Crosslinking and reinforcement of silica/silane-filled rubber compounds.

Since the development of the coupling agent Si 69 (Bis-[triethoxysilylpropyl]polysulfide) at the beginning of the 1970s and, in particular, since the introduction of the green tire in 1992, the silica/silane reinforcement system has been a fixed component in rubber technology. In addition to its main application in car tire tread compounds with reduced rolling resistance and considerably improved wet traction, silica is also used in technical rubber articles. Here, the compounds display improved tear resistance, low hysteresis loss and low electrical conductivity. The use of light fillers also permits the manufacture of transparent and colored components.

While the development and optimization of the silica/ silane reinforcement system was at first based on empirical findings and effects, major advances have been achieved in recent years in the understanding of the principle of operation of this filler system (ref. 1). The work here is concentrated, firstly, on the mechanism of the silica/silane and silane/rubber coupling and, secondly, on the reinforcement mechanism of this filler system in different polymers.

Reinforcement by means of active fillers

The addition of active fillers to a rubber matrix leads to a considerably increased reinforcement. This greater reinforcement manifests itself in a rise in the modulus, leading, among other things, to higher strain values, greater tensile strength and lower abrasion. In addition to the crosslinking of the rubber matrix, a number of filler-specific contributions, first described by Payne (ref. 2), are responsible for this rise in the modulus.

Firstly, the addition of a solid, non-deformable and dispersive filler leads to an increase of inner tension in relation to the external tension applied to the sample. This so-called "hydrodynamic" effect causes a modulus increase independent of deformation. Since, with the addition of an active filler, a part of the deformable rubber matrix is immobilized by physical adsorption or chemical coupling to the filler on its surface and in its structure, a further deformation-independent modulus contribution, the so-called "in-rubber structure" is found. Also observed, in addition to this filler/polymer interaction, is a filler/ filler interaction which leads to the formation of a filler network, the strength of which is determined by the type of filler, its surface and structure and the filler load. The formation of the filler network and the additional immobilization of rubber in filler clusters leads, in the case of minor deformations, to a clear increase in the modulus. Under deformation, this network is successively broken down, leading to a reduction in the modulus (refs. 3-5). This modulus difference is generally known as strain softening or the Payne effect ([DELTA][G.sup.*]). The breakdown of the filler network takes place with a loss of energy, leading to the known reduction in elasticity and a greater heat build-up in dynamically loaded, filled samples. Figure 1 shows the different modulus contributions in relation to shear.

[FIGURE 1 OMITTED]

It is also known that the formation of the filler network and, in consequence, the Payne effect increases with a higher filler load and greater specific surface, as the inter-aggregate spacing of the particles decreases, rendering filler-filler cross-linking more probable (ref. 6). Figure 2 shows this relationship. Consequently, the Payne effect can be directly adjusted by selection of the filler load and of the particle surface area (ref. 5). It should be noted, however, that, because of the hydrodynamic effect, a change in the filler load also influences the modulus and the hardness.

[FIGURE 2 OMITTED]

The silica/silane reinforcement system

Like carbon black, precipitated silica is a nano-scale filler with primary particle diameters of l0 to 80 nm. In contrast to carbon black, the surface of silica is polar, possesses reactive silanol groups and is hydrolyzed. Because of the high polarity, interaction with apolar rubber is low (ref. 7). This leads to a strong tendency to agglomeration of the silica, resulting in the formation of a well-developed filler network. In the filler network, a part of the rubber is immobilized, so that it is not available for deformation (trapped rubber). As deformation increases, the silica network breaks down, releasing the trapped rubber. This leads, as shown in figure 3, to a reduction in the modulus (Payne effect). Therefore, particularly in the case of high deformations, the reinforcement of unmodified silica is low.

[FIGURE 3 OMITTED]

In contrast to silica, the strong physical interaction of the carbon black surface with the rubber leads to an immobilization of the polymer in the filler structure, and thus to a reinforcement, even in the case of major deformations. In order to obtain comparable rubber reinforcement with silica in an apolar rubber, a bifunctional silane must be used as the coupling agent (ref. 8). This silane first reacts with the silica during the mixing, and then, during vulcanization, forms a chemical bond with the rubber. This silica-polymer coupling leads, as with carbon black, to an immobilization of the rubber on the particle surface, and thus to a marked increase in reinforcement (ref. 5). Figure 4 shows the influence of the filler-polymer coupling on the modulus and the loss lector tan [delta]. As can be seen, particularly in the case of high deformations, the modulus value is higher than in the case of compounds without silane or with monofunctional silane (VP Si 216), which does not lead to coupling. The lower tan [delta] in the shear range studied also indicates a weakening of the silica-silica crosslinking, the breakdown of which results in a corresponding loss of hysteresis.

[FIGURE 4 OMITTED]

The use of unmodified silica, particularly in combination with a reinforcing carbon black, is always of advantage if the tear and tear resistances are to be improved. For example, a combination of carbon black with 5 to 15 phr silica is used in earthmover tires, conveyor belts and drive belts (refs. 9 and 10).

Silica/silane coupling

The chemical modification of the silica surface with silanes generally takes place during the mixing process. This modification leads to a shielding of the polar surface, which itself leads to better rubber compatibility, and thus to reduced silica-silica crosslinking (ref. 11).

Figure 5 shows a fall in Mooney viscosity with increasing quantities of alkyl silane and a plateau at high dosages, representing total hydrophobization of the surface.

[FIGURE 5 OMITTED]

The silanization reaction may be described as shown, in simplified form, in figure 6 (ref. 12). The primary reaction, which causes chemical coupling of the silane to the surface, must be fully completed in order to ensure high reinforcement. The secondary reaction is a condensation reaction between two adjacent silane molecules, and is considerably slower than the primary reaction. It requires water as a reaction partner. In general, this condensation reaction is not fully completed during the mixing.

[FIGURE 6 OMITTED]

A homogeneous silanization of the silica surface is necessary for optimum reinforcement. To this end, the following points should be observed when composing the compound:

* The addition of silica and silane should take place as early as possible in the mixing process, in order to ensure good dispersion of the silica and liar fullest possible completion of the silanization reaction. Optimum dispersion is achieved by using highly dispersible silica.

* The silanization reaction must take place simultaneously with or immediately alter the dispersion in order to achieve a modification of the surface now made accessible by the dispersion. The modification of the freshly dispersed silica reduces reagglomeration of the particles.

* The mixing temperature selected must be high enough for completion of the silanization reaction during the mixing process and for the ethanol formed to be expelled from the compound. On the other hand, the temperature selected must not be so high that a premature reaction occurs between the silane and the rubber (pre-scorch), the viscosity of the compound falls to such an extent that the dispersion is impaired and an intermolecular condensation reaction of the silane (polysiloxane formation) leads to a deterioration of the hydrophobization.

A summarized description of the influences to be considered in the mixing process is given in (ref. 13) and can be represented, as shown in figure 7.

[FIGURE 7 OMITTED]

As described above, silica has a strong tendency to form a filler network, which leads to a hardening of the compound. This process, also known as flocculation (ref. 14), takes place, in particular, at higher temperatures, such as those occurring during mixing and vulcanization.

A homogeneous silanization of the surface inhibits this agglomeration activity. Figures 8 and 9 compare the flocculation of a pure silica compound, a compound with silanized silica (propyltriethoxysilane) and a compound with carbon black N 234. As can be seen, the flocculation is pronounced in the unmodified silica and the carbon black; the alkyl modified silica shows the least inclination to reagglomeration.

[FIGURE 8-9 OMITTED]

The longer the alkyl chain of the silane, the better the shielding of the polar silica and, consequently, the greater the inhibition of filler-filler crosslinking. Figure 10 compares the flocculation of silica compounds with a short-chained alkyl silane and a long-chained alkyl silane. As can be seen, the complete silanization of the silica surface with the long-chained alkyl silane effectively suppresses the filler-filler crosslinking. For this reason, alkyl silanes are suitable as silica-specific, active processing additives (ref. 11).

[FIGURE 10 OMITTED]

Silane-rubber coupling

During the vulcanization process, the silane-rubber coupling takes place, in addition to the matrix crosslinking. In the case of sulfur vulcanization with the poly- and disulfide silanes Si 69 (Sx = 3.8) and Si 266 (Sx = 2.2), it could be shown that the two crosslinking reactions take place simultaneously and compete for the added sulfur (ref. 16). It is therefore not possible to vary the two crosslinking processes independently of one another. As the sulfur-functional silanes require additional sulfur for their activation, an increase in the quantity of silane leads to an increase in silane-rubber coupling at the expense of the crosslink density of the matrix. However, if the quantity of sulfur in the compound is increased, both the matrix crosslinking and the silica-rubber coupling are increased, leading to higher moduli, but also to appreciably shorter elongations at break. Figure 11 describes the influences of the silane and sulfur quantities on the matrix and silane crosslinking and the resulting crosslink densities. As an experimental proof, figure 12 compares the delta torque (measure of crosslink density) for the silanes (ref. 16).

[FIGURES 11-12 OMITTED]

The silica-polymer coupling particularly increases the "high-strain" modulus (strain values 100% and 300%). The influence of the quantities of Si 69 and sulfur on the strain value 300% is shown in figure 13. As can be seen, increases in the quantities of sulfur and silane both lead to higher strain values. It can also be seen that, at a constant strain value, the proportion of silica-rubber bonds can only be increased if the proportion of silane is increased and the proportion of sulfur is correspondingly decreased; thus, there is a shift from matrix crosslinking to silane coupling (ref. 16). This can he of particular importance when a component is to be optimized with regard to abrasion resistance, as in the case of tire tread compounds.

A corresponding sulfur correction is necessary when adjusting the silane quantity to the filler surface if comparable strain and elongation at break are to be achieved. If, however, while increasing the silica surface, the number of filler-polymer couplings are to be kept constant and an increase in the Payne effect is to be compensated, the addition of an alkyl silane is to be recommended for additional hydrophobization.

Variation of the silica surface

In the following study, four silicas with different surface areas (table 1) were examined in an S-SBR/1.4-BR compound (table 2). The objective was to determine the influence of the surface on in-rubber data. For this purpose, three different formulation settings were selected, using Ultrasil 7000 GR with a CTAB surface of 167 [m.sup.2]/g as reference.

1. 80 phr silica with 6.4 phr Si 69 and 1.5 phr sulfur. The DPG quantity was adjusted to the CTAB surface, in order, as far as possible, to obtain comparable vulcanization kinetics. The filler-filler crosslinking was also reduced in a compound with high-surface silica by the addition of 1.6 phr VP Si 216 (HDTEO) as hydrophobizing agent.

2. 80 phr silica with a silane quantity adjusted to the CTAB surface. Here, in addition, the quantity of sulfur was adjusted in accordance with the quantity of silane, and the addition of DPG was adjusted to the surface.

3. The degree of silica filling was adjusted in accordance with the CTAB surface such that the surface of the filler introduced remained constant. The silane and sulfur quantities were also held constant.

Formulation setting 1

In this formulation setting, the quantities of sulfur and silane were held constant. Figure 14 shows the Payne effect curves and tan [delta] curves of the vulcanizates (RPA measurement: Strain sweep 0.28-42%, 1.6 Hz, 60[degrees]C) for the compounds. As expected, there is a marked increase in the low-strain modulus and, correspondingly, in the Payne effect, with an increase in surface (higher silica network). In view of the smaller interaggregate spacing as the surface increases (figure 2), this was to be expected. At high deformations, the moduli rise slightly as the surface is increased. This is probably attributable to the fact that, even at 42%, there is still not a complete breakdown of the silica-silica crosslinking, which still makes a corresponding contribution. The addition of the alkyl silane VP Si 216 to the high-surface silica, Silica III, produces a Payne effect similar to that of Silica II. The loss factor tan [delta] also rises with an increase in the surface. This is to be expected because of the breakdown of the more stable crosslinking. It is interesting to note, in this connection, that as the surface increases, the maximum of tan [delta] is shifted to higher elongations. This shift is not observed in the case of carbon black.

[FIGURE 14 OMITTED]

Selected rubber specifications of these compounds are listed in figure 15. As expected, with an increase in surface, the compound viscosity rises and the durometer A hardness (low-strain modulus), which is largely influenced by the filler network, also rises correspondingly. The strain value 50% also rises as the surface increases, while, at 300% elongation, comparable values are measured. Here, the crosslink density and the silica-polymer coupling are decisive. With increasing hardness, there is a marked improvement in DIN abrasion. The heat build-up, measured in the Goodrich flexometer (constant deformation: 0.25 inch stroke), increases appreciably, as the deformation is sufficient for a hysteresis-rich breakdown of silica network. Because of the increase in silica network with an increase in surface, an increase in dynamic stiffness E* is also observed at 0[pounds sterling]C and 60[pounds sterling]C (constant force: 50 N preforce, 25 N ampl. force, 16 Hz). This is accompanied by an increase in tan [delta] (60[degrees]C) and a reduction of the ball rebound value at 60[degrees]C (values not shown). On the other hand, the ball rebound rises at 0[degrees]C because the silica crosslinking is more pronounced at the lower temperature and does not break down completely. Noticeable in the case of the compound with added VP Si 216, apart from the expected reduction in the viscosity, is the similarity of the vulcanizate data to the compound with Silica II. An advantage is the lower DIN abrasion, possibly because of the smaller diameter of the primary particles of Silica III.

Formulation setting 2

In this series, the quantity of Si 69 was adjusted to the CTAB surface of the silica, and the quantity of sulfur was selected such that the compounds contained a constant proportion of mobile sulfur (elemental sulfur + polysulfidic sulfur in Si 69). Figure 16 shows the results of the RPA measurement on the vulcanizate. Here too, an increase in the Payne effect is seen with increasing surface, but because of the silane adjustment (comparable hydrophobization pro [m.sup.2]) the differences are smaller than those for setting 1.

With the decrease in inter-aggregate distances as the surface is increased, an increase in the Payne effect was also to be expected here. The tan [delta] curve also behaves in line with the Payne effect.

In figure 17, selected in-rubber data are compared. In comparison with the data in figure 15, it can be seen that an increase in the quantity of silane as the surface increases leads to better hydrophobization, which manifests itself in a reduction in Mooney viscosity, a reduction in hardness and lower dynamic rigidity values in the compound with Silica III. These improvements indicate that an adjustment of the silane quantity when using high-surface silica is beneficial. There are also advantages for Silica III in its DIN abrasion and tensile strength. The extent to which the smaller diameter of the primary particles in Silica III has a positive influence on road wear remains to be studied. However, the silane reduction in the case of the low-surface silica, Silica I, leads to lower strain values, reduced hardness and appreciably worse DIN abrasion in comparison with recipe setting I without silane adjustment. This is certainly due to the reduction in silica-polymer coupling, resulting in road abrasion losses.

[FIGURE 17 OMITTED]

Formulation setting 3

In this setting, the degrees of filling were adjusted to the silica surface, and the silane/sulfur content was held constant. The RPA measurements on the vulcanizate now show quite a different picture (figure 18) than in the preceding studies. The greatest Payne effect is shown by the compound with the higher filler load of Silica I. Here, the filler load has a noticeable influence, as shown in figure 2. The lowest filler network is found in Silica III with the reduced filler load. The tan c3 curves show an increase with a rise in the Payne effect.

[FIGURES 2, 18 OMITTED]

The in-rubber data in figure 19 are significantly poorer for the compounds with Silicas II and III, in which the filler load was reduced. In particular, the strain values, dynamic rigidity values and DIN abrasion all deteriorate, which renders this recipe setting impractical for use with a higher surface silica. On the other hand, the compound with the increased filler load of Silica I shows a positive picture with regard to the in-rubber data. In comparison with the values in figures 15 and 17, it shows higher strain values, improved DIN abrasion and moderate dynamic rigidity values with acceptable hysteresis behavior. However, the positive effect of the tan [delta] reduction when using a low-surface silica instead a standard tread silica is lost.

Conclusion

The aim of this article is to provide applied researchers and compounders an understanding of the mechanism of the silica/ silane reinforcing system and how to adjust this system properly to their requirements. It is shown that the variation of the amount of silane and sulfur can be used to find best the balance between static and dynamic properties. Furthermore, the consequences of a changed silica surface area on major in-rubber data are given. It is suggested that the use of a low surface area needs slightly higher amounts to maintain reinforcement, while in the case of a silica with a higher surface area, the amount of silane needs to be increased. Due to these possible variations in the formulation, the silica/silane reinforcing system is still an exciting and challenging tool to meet the customer demands.

Table 1--silicas

                        CTAB            DBP
                    [m.sup.2]/g     [m.sup.2]/g

Ultrasil 7000 GR        167             210
Silica I                148             255
Silica II               181             254
Silica III              222             239

Table 2--formulation

1. Stage
Buna VSL 5025-1                           96
Buna CB 24                                30
Silica                           [+ or -] 80
Si 69                            [+ or -] 6.4
ZnO 3; stearic acid 2; oil 10
6PPD 1.5; wax 1
2. Stage
3. Stage
Batch stage 2
DPG                              [+ or -] 2
CBS                              [+ or -] 2.5
TBzTD                                     0.2
Sulfur                           [+ or -] 1.5

References

(1.) H.-D. Luginsland, "Chemistry and physics of network formation in silica-silane-filled rubber compounds," paper F presented at the ACS meeting May 2002, H.-D. Luginsland, "A review on the chemistry and the reinforcement of the silica-silane filler system for rubber applications," Shaker Verlag, Aachen 2002.

(2.) A.R. Payne and R.E. Whittaker, "Low strain dynamic properties of filled rubbers," Rubber Chem. Technol. 44, 440 (1971).

(3.) M.-J. Wang, "Effect of polymer-filler and filler-filler interaction on dynamic properties of filled vulcanizates," Rubber Chem. Technol. 71, 520 (1998).

(4.) J. Frohlich, H.-D. Luginsland and W. Niedermeier, "Reinforcement mechanism in the rubber matrix by active fillers," paper no. 9 presented at the ACS meeting, April 2000.

(5.) H.-D. Luginsland, J. Frohlich and A. Wehmeier, "Influence of different silanes on the reinforcement of silica-filled rubber compounds," paper no. 59 presented at the ACS meeting, April, 2001; Rubber Chem. Technol. 55, 563 (2002).

(6.) M.-J. Wang, S. Wolff and E.-H. Tan, "Filler-elastomer interactions, part VIII. The role of the distance between filler aggregates in the dynamic properties of filled vulcanizates," Rubber Chem. Technol. 66, 178 (1993).

(7.) M.-J. Wang and S. Wolff, "Filler-elastomer interactions. part III. Carbon-black-surface energies and interactions with elastomer analogs," Rubber Chem. Technol. 64, 714 (1991).

(8.) S. Wolff, "Chemical Aspects of Rubber Reinforcement by Fillers," Rubber Chem. Technol. 69, 325 (1996).

(9.) T.A. Okel and W.H. Waddell, "Silica properties/rubber performance correlation. Carbon black-filled rubber compounds," Rubber Chem. Technol. 67, 217 (1994).

(10.) N.L. Sakhnovaskii, L.I. Stepanova, V.F. Evstratov and R.V. Dunaeva, "Dependence of the resistance to chipping of tread vulcanizates on their composition," Kauch. Rezina 30 (10), 30 (1971).

(11.) A. Hasse and H.-D. Luginsland, "Influence of alkylsilanes on the properties of silica-filled rubber compounds," paper presented at the RubberChem '01 Conference, April, 2001.

(12.) U. Gorl, A. Hunsche, A. Muller and H.G. Koban, "Investigations into the silica/silane reaction system," Rubber Chem. Technol. 70, 608 (1997).

(13.) H.-D. Luginsland, "Processing of silica/silane-filled tread compounds," paper no. 34 presented at the ACS meeting, April, 2000.

(14.) M. Gerspacher, L. Nikiel, C.P O'Farrell, G. Schwartz and S. Cerveny, "Carbon black flocculation in rubber. Uncured compounds," paper presented at the Intertech Conference Functional Tire Fillers 2001, January 2001.

(15.) J. Frohlich and H.-D. Luginsland, "RPA-studies into the silica/silane system," Rubber World 244, 28 (2001).

(16.) Andre Hasse, Oliver Klockmann, Andre Wehmeier and H.D. Luginsland. "Influence of the amount of di- and polysulfane silanes on the crosslinking density of silica-filled rubber compounds," Kautsch. Gummi Kunstst. 55, 236 (2002); paper no. 91 presented at the ACS Meeting, October: 2001