Reis
Nunes and Leila LeÂa Yuan Visconte Associate Professors at Instituto de MacromoleÂculas Professora
Eloisa Mano, Universidade Federal do Rio de Janeiro.
Corresponding author
Leila LeÂa Yuan Visconte, Centro de Tecnologia, Bloco
J , P. O. Box 6 85 25, 21945-970 Rio de Janeiro, RJ,
Brazil, e-mail: [email protected];
11. 03. ± 12. 03. 2003
Technische Kunststoffe mit
hochwertigen Eigenschaften
Frankfurter Str. 15-17
97082 WuÈrzburg
Tel. 0931/4104-123
Fax 0931/4104-177
19. 03. ± 20. 03. 2003
Innovationen in der Kunststofftechnik
Institut fuÈr Kunststofftechnologie
BoÈblinger Str. 70
70199 Stuttgart
Tel. 0711/641-2344
Fax 0711/641-2335
26. 03. ± 27. 03. 2003
Altdorf bei NuÈrnberg
Elastomere als Dichtungsmaterialien
Technische Akademie Wuppertal
Hubertusallee 18
42117 Wuppertal
Tel. 0202/7495-0
Fax 0202/7495-202
26. 03. ± 27. 03. 2003
Frankfurter Str. 15-17
97082 WuÈrzburg
Tel. 0931/4104-136
Fax 0931/4104-177
24. 03. ± 28. 03. 2003
Extrusion von Kautschukmischungen
Eupener Str. 33
30519 Hannover
Tel. 0511/84201-10
Fax 0511/8386826
Thermoplastische Elastomere
KGK Kautschuk Gummi Kunststoffe 56. Jahrgang, Nr. 1-2/2003