Characterization of physical properties of SBR/carbon black masterbatch.

By Kazuhiro Sone, M. Ishida, M. Ishiguro and T. Hatta, Mitsubishi Kasei and Kunihiko Fujimoto, Shibaura Institute of Technology

Among the particle fillers which reinforce elastomers, carbon black is one of the most widely used fillers. It is known that physical properties of carbon black filled rubber vulcanizates are affected by the dispersion level of filler. One of the improved methods to disperse filler into elastomers is the wet dispersion system. Mitsubishi Kasei has been improving the quality and production process of WMB, an SBR/carbon black masterbatch produced by co-coagulation of SBR latex, carbon black and extender oil under the wet dispersion process. WMB is widely applied to tire tread and industrial products due to its lower power consumption of rubber mixing, better processability and vulcanized properties, compared with the ordinary dry-mixing method (refs. 1 and 2).

In previous work (ref. 3), the physical properties of WMB were studied from the viewpoint of carbon black dispersion and polymer-carbon black interaction. And the carbon black dispersion of WMB-mix in micro-scale was found to be better than dry-mix, as well as in conventional macro-scale measurements.

In this article, the physical properties of compounds using WMB are compared with dry-mixing compounds, and the characteristics of WMB are investigated by means of the spin-spin relaxation time, the fracture surface energy and the dynamic creep.

Experimental

Materials

WMB was produced by co-coagulation of SBR latex, carbon black and extender oil under the wet dispersion process. The amount of carbon black was varied from 5 to 90 phr. The compositions of WMB used in this study are listed in table 1.

Table 1 - WMB description and test recipe
WMB description      Dry-mix       WMB-mix
Polymer type                      SBR 1500
Carbon black type                   N220
Extender oil type                   Hi-Ar.
Carbon black PHR                 vari. (5 90)(a)
Extender oil PHR                         8
Test recipe, PHR
WMB                              vari. (113 198)
SBR 1500               100              -
Carbon black (N220)    0 90             -
Extender oil (HI-Ar.)    8              -
Stearic acid           1.0             1.0
Zinc oxide             3.0             3.0
Anti ozonant S-13      1.0             1.0
Sulfur                 1.5             1.5
Acc (NOB)              1.0             1.0
Mixing stage             3               2
(a) - the amount of N220 were varied from 5 to 90 phr

Compounding

Compoundings were carried out in an OOC internal mixer based on the test recipes listed in table 1. Compounds using WMB were prepared by the two stage mixing method, while dry mixing consisted of three stages. After sheeting with a two-roll mill (254 [symbol omitted] x 500mm), uncured and cured properties of the compounds were evaluated. Table 2 lists the results of composition analysis on the mixed compounds.

Table 2 - results of composition analysis on
            mixed compounds
  Dry-mix/Amount of
carbon black (C.B) and oil
No.    C.B    Oil       Vf
     [phr]   [phr]   (C.B) %
D-0     0      8        0
D-1     5     12        2.4
D-2    11      8        4.9
D-3    16      8        6.9
D-4    21      8        8.9
D-5    31      9       12.9
D-6    49     10       18.6
D-7    75     10       25.1
D-8    96      8       31.3
 WMB-mix/Amounts of
carbon black (C.B) and oil
No.    C.B     Oil      Vf
      [phr]   [phr]   (C.B) %
W-1     5      6        2.6
W-2    10      7        4.7
W-3    15      7        6.6
W-4    22     10        9.2
W-5    31     10       12.5
W-6    50     10       19.0
W-7    69      8       24.6
W-8    89      9       29.7

Viscoelastic properties

The loss tangent (tan [delta]) and dynamic modulus (E') were measured with a viscoelastic spectrometer under simple tension mode at 10 Hz, 10% static strain. The specimen measured 2 mm thick, 5 mm wide and 25 mm long.

Spin-spin relaxation time ([T.sub.2])

Pulsed NMR measurements were carried out with a pulsed NMR spectrometer, operating at 20 MHz. The solid echo sequence provided a good approximation to the free induction decay, from which the proton spin-spin relaxation time ([T.sub.2]) was obtained (refs. 4 and 5).

[Mathematical Expression Omitted]

i = L, M, S f L: mobility component; M: intermediate state component; S: immobility component

[Mathematical Expression Omitted]

t: time

S(t): free induction decay signal

[S.sub.0i]: signal amplitude of each component at t = 0

[mu]i: the parameter which has range 1 [less than or equal to] [[mu].sub.i] [less than or equal to

[T.sub.2i] : spin-spin relaxation time of each component

S [(t).sub.t=0],: signal amplitude at t = 0

In this study, the width of 90[degrees] pulse and the pulse interval were adjusted to be 3.5 and 11 [mu] sec., respectively.

Fracture surface energy

The energy to form a unit area of fracture surface neglecting a factor of elongation, i.e., "fracture surface energy" was measured with the apparatus (ref 6) shown in figure 1.

A vulcanized sample was cut by a razor blade edge attached to a pendulum having a certain potential energy. Fracture surface energy ([[gamma].sub.i]) was obtained from the fracture surface area (A) and the energy loss of the pendulum by equation 1 (ref 6).

1 [[gamma].sub.i]= (1/A) [mgh {1 - [([V/[V.sub.0]).sup.2]} - Wd. critical]

where : [[gamma].sub.i] = fracture surface energy

A = fracture surface area

[V.sub.0] = velocity before cutting

V = velocity after cutting

Wd.critical = const.

Dynamic creep

In this study, plastic deformation under the continuous cyclic loading is named dynamic creep. Dynamic creep measurements were carried out with a dynamic creep tester (ref. 7). Cylindrical vulcanized specimen was rotated continuously under the constant shear loading, and the displacement of plastic deformation was measured. The shear stress ([tau]) on each specimen was varied at five levels: 0.40, 0.55, 0.70, 0.85, 1.00 MPa. Revolution frequency was 0.84 Hz. The temperature of specimen was measured by a thermocouple set in the center of specimen through a slip ring, and was controlled to keep between 23 and 27[degrees]C by a vortex cooler. Normalized dynamic creep strain [delta] [gamma] (t) was defined by equation 2.

2 [delta] [gamma] (t) = ([gamma] (t) - [[gamma].sub.0]) / [[gamma].sub.0]

where: [gamma] (t) is a dynamic creep strain t sec. after loading, [[gamma].sub.0] is a dynamic strain immediately after the loading of shear stress (t = 0). Test samples used in this measurement were D-6 (Vf = 18.6%) and W-6 (Vf = 19.0%) vulcanizates listed in table 2.

Results and discussion

Carbon black dispersion and amount of bound rubber

Carbon black dispersion was evaluated with ASTM D-2663 test method, and the results are listed in tables 3 and 4. Though the number of mixing stage of WMB-mix is less than dry-mix, the carbon black dispersion levels of WMB-mix (WMB-mixed compounds) are higher in all range of particle concentration.

[TABULAR DATA 3 and 4 OMITTED]

Figure 2 shows the dependence of amount of bound rubber on carbon black concentration. The amounts of bound rubber in WMB-mix are larger than dry-mix (dry-mixed compounds) in all range of particle concentration, in spite of the shorter mixing time of WMB first stage. It seems that the carbon black in WMB-mix forms stronger interaction with polymer than in dry-mix.

Physical properties

Physical properties of dry-mix and WMB-mix are listed in tables 3 and 4. The results of abrasion resistance are shown as the indices, compared with a carbon black un-filled vulcanizates (Vf = 0). Among these properties, the dependence of tensile, abrasion and fatigue properties on particle concentration are shown in figures 3-7. As shown in these figures, the tensile strength at break, elongation at break, Lambourn abrasion resistance index and elongation fatigue life exhibited upword convex curves with an increase in particle concentration, and these properties of WMB-mix were superior to dry-mix in all range of particle concentration.

In this study, based on some analytical and mechanical measurements, these characteristics of WMB-mix were investigated and discussed.

Spin-spin relaxation time ([T.sub.2]

The polymer-carbon black interaction was evaluated by the measurements of spin-spin relaxation time in pulsed NMR spectroscopy. Proton spin-spin-relaxation time [T.sub.2] is an index of polymer molecular mobility. In this study, [T.sub.2] decay signal could be resolved into three components, [T.sub.2L] [T.sub.2M] and [T.sub.2S]. The long [T.sub.2] ([T.sub.2L]) Component is supposed to correspond to relatively free regions in the network chains, and the short [T.sub.2] [T.sub.2M], [T.sub.2S]) component to motionally constrained regions. The results of [T.sub.2] measurement are listed in tables 5 (non-vulcanization rubber) and 6 (vulcanized).

[TABULAR DATA 5 and 6 OMITTED]

Figures 8 and 9 show the dependence of [T.sub.2L] in non-vulcanized and vulcanized compounds on carbon black concentration, respectively. From these figures, it is recognized that the polymer molecular movement of vulcanized compound is restricted by the crosslinking between molecular chains, because [T.sub.2L] of vulcanizates showed smaller values compared with that of non-vulcanized, while the [T.sub.2L] decreased as the carbon black concentration increased. The decrease of [T.sub.2L] was more remarkable in WMB-mix than in dry-mix. It means the degree of restriction of the SBR molecular motion by carbon black is stronger in WMB-mix than in dry-mix. This indicates that the carbon black in WMB-mix has stronger interaction with polymer than in dry-mix.

The short [T.sub.2], which corresponds to the rubber phase of low mobility formed around carbon black surface, was resolved into two components, [T.sub.2M] and [T.sub.2S], having about 40 to 55, [mu] sec. and 10 [mu] sec. spin-spin relaxation time, respectively. The effects of carbon black concentration on these parameters showed the [T.sub.2M] in WMB-mix was slightly shorter than in dry-mix, while there was no difference of [T.sub.2S] between them. The [T.sub.2S] component seems to correspond to quasi-glassy state.

Fracture surface energy

The amount of energy required to form a unit area of fracture surface, i.e., "fracture surface energy" was measured. The results showed the dependence of fracture surface energy ([[gamma].sub.i]) on carbon black concentration. [[gamma].sub.i] exhibited an upward convex curve with an increase in particle concentration. And carbon black volume fraction at maximum [[gamma].sub.i] agreed with the turning point in the relation between particle concentration and elongation at break shown in figure 4. It is supposed that these phenomena are caused by a phase transition of particle-dispersed structure. In a system which contains a large amount of strong-reinforcing particles like carbon black, it was reported (refs. 8 and 9) that particle-dispersed structures exhibit a phase transition from random to ordered one, while the elastic properties change entropic to energetic state.

Besides it was found that the fracture surface energy ([[gamma].sub.i]) of WMB-mix was larger than that of dry-mix in all range of carbon black concentration. It seems that this result comes from the superior micro-scale dispersion of carbon black in WMB-mix, as well as good macro-scale dispersion, which was reported in our previous paper (ref. 3).

Dynamic creep

In this study, plastic deformation under the continuous cyclic loading is named "dynamic creep."

Carbon black filled vulcanizates form a heterogeneous structure by dispersion of particles and at the same time by dispersion of polymer chains which are in different thermodynamic states (ref. 10). When such a heterogeneous system is stimulated dynamically, change of interaction between the polymer chains in the micro-structure (A, C phase) and restructuring of the macro-structure element (composite structure of a particle and C- phase) make its dynamic properties non-linear and non-equilibrium. It was proposed (ref. 7) that the dynamic creep process accompanies two physical phenomena: Breaking of weak physical bonds between rubber molecular chains and transition to a stable structure of filler within matrix when there occurs no chemical reaction in the creep at room temperature. Above proposed reaction coordinate model (ref 7) and the typical dynamic creep process are shown in figures 10 and 11. In this case, the amount of dynamic creep is explained by the combination of (a) a dynamic creep induced by the break of weak physical bonds consisting of interactions between polymer chains (first creep) and (b) a creep induced by the transformation of filled particles structure into stable one (second creep). And it was reported (refs. 7 and 11) that the amount of first and second normalized creep strain are proportional to the amount of physical bonds and particles transited to a stable structure, respectively.

Figure 12 exhibits the results of normalized dynamic creep curves of dry-mix and WMB-mix compound in this study. As shown in this figure, the dynamic creep strain increased rapidly after imposing shear stress and reached a plateau temporarily. After that it increased again and then reached constant. This two stage increase pattern of dynamic creep was more emphasized when the shear stress level was high. The amount of first and second normalized\convergent creep strain, ([[delta][gamma].sub.1])[infinity] and [[delta][gamma].sub.2])[infinity], were estimated from the approximate equation proposed by Fujimoto (ref 7). To summarize, the amounts of first and second dynamic creep in WMB-mix are 90% and 75% of that in dry-mix. It suggests that the amount of breaking weak physical bonds of polymer chain interaction and the amount of carbon black which transits to a stable structure within matrix are smaller by 10% and 25% in WMB-mix than in dry-mix, respectively. Particularly, there is a significant difference in the latter, which indicates that WMB-mix system has more stable structures originally. It means that the amounts of carbon black arranged in stable state energetically are more in WMB-mix than in dry-mix. This proposal is also supported by the results of our previous study (ref. 12). which revealed the distributions of the distance between carbon black aggregates, the distance distribution for WMB-mix is sharp, on the other hand, dry-mix has broad distribution with longer distances. Furthermore, it is reasonably expected that these stable structures in WMB-mix would bring on longer fatigue life as shown in figure 8.

Summary and conclusions

The characteristics of WMB, SBR/carbon black wet masterbatch are discussed. The WMB-mix (WMB-mixed compounds) showed superior reinforcement and longer fatigue life compared with dry-mix (dry- mixed compounds). By some analytical and mechanical measurements, these features of WMB-mix were studied.

* From the relation of long spin-spin relaxation time ([T.sub.2L]) to particle concentration, it was found that the degree of restriction of the SBR molecular motion by carbon black was larger in WMB-mix than in dry-mix. It might be caused by the excellent micro-scale dispersion of carbon black in WMB- mix. Together with the results of [T.sub.2M] and bound rubber measurements, it is concluded that the carbon black in WMB-mix has stronger interaction with polymer than in dry-mix.

* The fracture surface energy in WMB- mix was larger than in dry-mixx in all range of carbon black concentration. It seems that this result comes from the better micro-scale and macro-scale dispersion of carbon black in WMB-MIX.

* The results of dynamic creep measurements indicated that the amount of carbon black in WMB-mix transited to a stable structure within matrix under shear loading was 75% of that in dry-mix. Therefore, the amounts of carbon black arranged in stable state energetically from the beginning are more in WMB-mix than in dry-mix, i.e., WMB-mix has more stable structures. The superior micro-scale dispersion of carbon black would relate to this feature of WMB-mix.

References

(1.) T. Sawanobori, Rubber Industry (Japan), 18, 1 (1982). (2.) K. Honda, International Seminar on Elastomers (Akron), October 1988. (3.) K. Sone, et al, Rubber World, 206, No. 1, 29, April 1992. (4.) K. Fujimoto, T. Nishi, Nihon Gomu Kyokaishi, 45, 640 (1972). (5.) J. O'Brien, E. Cashell, G.E. Wardell, V.J. McBrierty, Macromolecules, 9, 653 (1976). (6.) K. Fujimoto, et al, Nihon Gomu Kyokaishi, 59, 671 (1986). (7.) K. Fujimoto, et al, Nihon Gomu Kyokaishi, 62, 448 (1989). (8.) K. Fujimoto, et al, Nihon Gomu Kyokaishi, 61, 284 (1988). (9.) K. Fujimoto, et al, Nihon Gomu Kyokaishi, 62, 98 (1989). (10.) K. Fujimoto, T. Nishi, Nihon Gomu Kyokaishi, 43, 465 (1970). (11.) K. Fujimoto, et al, Nihon Gomu Kyokaishi, 59, 282 (1986). [12.] K. Sone, et al, Paper No. 57 Presented at ACS Rubber Division Meeting, Toronto, Canada (May 1991).