Effects of filler on compatibility of NR.

NR has been studied and reported on extensively because of its superior performance in tire applications. So, there are many studies of NR and its blend compounds.

The microstructure of NR has been defined by the ozonolysis - GPC method (refs. 1-3). NR has a functional group on its terminal and the microstructure is a highly-continuous cis-1, 4 unit. Therefore, NR has stronger tensile strength than synthetic rubbers for its crystallization by orientation.

We have conducted many trials to get the same or better performance as with NRs. Chemically modified polyisoprene rubber (IR) has better green strength than IR. However, its physical properties were not sufficient and its compound viscosity was too high to mix with carbon black.

On the other hand, polymer blend with NR has been studied for the same reasons. Butadiene rubber (BR) with less than 60 wt% vinyl content (ref. 4), styrene-butadiene rubber (SBR) with more than 15 wt% styrene content (ref. 5), and isoprene rubber (IR) with more than 20 wt% 3,4-microstructure are incompatible with NR (ref. 6). These results suggested that the [Pi]-electron density of the side-chains is strongly affected by their compatibility with NR, but there is no evidence to explain this phenomenon.

Because of environmental pollution concerns, the auto industry must find ways to lower fuel consumption. Tire and car manufacturers need more low rolling resistance (LRR) and high abrasion resistance. Silica-filled compound is suitable for the LRR. Michelin (with Rhone Poulenc) developed the first "green tire." Silica was not always suitable for tire applications, its performance was worse than with carbon blacks (CBs). In its quest for lower rolling resistance than CBs, Michelin developed the silica tire. Basically, polymers with non-dipole moment have less compatibility to silica, so general-purpose rubbers are difficult to mix with silica. For this reason many silicas have been developed, for example surface modified silica, methyl or phenyl modification and fumed silica (Aerosil). But these fillers cannot bond to polymers, so the polymer was not reinforced. As a result, these compounds have no merit relative to one reinforced with carbon black. This suggested that improvement of physical properties needed both good dispersion of the filler and a strong bond to the polymer, like a covalent bond.

As above, in a polymer blend compound, if the two polymers were not compatible to each other, it would show poorer physical properties than the additivity of configurational contribution. This happens because of phase separation of each polymer

From the above discussion, it seems that the most important issue was control of compatibility.

In this study, regarding controlling the compatibility, we have found that the polymer's compatibility with NR is changed by the filler. Its behavior was closely related to the amount of filler, silane coupling agent and cure.

Experimental

Mixing

Polymer, reinforcing filler and other additives were mixed by a 0.25-liter type internal mixer at 80 [degrees] C (CB) and 120 [degrees] C (silica). The compounds were mixed for 5.5 minutes. In the case of CB, polymers were mixed for 30 seconds at 80 [degrees] C and CB and others were put into the mixer, and dumped after five minutes. The temperature was less than 140 [degrees] C. In the case of silica, polymers were mixed the same way, then one-half weight of silica, silica coupling agent and diethylene glycol were put into the mixer and mixed for 1.5 minutes. After that, others were put into the mixer and mixed about 2.5 minutes longer until they reached 160 [degrees] C.

Each compound was mixed with sulfur and accelerator in an open mill at 50 [degrees] C. After that the compounds were cured at 160 [degrees] C x 15 minutes (CB), and 160 [degrees] C x T90 minutes (silica).

Differential scanning calorimetry (DSC)

Thermal properties were measured by DSC (Perkin-Elmer DSC7) at a heating ratio of 80 [degrees] C/ min., and at a range of temperature from -120 [degrees] C to 30 [degrees] C.

Properties

Rheological properties were measured by an RDA-II (Rheometric Scientific Inc.), the temperature range from 120 [degrees] C to 100 [degrees] C at 125.7 rad. per second (= 20 Hz). The heating ratio was about 4 [degrees] C per minute. Physical properties were measured in accordance with JIS K6301.

Characteristics of polymers

Polymers were characterized by GPC, infrared spectroscopy. Molecular weight was determined by being referred to standard polystyrene in THF. The coupling ratio was calculated from the area of GPC curve. Microstructure of polymers was measured by infrared spectroscopy and calculated by the Hampton method (table 3).

Table 3 - details of microstructure by infrared spectroscopy

Name       ([dagger])Styrene   1,2-vinyl
                 mole %           mole
                 (wt%)           %(wt%)

SBR(*) 1      12.2 (21.1)      62.3 (56.0)

SBR(*) 2      13.7 (23.5)      32.5 (28.9)
BR(*)  1         0 (0.0)        7.5 (7.5)

Name             1,4-cis      1,4-trans
                 mole %         mole %
                  (wt%)         (wt%)

SBR(*) 1         6.9 (6.2)   18.6 (16.7)

SBR(*) 2        14.5 (16.4)   37.1 (32.9)
BR(*)  1        35.7 (35.7)   56.8 (56.8)

(*) The Hampton method

Polymerization of SBR1

An autoclave equipped with a stirrer was charged with 8.000 grams of cyclohexane, 360 grams of styrene and 720 grains of butadiene, followed by an addition of 20 mmol of TMEDA (tetramethylethylene diamine) and 13 mmol of n-BuLi to initiate a polymerization at 50 [degrees] C. After the conversion had reached about 20%. 860 grams of butadiene and 60 grams of styrene were continuously added over a 60-minute interval by pump. After the conversion was almost complete-ly perfect. 0.5 grams of butadiene were added.

When the monomer was completely polymerized, about 5.2 mmol of tin tetrachloride (SnC14) was added and reacted with living polymer for 15 minutes. As a shortstop. 20 mmol of isopropyl alcohol was added and 2 grams of Irganox-1520 were added. SBR was coagulated by a steam-stripping method and vacuum dried for 24 hours. The polymer was characterized by infrared spectroscopy and GPC. SBR2 and BR1 were produced the same way.

Results and discussion

Physical properties

Characteristics of fillers are shown in table 1. SMR-CV60. solution polymerized butadiene rubber (BR1). and SBR (SBR1 and SBR2) were used. Characteristics of the polymer are shown in table 2.

Table 1 - filler properties

Filler               BET           Primary     Characterization
                     [m.sup.2]/g   particle
                                   diameter
                                     (nm)

Seast(*) KH              93           24                HAF black
Seast 6                 117           22               ISAF black
Seast 9                 142           19                SAF black
Ultrasil VN-2G(**)      125           20      precipitated silica
Ultrasil VN-3G          175           17      precipitated silica
Nipsil AQ(***)          200           16      precipitated silica

(*) The CB of Seast series are produced by Tokai Carbon K.K.

(**) The silica of Ultrasil series are produced by Degussa.

(***) AQ is produced by Nippon Silica K.K.

Table 2 - characteristics of polymer

Name        Bound    1,2-unit in    Average
           styrene    butadiene    molecular
             %          mole %     weight Mw
                                    X 1,000

SBR(*) 1    21.1         71.0        42.5
SBR(*) 2    23.5         33.4        43.5
BR(*)  1     0.0          7.5        45.5

Name              Coupling     Coupling
                  efficiency     agent
                     %

SBR(*) 1            37.8        SnC14
SBR(*) 2            36.8        SnC14
BR(*)  1            38.4        SnC14

(*) The Hampton method

Physical properties of these compounds are shown in table 4. Each compound was cured at 160 [degrees] C x T90 min. The relative values on abrasion test were given in the last column of each table. Abrasion value of sample 3 was defined as 100.

[TABULAR DATA 4 NOT REPRODUCIBLE IN ASCII]

Physical properties of various silica compounds are shown in table 4. Sample 1 shows higher tensile properties, but lower rebound resilience at 60 [degrees] C. These data indicate that the silica compounds are not superior to the carbon black compound (samples 6-8). Although samples 1-3 include more filler then sample 4, sample 4 has higher hardness and higher rebound resilience than samples 1-3 because it includes carbon black as filler. This result seemed to indicate that silica compound has higher performance in tire tread than CBs. But compound viscosity of silica compound was far higher than carbon blacks. This is the most significant problem with the silica compound.

The effect of filler blend, silica (VN-3G) and carbon black (HAF grade), are also shown in table 4. When the carbon black ratio was increased, although the tensile properties and abrasion resistance were improved slightly, the rebound resilience became low, and it has no effect on rolling resistance. Half of the silica is replaced by carbon black (sample 9).

Physical properties on effect of polymer are shown in table 5. Characteristics of polymer are shown in table 2. Sample 10. high vinyl SBR*1(HV-SBR): NR=70:30 by weight ratio, had a better balance of rebound resilience at 60 [degrees] C and at 23 [degrees] C than NRs (sample 14).14V-SBR showed a different behavior from medium vinyl SBR2 (MV-SBR). HV-SBR blended with NR had much different physical properties (hardness at -10 [degrees] C, rebound resilience and abrasion resistance) from HV-SBRs. On the other hand, MV-SBR blended with NR showed almost the same properties as MV-SBRs. It seemed that the compatibility of HV-SBR with NR was different from that of MV-SBR.

Table 5 - physical properties (ref. 4) effect on microstructure of
polymer

Component/property           10      11      12      13     14
SBR(*) 1                     70     100
SBR(*) 2                                     70     100
BR(*) 3                      30
SMR-CV60                                     30            100
Silica VN-3G                 60      60      60      60     60
Compound ML
 (1000C 1+4)               92.0   104.7   101.2   114.7   77.4
ODR test@160 [degrees] C
 Vmax - Vmin kg-cm         54.6    49.1    50.9    51.2   43.4
 T90 min                   23.1    35.0    17.9    21.5   11.0
Tensile test (cured at
160 x T90 min)
 Ultimate strength MPa     15.7    15.3    21.1    18.1   27.1
 Elongation at break %      410     340     490     400    620
 100% modulus MPa           2.9     3.5     2.9     3.1    2.1
 300% modulus MPa           9.9    12.9    10.4    11.5    8.9
Hardness Shore A
 At -10 [degrees] C        83.7    93.4    74.6    78.6   77.4
 At 23 [degrees] C         73.7    73.6    69.7    73.6   73.3
 At 80 [degrees] C         67.3    66.4    65.8    67.6   67.1
Rebound resilience
(Lupke %)
 At 23 [degrees] C         50.2    28.2    54.3    53.2   53.0
 At 60 [degrees] C         64.5    61.5    66.3    65.8   60.5
Pico abrasion test
Index                       115     100     107     109    116

NR compound with silica showed poor physical properties relative to the one with carbon black. It was reported that NR was degraded when mixed with silica (ref. 3). chain was broken during mixing with silica at cis 1,4 microstructure. Therefore. NR with silica has higher hysteresis loss than that with carbon black. Usually silica compounds were characterized by low hysteresis loss. Only NR had an exceptional result. Sample 11 (HV-SBR) and Sample 14 (NR) showed a low rebound resilience at 60 [degrees] C. However, sample 10 (HV-SBR/NR) showed higher rebound resilience than samples 11 and 14, although sample 10 is composed of the same materials as sample 11 with 14. This suggested that HV-SBR was compatible with NR. On the other hand, the same SBR reinforced with carbon black, table 4, showed lower rebound resilience at 60 [degrees] C than sample 10 reinforced with silica. And their rebound resilience at 23 [degrees] C was significantly different from the silica compound.

Thermal properties

In order to confirm the compatibility, the compound was measured by DSC. A thermogram of NR/HV-SBR, 100/0.70/30,-30/70,0/100, is shown in figure 1. Each compound had a different Tg, so these systems were compatible to each other. By using MV-SBR, it had almost the same Tg as the original Tg. This seems to suggest that NR is incompatible with HV-SBR. The thermogram of samples 4-6 is shown in figure 2. These had two different Tgs. It is a very interesting result that the compatibility with NR was not changed by any carbon blacks.

[Figures 1-2 ILLUSTRATION OMITTED]

A thermogram of samples 7-9 is shown in figure 3. With the increasing ratio of carbon black, Tg began to show two positions clearly. When the blend contains equal parts of silica and carbon black, there are two Tgs and HV-SBR, and they therefore become incompatible with NR. These results suggested that the change of compatibility with NR was caused only by the existence of silica. In addition to the above experiments, we prepared the sample various blend ratios of SBR reinforced with 50 phr of HAF carbon black. The thermogram of these compounds is shown in figure 4. The blend compounds had two Tgs, and the Tg was almost the same value of the original Tg for both HV-SBR and NR. For all experiments, it was confirmed that the phenomena are characteristic of silica.

[Figures 3-4 ILLUSTRATION OMITTED]

It has been reported that SBR is compatible with BR after vulcanization (ref. 9). The thermograms of vulcanized and unvulcanized compound are shown in figure 6. The compound before cure had two Tg, so before cure, it was suggested that SBR was not compatible with NR. Generally Si-69 is used when mixing with silica for tire applications. The role of Si-69 was investigated; the compounds mixed with Si-69, silane-coupling agent, and without Si-69 were measured by DSC. The thermograms are shown in figure 6. The compound without Si-69 had two Tgs.

[Figure 6 ILLUSTRATION OMITTED]

Effect of silica

A G' vs. G" plot of samples 1. 2 and 3 was also conducted. The smallest surface area of silica, VN-2G, showed clearly two relaxations. Although VN-2G was covered with Si-69 more completely than VN-3G, the compound did not have single relaxation. It was considered the reason that particle size and surface activity are different from each other. For the same mason, sample 1 had two relaxations.

Microstructure

MV-SBR did not show this phenomenon. Ms result suggests the 1,2-vinyl contents of SBR were the most important factor in this phenomenon. Even if SBR were mixed with NR. SBR would not be compatible with NR. Since HV-SBR was compatible with NR in existence of silica, the compatibility is strongly related between silica and the density of the 1.2-vinyl unit.

Silica has strong activity on its surface, so it could be supposed that the 1,2-vinyl unit had a great affinity for silica on its surface. But this was not true according to the following experiments. Effects of silane coupling agent are shown in figure 3. The HV-SBR was starting to be incompatible with NR without silane coupling agents. This result suggested that SBR was bonded with NR by both silica and silane coupling agents. It is known that silane-coupling agents are effective in lording silica to NR and that NR has a strong affinity to silica (ref. 8).

Conclusions

It was found for the first time that the compatibility of NR is changed by the type of filler The effect of some silicas, some polymers, silane coupling agents and cure were studied. It was made clear that the compatibility is improved under the following conditions:

* Filler is silica;

* BET surface area of fillers is larger than about 150 [m.sup.2]/g;

* cure Pipe;

* polymer is HV-SBR; and

* silane coupling agent.

It was closely related to the 1,2-vinyl unit, but the relationship between a compatibility of NR and the surface activity, amount of coupling agents and microstructure was not clear.

[Figure 5 ILLUSTRATION OMITTED]

References

(1.) H. Sato and Y. Tanaka, Gomukyoukai-shi (Japan), 54 (9), 52(1981).

(2.) Y Tanaka and H. Sato, Polymer, 17, 413 (1976).

(3.) Y. Tanaka, H. Sato, et al., Koubunsi-gakkai-preprints (Japan), 29 (9). 2055 (1980).

(4.) I. Furuta, Textbook of a 35th fall school produced by Japanese Society of Rubber, 28 (1994).

(5.) A. Ueda, H. Watanabe and S. Akita, IRC '85, Kyoto 16A09, 199-204 (1985)

(6.) A. Yoshioka, K. Komuro, A. Ueda, H. Watanabe, et al., Pure & Appl. Chem. 58 (12), 1697 (1986).

(7.) I. Furuta, I. Hattori, et al., presented at the 137th ACS Rubber Div. Meeting, Las Vegas, NV (1990).

(8.) M. Ito, et al., Gomu-kyoukai-shi (Japan), 58 (7), 468 (1985).

(9.) J.E. Callan, W.M. Hess and C.E. Scott, Rubber Chem. Technol. 44, 814 (1971).