Reinforcement of BIIR with silica.

Tire tread compounds containing butyl elastomers are known to have very good wet traction, but very poor abrasion resistance. Grafting living solution BR or SBR onto chlorobutyl (CIIR) in solution resulted in a decrease in DIN abrasion loss for elastomer blends containing (S)BR and CIIR grafts (ref. 1). This improvement is probably due to improved elastomer and/or cure compatibility resulting in an improved network. However, the process is prohibitively expensive on a commercial scale.

There are at least two other factors besides compatibility which have a direct impact on abrasion resistance:

* The degree and type of crosslinks in the BIIR phase; and

* the elastomer-filler interaction.

Crosslink density and structure

Crosslink density is known to affect the physical properties of the vulcanized elastomer. BIIR is unique in that it cures with either sulfur or ZnO alone. The addition of MBTS, which is utilized to enhance scorch safety, acts as a cure retarder, not as an accelerator, and reduces the crosslink density. Conversely, the addition of MgO to BIIR cured with ZnO + S + MBTS can increase the crosslink density.

This article briefly describes the impact of S, ZnO and MgO levels on the properties of BIIR containing silica.

Elastomer-filler interaction

The interaction between butyl elastomers and carbon black is poor. This is probably due to the very low level of C=C bonds in butyl. Addition of any other elastomer to a blend of (H)IIR and carbon black results in the migration of the carbon black to the other elastomer. However, the advent of silica technology for tire treads has provided a potential avenue to obtain much better butyl to filler interaction by coupling the elastomer and silica through the use of silanes. This technology is utilized in tire treads to lower rolling resistance (the so called "green tire"). BIIR is a very reactive elastomer, it is unique in that it can be cured by sulfur (figure 1) in the absence of any accelerators or activators, and it can also be cured with amines (figure 2) (ref. 2). The most common agents used to couple silica to elastomers are bis(triethoxysilylpropyl) tetrasulfide (TESPT) and bis(triethoxysilylpropyl) disulfide (TESPD). Several silanes containing amine groups such as 3-amino propyltriethoxy silane (APTES) are also commercially available. All these silanes are expected to more readily react with BIIR than with CIIR or IIR. If the reactions occur as envisaged, it may be possible to design butyl compounds with much better properties, including tire treads with a better balance of properties.

[FIGURES 1-2 OMITTED]

This article examines the effect of APTES, TESPD and TESPT on some of the most important tire tread properties, particularly abrasion resistance, and looks at some of the underlying chemical mechanisms.

Experimental details

Statistically designed experiments

Two series of experiments were run to determine (a) the optimum levels of S, ZnO, MgO and (b) the optimum levels of silica and silane to give good wear resistance. Experiments were designed and analyzed using JMP statistical software. A response surface - (3 variable) - Box- Behnken design was used; this calls for 15 experiments.

Optimization of S, ZnO, MgO contents

The compound formulation used was:
  BB2040                    100
  Precipitated silica       40
  Paraffinic oil            20
  Stearic acid              1

Four batches of these ingredients were mixed in a "B" internal mixer and blended on a mill. The resulting compound was then divided into 15 portions and different amounts of NBS S (0.5, 1 or 2 phr), ZnO (0, 1 or 3 phr) and magnesium oxide (0, 1 and 2 phr) were added on a mill at 40 [degrees] C.

Optimization of silica and silane contents

The amounts of precipitated silica (60, 80 and 100 phr), AFFES (4, 8 and 12 phr) and sulfur (0.5, 1 and 2 phr) were varied. The compound also contained l0 phr of paraffinic oil. One phr each of ZnO, MgO, and stearic acid were added to all the compounds together with the variable sulfur on a cool mill.

Effect of APTES/TESPD ratios on the reaction rate with silica and the properties of the BIIR vulcanizates

APTES has some processing advantages over TESPD, but also some property disadvantages, such as a high compound Mooney. A series of compounds was prepared to compare the combination of APTES + TESPD with APTES and TESPD alone.

Effect of different silanes containing nitrogen

The interaction of APTES with silica and BIIR is complex. A series of compounds with different silanes was produced to probe the reaction of the amino groups with the silica.

Mixing conditions

Three types of mixers were used:

* Mill 150 x 300 mm. The elastomer, silica, silane and oil were added on a cool mill and subsequently heat treated for 10 minutes at 110-125 [degrees] C. The curatives were then added on a cool mill.

* An internal mixer ("B"). Fill factor nominally 67%, 77rpm, cooling water at 40 [degrees] C. 0' BIIR + 1/2 silica + 1/2 silane; 1' 1/4 silica; 2' 1/4 silica + 1/2 silane; 3' brush chute; 6' dump (<150 [degrees] C). Rubber was refined on a 250 x 500 mm mill, and the curatives added on a cool mill.

* A miniature internal mixer (with 70 ml head); fill factor 67%, 60 rpm, start temperature 90-110 [degrees] C. Conditions were the same as for the larger mixer, apart from dumping after 6-15' with a dump temperature of 110-130 [degrees] C.

Testing

The compounds were tested for cure rate and state (MDR), stress strain, DIN abrasion and dynamic mechanical properties (Rheometrics RSA-2 or GABO).

Results and discussion

Optimization of the curative package

This study was conducted in the absence of a silane and with 40 phr of silica, 20 phr of oil and 1 phr of stearic acid. No accelerator was used. Sulfur was varied from 0.5-2 phr, ZnO from 0-3 phr, and MgO from 0-2 phr. The optimum curative combination to give the lowest DIN abrasion loss was predicted to be 2 phr of S + 2 phr of MgO with zero ZnO. This combination also predicted the highest tensile strength. There was a reasonable correlation between tensile strength and DIN ([R.sup.2] = 0.83). However, the cure reversion was high, thus a ZnO level of 1.0 phr was selected for further work to lower the cure reversion. Also, there was only a minor improvement in adding 2 vs. 1 phr of MgO, and 1 MgO was selected for further work.

Optimization of silica and silane contents in BIIR

Abrasion resistance as measured by DIN was used to optimize the silica (40-80 phr), APTES (4-12 phr) and sulfur (0.5-2 phr) contents in a recipe, which also included 10 phr of oil and 1 phr each of ZnO, MgO and stearic acid. The optimum levels were predicted to be as follows, with the silica and APTES having the largest effect on DIN abrasion loss:

BIIR - 100; paraffin wax - 10; precipitated silica - 60; APTES - 8; S - 0.5; ZnO - 1; MgO - 1; stearic acid - 1.

However, the optimum tensile strength was obtained with 1.2 phr S. Further work using BIIR/silica/APTES/oil of 100/60/8/10 showed that increasing sulfur from 0.5 to 1.0 phr significantly increased the tensile strength, with no discernible effect on DIN abrasion loss. Levels of sulfur above 1.0 phr did not improve the tensile strength. In subsequent experiments, the ZnO was increased to 1.5 phr to further reduce the cure reversion as observed on cure rheographs. Additionally, the MgO was removed as no benefit was observed in having it in the compound containing APTES.

The optimized recipe is:

BIIR - 100; precipitated silica - 60; paraffin wax - 0-10; silane - 8; S - 1; ZnO - 1.5; stearic acid - 1.

Chemistry of the reaction between APTES, silica and bromobutyl

Figure 3 gives the structure of the different silanes used in this study. 3-amino propyl triethoxy silane has the structure:

[H.sub.2]N-[(C[H.sub.2]).sub.3]-Si[(O[C.sub.2][H.sub.5]).sub.3]

Figure 3 - silica interaction - chemistry

3-amino propyl triethoxy silane (APTES) has the structure:

[H.sub.2]N-[(C[H.sub.2]).sub.3]-Si[(O[C.sub.2][H.sub.5]).sub.3]

BIS (triethoxysilylpropyl) tetrasulfane (TESPT, Si69) has
the "general" structure:

[[([C.sub.2][H.sub.5]0).sub.3] Si-C[H.sub.2]).sub.3]
-S-S-S-S-[(C[H.sub.2]).sub.3]-Si[(O[C.sub.2][H.sub.5]).sub.3]

TESPD (Silquest 1589) is the same as TESPT, but with
two (on average) fewer S atoms

One or more of the ethoxy groups react with the silanol groups on silica as follows:

-Si[(O[C.sub.2][H.sub.5]).sub.3] + Silica-OH [right arrow] -Si-O-silica + EtOH

It is anticipated that the amino group will react with an allylic bromine as shown in figure 4. Figure 5 is a pictorial representation of these reactions.

[FIGURES 4-5 OMITTED]

Chemistry of the reaction between TESPD or TESPT and bromobutyl

Figure 6 depicts the expected reaction between the silica and ethoxy groups of the silane and the BIIR and the S-S groups in TESPT (or TESPD). It is anticipated that BIIR will react with a S-S bond in the silane to form a linkage between the silane and BIIR (figure 5). An expected side reaction is the liberation of free sulfur from the silanes containing more than two S atoms in the molecule. This could result in premature curing during the mixing process. Initial experiments using the B internal mixer with the different silanes resulted in differences in mill shrinkage. APTES compounds had very little shrinkage, TESPD compounds had less than 10% shrinkage, but TESPT compounds had >30% shrinkage. It is theorized that the mill shrinkage was caused by release of some sulfur from the TESPD and TESPT during the mixing process. TESPT would be expected to release a higher amount of free sulfur because of the higher sulfur rank.

[FIGURE 6 OMITTED]

Effect of silane on the reinforcement of BIIR with silica

Figure 7 shows the stress strain curve for compounds containing APTES, TESPD, TESPT and no silane. There is a significant improvement in the reinforcement of BIIR with all three silanes. These compounds were mill mixed and subsequently heat treated for ten minutes at temperatures between 110 [degrees] C and 125 [degrees] C. Table 1 gives additional data for these compounds. Dynamic properties are also improved by the addition of a silane, especially tan [delta] at 0 [degrees] C (higher) and E" at 60 [degrees] C (lower).

[FIGURE 7 OMITTED]

Effect of internal mixer versus mill mixing

Comparison of a compound containing APTES prepared in an internal mixer using a six minute mix with a mill mix is shown in figure 8 and table 2. The reinforcement as shown by the stress strain curve and DIN abrasion loss appears to be practically identical. However, table 2 shows significant differences in both the M300/M100 ratio and the dynamic properties. ML is higher for the internally mixed compound. These data indicate that better dispersion was obtained on the mill, an alternative explanation could be lower scorch.

[FIGURE 8 OMITTED]

Effect of oil content and elastomer Mooney

Figure 8 and table 2 also show the effect of oil content and raw polymer Mooney on these properties. Removing the oil increases wear resistance and improves tan [delta] at 0 [degrees] C. There appears to be no significant impact of elastomer Mooney on vulcanizate properties. ML is higher for the higher Mooney grade as expected.

Effect of mixtures of APTES and TESPD

Figure 9 compares the relative reaction rates of APTES and TESPD. The graph in figure 9 shows the weight loss from a compound, which was prepared in a three minute internal mixer mix. It is assumed that very little reaction takes place in the initial mixing stage in the internal mixer. Following this three minute mix, 400 g of the compound was placed on a 150 x 30 mm mill for 40 minutes. The rubber temperature was 115 [degrees] C. The sample was removed from the mill at 5, 10, 20, 30 and 40 minutes and weighed. The weight loss is primarily due to water from the silica and ethanol evolution as a result of the reaction between the silica and the ethoxy groups. The difference in the curves for zero silane and with a silane is assumed to be due to loss of ethanol. The graph shows that APTES reacts much more quickly than TESPD. After 40 minutes at 115 [degrees] C, the differences in weight loss between no silane and 100% APTES and 100% TESPD indicate that 30% of the ethoxy groups have reacted in the APTES and 15% in the TESPD. It is noteworthy that a mixture of 50% APTES + 50% TESPD reacts as quickly as 100% APTES.

[FIGURE 9 OMITTED]

Table 3 shows a comparison of properties for compounds containing APTES, TESPD and 50% APTES + 50% TESPD. All the compounds contained the same amount of ingredients.

BIIR - 100; precipitated silica - 60; paraffin wax - 5; silane - 8; stearic acid - 1; sulfur - 1; ZnO - 1.5.

These compounds were mixed in an internal mixer. Table 3 gives the mixing times and initial temperature of the lab mixer with the final rubber temperature. The data in table 3 show that a 50/50 mixture of APTES + TESPD gives better wear and dynamic properties than either of the individual silanes.

The APTES compounds have much higher compound Mooney and minimum torque (cure rheometry), suggesting that the APTES is reacting to some extent with BIIR during the mixing process, even though the uncured compound produces a nice sheet when passed through a mill. The TESPD compounds have much lower compound Mooney than those with APTES, and the wear resistance is much better. However, the combination of APTES + TESPD gives the best wear resistance and dynamic properties (both higher tan [delta] at 0 [degrees] C and lower tan [delta] at 60 [degrees] C). The compound Mooney is higher than with TESPD alone. It may be possible to improve the data for the 100% TESPD compounds by mixing for a longer time. However, the TESPD compounds were mixed for much longer times than those containing APTES, 15 versus six minutes.

The complex modulus at low strain can be an indicator of filler dispersion. However, if some of the APTES has reacted with the BIIR this will raise G*. It is noteworthy that G* for the silane combination is very similar to that for TESPD, both are much lower than for APTES alone.

Mechanism of the reaction of APTES with silica and BIIR J.P. Blitz (ref. 3) has shown that amines catalyze the reaction of alkoxy groups with the silanol groups on the surface of silica. It is also known that APTES will cure BIIR readily in the absence of any fillers. The reaction rate of APTES with BIIR appears to be reduced by the presence of silica. It is hypothesized that some of the amino groups on the APTES hydrogen bond to the silica surface, thereby reducing the reaction rate. Some experimental support for this was obtained by examining a number of different silanes containing N atoms. Figure 10 shows that the cure of a silane containing three nitrogen atoms (triamino functional silane A-1130) is much slower than APTES. A possible explanation for the slower cure rate is a much stronger interaction of A-1130 with the silica surface. Conversely, a silane containing a N atom with a bulky phenyl group on the carbon atom adjacent to the nitrogen (N-phenyl gamma aminopropyltrimethoxy silane, Y-9669) is much faster. This supports the proposed hydrogen bonding hypothesis but is by no means conclusive evidence.

[FIGURE 10 OMITTED]

A-1100 and A-1110 are 3-aminopropyl triethoxy silane (APTES) and 3-aminopropyl trimethoxy silane, respectively.

Conclusions

Good reinforcement of a butyl vulcanizate has been achieved through the interaction of BIIR, silica and a silane coupling agent.

This improvement in elastomer reinforcement has resulted in a significant improvement in many properties, including tensile strength (>15 MPa), modulus (M300/100M up to 7), wear resistance as measured by DIN abrasion loss, and dynamic properties such as the increase in tan [delta] at 0 [degrees] C and reduction in E" at 60 [degrees] C.

APTES reacts faster with silica than TESPD, but results in a higher compound Mooney.

A combination of APTES and TESPD can provide superior properties.

Table 1 - effect of silanes on the reinforcement
of BIIR + silica

Silane (8 phr)                           0   APTES   TESPD   TESPT
Paraffinic oil (phr)                    10      10      10      10

MDR (3 [degrees] arc, 60' @ 170
  [degrees] C)
[M.sub.H](dN.m)                       22.8    38.4    34.5    32.8
MI                                    14.7    11.8     6.7     7.3
[T.sub.S]2 (min.)                      2.0     0.8     0.9     0.7
Durometer A2 (pts.)                     61      51      59      62
Tensile (MPa)                         6.59   15.20    16.2    16.9
Elongation (%)                         890     309     431     494
100% modulus                          0.87       2    1.83    1.59
200% modulus                          1.01    6.89    4.80    4.23
300% modulus                          1.39    14.5    9.94    8.88
M300/M100                              1.6     7.3     5.4     5.6
DIN (60 grit)                     Too soft     230     235     198
Tan [delta] 9 0 [degrees] C           0.32    0.77    0.66    0.66
Tan 6 [delta] 60 [degrees] C          0.10    0.15    0.16    0.16
E" @ 60 [degrees] C                   1.90    0.45    0.79    0.63

Compounds mill mixed at ambient temperature followed by 10
minutes heat treatment on a mill with the rubber temperature
between 110 and 125 [degrees] C
Table 2 - internal mixer vs. mill and effect of oil
and BIIR Mooney on properties of BIIR + 60 phr
silica + 8 phr APTES

BIIR                               BB2040   BB2040   BB2030    BBX2

Aminosilane (p hr)                      8        8        8       8
RPML(1+8@125 [degrees] C)              41       41       33      48
Paraffinic oil (phr)                   10       10        0       0
Mixer                                Mill       "B" internal mixer
Durometer A2 (pts.)                    51       56       63      64
Tensile (MPa)                        15.2     14.9    15.39    16.9
Elongation (%)                        309      331      246     271
100% modulus                          2.0     2.69     3.78     3.6
200% modulus                         6.89     7.04    11.12   10.57
300% modulus                         14.5    13.22
M300/M100                             7.3      4.9
M200/M50                              7.3      5.5      6.4     6.0
DIN (60 grit)                         230      226      149     148
Tan [delta] @ 0 [degrees] C          0.77     0.62     0.76    0.72
Tan [delta] @ 60 [degrees] C         0.15     0.11     0.11    0.11
E" @ +60 [degrees] C                 0.45     0.75     0.76    0.76
MDR (3 [degrees] arc, 60' at 170
  [degrees] C)
[M.sub.H] (dN.m)                     38.4     41.9     52.1    51.6
[M.sub.L] (dN.m)                     11.8     15.1     19.3    22.9

References

(1.) U.S.P. 5,264,494. Halogenated butyl rubber graft copolymers, C.H. Ho, and W. Hopkins, Polysar Rubber Corp.

(2.) Cure reactivity - a route to improved performance in halobutyl applications, J. Walker and W. Hopkins. Presented to the Australian PRI, October 1986.

(3.) The role of amine structure on catalytic activity for silylation reactions with Cab-O-Sil, J.P. Blitz, J. of Colloid and Interface Science, Vol. 126 (2) pp. 387-391.