PROCESS FOR THE PRODUCTION OF A PRECIPITATED SILICA SLURRY WITH A CONTROLLED AGGREGATE PARTICLE SIZE DISTRIBUTION In one of its aspects, the present invention relates to a process for the production of a precipitated silica slurry. In another of its aspects, the present invention also provides a means of controlling the average sizes of the silica particle clusters or aggregates within this slurry. In still another of its aspects, the present invention relates to a monodisperse distribution of silica aggregates and moreover yields clusters or aggregates which have both a fractal morphology and low mechanical strength. As used throughout this specification,"monodisperse"is intended to mean aggregate particles that have a narrow size distribution characterised by the standard deviation about the mean value of their radius. As used throughout this specification, the term"fractal morphology" is intended to mean that each of the particle clusters is a rough or fragmented geometric shape as opposed to"spheroidal morphology"wherein it is meant that the particles have a substantially constant internal radius. In yet another of its aspects, the present invention relates to precipitated silica slurries which can advantageously be used in the production of rubber/silica masterbatches.

Precipitated silica is well known as a filler material useful in a variety of products. One of the applications in which precipitated silica has achieved particular interest recently is as a filler for polymers. It is particularly useful for the reinforcement of articles made from vulcanized elastomers (e. g., butadiene rubber, styrene-butadiene rubber, natural rubber, EPDM and the like), including tire treads.

One requirement for the achievement of optimum reinforcing in a cured rubber article is that the silica or other filler be homogeneously dispersed throughout the elastomer compound in the form of discreet small particles. Dispersed filler particle sizes in the same order as polymer chain molecular dimensions (i. e., from a few tens of nanometers to a hundred nanometers or so, nanometer = 10-9 meters) provide excellent reinforcement potential while larger particles are generally viewed as being less advantageous in this respect. Dispersed particles with diameters in excess of one micron (103 nanometers) are, as a rule, non-reinforcing, and rubber articles containing such large particles may even have less utility than those without any added filler.

Many methods exist for the characterization of dry silica powders. The most important silica properties for rubber reinforcement are the average particle size (ASTM C721 and D1366 ; ISO 787 Part XVII), the specific surface area (ASTM D1933 and D5604; ISO 5794 Part I), the mean projected area of the aggregates (ASTM D3849), and the oil absorption (ASTM D2414; ISO 787

Part V). In the case of wet (i. e., never-dried) silicas, the average particle size may be determined from the original surface area measurement method of Sears, (G. W. Sears Jr., Analytical Chemistry, Vol. 28, No. 12 (December, 1956) pages 1981-1983), or modifications thereof as described in Iler, pages 203-205 or by an improved test as set forth and detailed in United States patent 5,739,197, col.

7. The CTAB surface area is also an important parameter and may be used to further characterize non-dried silicas. The CTAB area is defined as the external surface area, as evaluated by absorption of cetyl trimethyl ammonium bromide with a pH of 9, following the method set forth by Jay, Jansen and Kraus in Rubber Chemistry and Technology, 44, pages 1287-1296 (1971). A full procedure for this test is set forth in United States patent 5,739,197, col. 5-7.

Methods also exist for assessing the dispersibility of silica powders without the necessity of mixing them with rubber (see for instance US Patent 5,403,570 or A. Blume,"Analytical Properties of Silica-a Key for Understanding Silica Reinforcement", presented at the Rubber Division Meeting, ACS, Chicago, April 13-16,1999). If mixing is used, then optical or electron microscopy may be used to assess the resulting degree of filler dispersion; a full procedure is described in United States patent 5,739,197, col. 1-5.

Commercially available dry silica powders are composed of aggregates of primary particles.

Larger structures termed"agglomerates"may also be present, usually as a result of the drying process used during the manufacture. In general, superior reinforcing properties in rubber compounds are obtained from the use of silicas which have both a high specific surface area (i. e. > 200 m2/gm) coupled with a small average aggregate particle size. However, such silica materials are difficult to disperse during rubber mixing ("dry mixing") and also yield compounds with high viscosities that are difficult to process further by operations such as calendaring, extrusion or molding which are needed to form the desired end-products. Extended mixing times and high levels of ancillary agents are normally required to attain a good dispersion with such silica materials, making their use uneconomical or impractical for most rubber applications. Fine particle silica materials moreover present difficult in-plant material logistics due to poor material flow characteristics and concomitant dust generation.

The ease of dispersion of a dry silica powder during rubber mixing is recognized as being dependent on the porosity of the aggregate structure. The latter is in turn governed to a large extent by the size and packing of the primary silica particles (see for example, Iler, Chapter 5). The degree of aggregate porosity is usually measured by a form of absorption test. Oil is commonly employed for this purpose (ASTM D2414). Silica particle aggregates with high degree of oil absorption have a high internal void volume; i. e., they contain numerous suitably-sized pores into which the oil

molecules can migrate. Pores of sizes approaching molecular dimensions (2-200 nanometers) have been classified as"mesopores"by Unger (Klaus Unger,"Structure of Porous Adsorbents", Angew.

Chem. Int. Ed., 11 (4) 267 (1972)); during the dry mixing process, the polymer component may also enter these pores. The shearing and elongational forces of dry mixing cause the rubber-entrained silica particles to fracture and disperse in the bulk polymer. A porous particle also exhibits a diminished fracture strength compared to a similar sized solid particle, and is thus more easily fractured by the forces inherent in the mixing process. However, practical limits are imposed on the particle porosity by the rigors of the dry silica production process itself, as will be discussed later.

Considerable efforts have thus been made to develop processes which yield silicas with improved dispersibility (see for example, United States patent 5,929,156 and United States patent 5,587,514) by controlling the various manufacturing variables.

As a second requirement to achieve maximum development of reinforcing properties in elastomers, filler particles not only need to be well dispersed but their surface also needs to be substantially chemically compatible with the polymers in which they are used as fillers. Special chemical agents are normally employed to reduce the surface polarity of the silica particles when these are used with non-polar hydrocarbon rubbers. In standard practice, these surface-reactive agents (usually bifunctional sulfur-containing silanes) are added during the mixing cycle, i. e. during Banbury mixing. Chemical reactions take place between these added agents and the silanol groups (Si-OH) on the silica surface during the mixing step and later between the agent and the polymer at the compound curing step. This allows the silica particle to be linked to the polymer backbone.

Special thermomechanical mixing regimens are required to complete and maintain separation of these two different reactions on a practical time scale, making the overall mixing process quite difficult to control. In this regard, recent efforts have been made to develop new techniques for dispersing silica aggregates (normally hydrophilic) into such elastomers (normally hydrophobic) which circumvent the problems associated with the dry mixing process. See, for example, any of the following references: published International patent application WO 98/52954 [Koski]; published International patent application WO 98/53004 [Koski]; and published International patent application WO 99/15583 [von Hellens]; hereinafter collectively referred to as"the Bayer patent applications".

As it is well known that primary particle size and surface area are directly related, silica materials composed of aggregates of very small primary particles have a relatively high surface area.

To ensure maximum compatibility of such silica materials with non-polar polymers, as well as to prevent absorption of curative chemicals on the surface, high ratios of expensive surface treating agents per unit silica weight are therefore necessary. Practical limits thus exist on the maximum surface area that can be economically used.

There are two general approaches for the production of precipitated silicas known in the prior art.

The first, more recent approach relies on the controlled hydrolysis ("hydrolytic polycondensation") of a hydrolysable silane (i. e., tetraethoxysilane, TEOS or ethyl silicate) hereafter called"the Stober processes" (W. Sober et al., J. Colloid Int. Sci., 26,62-69, (1968)). The second, more classical approach entails the reaction of an acid (e. g., H2SO4) with a water solution of alkali silicate (e. g., Na2O (SiO2) 25334) There are many patents which describe variations of the Stober process to produce silica materials of controlled particle size, narrow particle size distribution and degree of spheroidal morphology (see, for example any one of United States patent 4,775,520; United States patent 4,861,572 and United States patent 5,425,930). While some of the silaceous materials produced by these Sober process variants may be technically acceptable as reinforcing rubber fillers, they are necessarily expensive owing to the complexity of the processes and above all the high cost of the raw materials. These products have commercial utility only in special applications where their cost/performance ratio is acceptable, such as in chromatography packings, electronic applications and as catalyst supports. Further references will thus not be made to silaceous materials produced by such Stöber-type processes in the context of the present specification since these materials are generally cost-prohibitive as primary rubber fillers.

Process variants utilizing the second approach are manifold as evidenced by the large number of patents on precipitated silica materials. The classical precipitation processes are understood to consist of the following steps (Encyclopedia of Chemical Technology, Volume 21, p. 1024): * formation of colloidal particles through nucleation and growth of these particles to form the primary particle; * coagulating the primary particles into aggregates to form a precipitate; and

reinforcement of the aggregate particle In practice, the processes are usually conducted by adding an acid at a predetermined rate to the heated alkali silicate solution contained in a first vessel. The first vessel is agitated throughout the reaction phase and relatively high reaction temperatures (i. e., 60°C to 85°C) are usually employed. Large vessel volumes are typically used. Addition of acid may be interrupted at a certain pH (ie.-8-8. 5) and additional silicate may also be added then, or the precipitation may be taken to completion and additional silicate and acid may be added later in a post-treatment (United States patent 4,336,245). These interruptions or post-treatments serve to improve particle aggregate strengths, and are usually referred to as"building"steps. Single or multiple building steps are possible (United States patent 4,243,428 and United States patent 5,911,963). The amount of reinforcement thus conferred to the aggregate is termed the"build-up ratio" (Iler, p. 557). Build-up ratio is further defined as the ratio of the final weight of silica in the system to the weight of the aggregated silica. Build-up ratios of 4: 1 or less are usually required to attain a silica that can be dispersed during dry mixing (Iler, p. 557). A certain amount of"build-up"is required to prevent collapse of the particle structure during the ensuing drying step as a result of capillary forces and those of surface tension as water is removed (Iler, p. 534-536). However, this build-up step also decreases the pore size and strengthens the particle thereby diminishing its intrinsic utility as a filler.

One embodiment of a silica particle during"build-up"is illustrated in Figure 2b.

After a period of time to allow for the desired amount of build-up, acid is again added to complete the desired degree of neutralization. The contents of the first vessel are then transferred to a ripening tank (e. g., a second vessel equipped with an agitator) wherein particle growth is completed over the course of several hours or days (known within the art as"Ostwalt ripening').

During this stage, many of the smaller particles dissolve and are redeposited on the surfaces of, or at the contact points between larger particles-see Figure 2a. The particle size distribution thus narrows and the average particle size increases. This ripening has the further effect of increasing particle strength thereby improving filterability by reducing the number of very small particles. It is these smaller particles which generally pass through the filter and are lost, or if their volume fraction becomes excessive, may clog the filter substrate. Filtering is still the most common process used to separate the silica from the mother liquor.

The wet filter cake resulting from the filtering step may be further processed by pressing to remove entrained mother liquor, washing to remove soluble salts, drying (i. e., in a tunnel drier) and finally grinding (i. e., using a hammer mill, jet mixer, ball mill or similar apparatus). These steps

yield the finished product as a powder having a broad particle size distribution and an average agglomerate particle size of from about 1 um to about 100 um, or in some cases even larger.

Moreover, because the particles are shattered by the final grinding process, their shapes tend to be angulated rather than fractal or spheroidal, and the flow characteristics of the powders are thus poor.

As a known means of improving the powder flow characteristics, the washed filter cake may be reslurried and then finished by a spray drying process. This process modification yields a more free flowing powder composed of spheroidal particle agglomerates; however because the so- produced dried particles are generally of larger size (i. e. > 50 microns) than those produced by grinding, longer mixing times are generally required to attain the same degree of filler dispersion when such spray-dried powders are used in the rubber compound. However, the spray drying process does produce a silica with much reduced tendency to dust, which is a desirable feature.

Several variations of the above general acid precipitation process are known in the prior art: For instance, an acidic gas may be employed as the precipitating agent. The acidic gas may be added directly to the heated silicate solution or it may be generated in-situ, for instance by the combustion of hydrocarbon (i. e., natural gas, in a submersible bumer as described in United States patent 3,372,046) to provide CO, and heat. Thereafter, the formed silica particles are ripened and the slurry is then filtered to remove the particles in the form of a wet filter cake. The filter cake may be further washed free of soluble salts, dried and ground to provide an end product having an average agglomerate particle size of from about 0.5 Rm to about 100 pM. Silicas produced by the above modified processes share the same disadvantages as those produced from direct addition of mineral acid to soluble silicate.

In yet another modified approach to the above acid gas technique, Derleth et al (United States patent 5,232,883) describe a process whereby an electrostatically charged alkali silicate solution is sprayed into a chamber containing an acidic gas (i. e. HCl). This process is said to produce a silica comprised of microspheroidal particles of narrow particle size distribution of between 50 to 200 micrometers and with a specific surface in excess of 200 m2/g, a pore volume of between 1 and 3.5 cm3/g, a roundness factor lower than 1.40, and a variation coefficient of the particle size distribution lower than 60%. However, these silica particles are designed specifically for use as catalyst supports for alpha-olefin polymerization. The accepted requirement of crush resistance in this application would suggest that these particles would be too hard for application as reinforcing fillers since reduction of the 50-200 micrometer agglomerates to the nanometer range necessary for reinforcement would be a lengthy and likely impractical process. The large surface area of these particles (i. e. up to 700 m2/g) would necessitate the use of large quantities of expensive surface-

reactive agents to ensure polymer/filler compatibility. Finally, the production of such particles requires a special apparatus of complicated design and would not be suitable for the large volume production of fillers.

Spheroidal silica particles in the 0.5 to 20 micrometer diameter range can also be obtained by agglomeration of colloidal silica in a polymeric matrix of urea and formaldehyde (United States patent 4,010,242). The colloidal silica, in this case"Ludox HS", is made by a precipitation of an alkali silicate utilizing an acid-form ion exchange resin. This patent further directs that the organic matrix must be removed by burning; this process thus suffers from unnecessary complexity and is wasteful of organic materials.

Regardless which modification of the acid precipitation prior art approaches is used, (with the exception of United States patent 5,232,883), the silica at the wet filter cake stage typically has a wide particle size distribution (PSD) of aggregates. Since each silica aggregate is made up of a number of smaller primary (ultimate) particles of varying sizes, a wide PSD means that, for a given aggregate : there are a relatively large number of contact points (i. e., between adjacent agglomerated particles) resulting in a particle aggregate with relatively high mechanical strength-see Figure lb, for example. While this strength is detrimental to ease of processing, it is nevertheless necessary to prevent collapse of the particle and the formation of a solid sintered mass during the drying process.

Thus, the requirements of the silica for a useful manufacturing process and those which provide for ease of processing during rubber mixing are somewhat at odds.

Further, for dried silica materials, a relatively high shear energy is needed to de-agglomerate the particle, and the particle porosity is low. Thus, it is somewhat difficult to conduct surface chemistry on all of the agglomerated particles in the sample.

The latter property is particularly problematic when employing a conventional silica preparation in the hydrophobicizing treatment set out in the Bayer patent applications. Specifically, when dispersing silica particles in an elastomer, it is desirable to treat silica particles prior to use to facilitate dispersion thereof into the elastomer. While the treatment of silica particles as described in the Bayer patent applications represents a significant advance in the art, there is still room for improvement.

For example, the Bayer patent applications refer to treatment of conventional silica preparations to facilitate dispersion thereof in the elastomer via a masterbatch technique. While the treatment technique is adequate on a small scale, it would be beneficial to have an improved silica preparation, inter alia, which would facilitate treatment of the surface of the silica particles and ultimately facilitate dispersion of the treated silica particles in the elastomer.

It is an object of the present invention to obviate or mitigate at least one of the above- mentioned disadvantages of the prior art.

It is another object of the present invention to provide a novel process for producing a precipitated silica slurry.

It is yet another object of the present invention to provide a novel precipitated silica slurry.

Accordingly, in one of its aspects, the present invention provides a process for production of a precipitated silica slurry comprising the steps of : (i) continuously feeding to a mixing zone an aqueous metal silicate compound, an acid and an electrolyte to produce the reaction mixture; (ii) maintaining the reaction mixture at a temperature of less than about 100°C in a quiescent zone; (iii) continuously removing the reaction mixture from the quiescent zone.

The present inventor has unexpectedly discovered that silica particles useful as fillers may be grown without the need to utilize a"building"step as described hereinabove. It is surprising then that easily dispersible silicas may be produced by this invention. Preferably, the silica particles produced in the process have a volume average particle size of at least 5 m, more preferably in the range of from about 10 pm to about 40 J. m.

Embodiments of the present invention will be described with reference to the accompanying drawings, in which: Figures 1 and 2 illustrate schematic views of various forms of silica particulate material; Figure 3 illustrates a schematic of a reactor used in the example described herein below and which is the subject of a co-pending patent application; Figures 4 and 6 illustrate particle size distribution curves for a pair of silica slurries based on volume fraction; Figures 5 and 7 illustrate particle size distribution curves for a pair of silica slurries based on number fraction; Figure 8 illustrates the result of an ultrasonic dispersion test on the number average particle size for a pair of silica slurries ; Figure 9 illustrates an electron microscope image of a commercially available silica material; and Figure 10 illustrates an electron microscope image of a silica material made in accordance with the present process.

The precipitated silica slurry produced in accordance with the present process has a relatively narrow PSD and moreover the particle aggregates have a fractal morphology. Specifically, since each silica particle in the precipitated slurry produced in accordance with the present process is itself a fractal aggregate of a number of smaller particles, a narrow PSD means that, for a given silica particle: (i) there are a relatively small number of contact points (i. e., between adjacent particles within the aggregate), (ii) only a relatively low shear energy is needed to de-agglomerate the particle since only a few contact points need to be broken, and (iii) it is possible to conduct surface chemistry on substantially all of the surfaces of the aggregate since these are readily accessible (i. e., compared to a slurry having a relatively wide PSD of silica particles).

Thus, the precipitated silica slurry produced in accordance with the present process may be used advantageously in the hydrophobicizing treatment set out in the Bayer patent applications. In other words, in a preferred embodiment, the precipitated silica slurry produced in accordance with the present process is not filtered and dried; rather it is subject to further treatment whilst in the nascent slurry form. This allows for production of improved treated silica particles which leads to improved dispersion thereof in an elastomer via the masterbatch technique described in the Bayer patent applications.

Step (i) of the present process comprises continuously feeding to a mixing zone an aqueous metal silicate compound, an acid and an electrolyte to produce the reaction mixture.

Preferably, the aqueous metal silicate comprises an alkali metal silicate. More preferably, the alkali metal silicate has the formula M2O. (SiO2) x wherein M is an alkali metal and x is at least 2. Preferably, x is in the range of from about 2 to about 4 (including fractional numbers). Non-limiting examples of useful alkali metal silicates may be selected from the group comprising sodium silicate, potassium silicate, ammonium silicate and mixtures thereof. Alkaline earth silicates such as calcium silicate or magnesium silicate may also be used. For economical reasons, the most preferred alkali metal silicate is sodium silicate, particular a sodium silicate comprising a SiO, : Na2O ratio in the range of from about 0.5 to about 4.0, also referred to as'water glass.' The acid used in Step (i) is conventional. Non-limiting examples of such acids may be selected from the group comprising hydrochloric acid, phosphoric acid, sulfuric acid, nitric acid and mixtures thereof. Acidic gases such as CO2, HCI, SO2 and the like may also be used.

The electrolyte used in Step (i) is conventional. Preferably, the electrolyte comprises an alkali metal salt. More preferably it is a sodium salt. Non-limiting examples of useful sodium salts may be selected from the group comprising sodium chloride, sodium nitrate, sodium carbonate, sodium bicarbonate, sodium chlorate, sodium bromide and mixtures thereof.

The reaction mixture produced in Step (i) of the present process is retained in the mixing zone for a period of time sufficient to achieve a uniform distribution of reactants. Preferably, the residence time of the reaction mixture in the mixing zone is less than about 15 minutes, more preferably less than about 10 minutes, even more preferably less than about 8 minutes, even more preferably less than about 4 minutes, most preferably less than about 1 minute.

The ratio of acid to alkali silicate is such that the pH of the reaction mixture in the mixing zone is maintained in the range of from about 6.0 to about 10.0. More preferably, the pH of the reaction mixture in the mixing zone is maintained in the range of from about 7.0 to about 9.0. Most preferably, the pH of the reaction mixture in the mixing zone is maintained in the range of from about 8.0 to about 9.0. Appropriate pH control can be achieved by regulating the amount of acid used in Step (i) of the process whilst keeping the flow of alkali silicate constant.

Step (ii) of the present process comprises maintaining the reaction mixture in a quiescent zone at a temperature of less than about 100°C. Preferably, Step (ii) comprises maintaining the reaction mixture in the quiescent zone at a temperature in the range of less than about 90°C, and most preferably from about 30°C to about 50°C.

Step (iii) of the present process comprises continuously removing the reaction mixture from the quiescent zone. This allows continued introduction of silicate, acid and electrolyte in Step (i) thereby rendering the present process continuous.

Preferably, the apparatus for use in the present process comprises a substantially vertical reactor assembly equipped with at least one reactant inlet disposed below a reaction mixture outlet.

The preferred reactor assembly is equipped with two reactant inlets disposed substantially near the bottom of the reactor assembly and a single reaction mixture outlet disposed near the top of the reactor. Preferably, the reactor is equipped with an intensive mixing zone and an intermediate substantially quiescent zone between the mixing zone and the outlet.

The preferred reactor assembly further comprises a means to achieve turbulent mixing (e. g., an impeller or the like) disposed in the mixing zone at or near the reactant inlet (s) to achieve essentially complete axial mixing thereof, and a means to reduce any backmixing from the quiescent zone into the mixing zone.

In this preferred form of the process, Step (i) preferably comprises feeding to the mixing zone: a first feed comprising an aqueous solution of the metal silicate and the electrolyte; and a second feed comprising an aqueous solution of the acid and the electrolyte. Practically, it is preferred that the aqueous solution in at least one, preferably both of, the first feed and the second feed comprises a brine solution.

Thus, in this preferred embodiment, the first feed and the second feed are continuously pumped into the reactor via their respective reactant inlets to provide a reaction mixture in the mixing zone. In the mixing zone, the reaction mixture is subject to intense axial mixing. The occurrence of vertical mixing in the reaction zone is preferably minimized or most preferably avoided by means as described in co-pending Canadian patent application CA 2,292,862. With continued pumping of the first feed and the second feed, the reaction mixture exits the mixing zone and rises toward the top of the reactor, into a substantially quiescent zone in a plug-flow manner.

Once the reaction mixture reaches the outlet, it exits the reactor assembly. Preferably, the reaction mixture, now in the form of a precipitated silica slurry, may then be transferred to a settling tank wherein the pH of the slurry is further adjusted, as necessary. If desired, the silica particles may be allowed to settle to the bottom of the settling tank after which the mother liquor may be decanted.

The silica particles may be further filtered, washed and dried in a conventional manner.

Alternatively, the decanted aqueous silica slurry may be used directly in the above-mentioned hydrophobicizing reactions described in the Bayer patent applications-i. e., without the need to filter and dry the silica particles. The decantate, which consists of the electrolyte and by-product salts may be recycled.

Embodiments of the present invention will be illustrated with reference to the following Example, which should not be use to construe or limit the scope of the present invention.

EXAMPLE A silicate solution was prepared by dissolving sodium silicate [Na2O (SiO2) 3 25] (28.93 kg) in water (163.5 kg). The ingredients were stirred and heated to 40°C, as required, to produce an aqueous solution of sodium silicate. Thereafter, rock salt (5.13 kg) was added with stirring to the aqueous solution of sodium silicate until all solids had dissolved to produce a brine solution of sodium silicate.

An acid solution was prepared by slowly adding with stirring concentrated hydrochloric acid (17.2 kg; 36-38wt% HCl) to tap water (152.8 kg). Thereafter, rock salt (5.13 kg) was added with continued stirring until all solids had dissolved.

The silicate solution and the acid solution were independently fed by means of pumps to a reactor shown schematically in Figure 3. The feed rate for each solution was adjusted by means of flow controllers and metered so as to provide a feed rate of 1 to 2 litres per minute for each solution.

The solutions were independently fed to the bottom of the reactor assembly where they were intensively axially mixed in a mixing zone. With continued pumping of each solution to the mixing zone, the reaction mixture flowed into a quiescent zone and thence though an outlet into an overflow tank. The pH at the outlet was maintained at 7.3 by adjusting the flow of acid to the acid inlet at the bottom of the reactor assembly. The product exiting the reactor assembly via the overflow was stored in a collection tank. Total residence time in the reactor assembly was between 10 and 25 minutes.

A sample of the silica slurry produced in this Example was filtered, washed and then adjusted with water to give a 2.5 wt. % slurry. For comparative purposes, a 2.5 wt. % slurry of HiSil 233 (a commercially available particulate reinforcing silica material) was prepared by adding the dry material to water and shaking manually to disperse.

The particle size distribution (PSD) for each slurry was determined by means of laser light scattering using a Malvern Mastersizefrm instrument and the manufacturer's directions. The results are illustrated in Figure 5 in terms of the volume average PSD and in Figure 4 as the number average PSD. Figure 5 shows that the particle size distribution of the Example slurry is substantially narrower than that of the HiSil 233 slurry. Figure 4 shows that the number average particle size of the Example slurry (1.45 pm) is much smaller than that of the HiSil 233 slurry (4.17 pm).

The ease of dispersion of the Example silica and HiSil 233 slurries were determined by the following protocol: Exactly 80 milliters of the slurry was placed into a 100 ml. beaker and a small stirring bar was introduced. The beaker was then placed on a magnetic stir plate and the sounding horn of a Branson Sonifier Model W-350 was immersed into the slurry to a depth of 4 cm. The magnetic stirrer was activated at a medium setting and the slurry was stirred for exactly 2 minutes. The sonifier was then activated at a setting of 140 Watts/cm on the dial over a period of 28 minutes. Small samples of slurry were periodically withdrawn during sonification and the PSDs of each slurry were then measured as previously noted using the Malvern Mastersizefrm. The analysis results for the slurry after 22 minutes of sonification are shown in Figure 6 (number average PSD) and Figure 7 (volume average PSD). The results in Figure 6 show that the number average particle size distribution for the

Example slurry is substantially smaller (1.12 pm) than that of the comparative material (2.54 u. m) indicating that the Example silica will give a dispersion of finer particles and hence better reinforcement. The Example silica also develops a distinct bimodality whereas the only indication of such that is evident in the comparative silica is a small shoulder on the left side of the main peak.

In accordance with the interpretation offered by Blaume, (see"Analytical Properties of Silica-a Key for Understanding Silica Reinforcement", Presented at the Rubber Division Meeting, ACS, Chicago, April 13-16,1999) the ratio of the peak height of the original agglomerate peak to the peak height of the decomposed agglomerates may be taken as an indication of dispersibility in rubber.

This ratio, termed the WF coefficient, is 1. 1 for the Example silica and 3.4 for the comparative silica, indicating that the agglomerates in the Example silica are much easier to break down mechanically than those in the comparative silica and the Example silica should thus disperse more readily during rubber mixing.

The change of the number average particle size with time of sonification for the Example silica and the comparative silica are both shown in Figure 8. While the initial rate of change for the comparative silica is initially quite fast (as indicated by the slope of the line) at 0.13, um/min, it slows to 0.035 pm/min during the last minute or so of sonification and by graphical extrapolation it appears that it will eventually reach zero at around 40 minutes. On the other hand, while the slope of the curve for the Example silica is flatter at 0.021 pm/min, there is no indication that it changes during the entire 28 minute course of sonification; the Example silica should thus continue to disperse at the same rate as mixing is continued.

Figure 9 illustrates an electron microscope image of the HiSil 233 silica material and Figure 10 illustrates an electron microscope image of the Example silica material. A comparison of these two Figures clearly illustrates that the Example silica material is made up of a number of similarly sized aggregate particles (i. e., a narrow PSD) while the comparative silica is made up of particles having a larger variety of particle sizes (i. e., a broad PSD). The smaller average particle size for the Example silica is easily seen when compared to the HiSilTM 233. The fractal nature of the aggregates in the Example silica are distinct from the angulated particles of HiSil 233.