Carbon black treated with silica in aqueous systems for tire tread compounds.

Rolling resistance, wet traction and abrasion resistance of a tire are the most important properties among the many kinds of requirements for a tire. Properties of the tire tread compound significantly contribute to these three properties. The viscoelastic properties of the fire tread compound have been controlled in order to improve the rolling resistance and wet traction. Namely, efforts of enhancing the tan 8 near [Delta] [degrees] C and reducing it near 60 [degrees] C have been continued in order to enhance the wet traction and reduce the rolling resistance, respectively.

Thanks to the recent remarkable development in the field of the synthetic robber, which includes the development of the anionic living polymerization for example, we can use many SBR and BR types. Polymer blending is one of the important methods to design and control the temperature dependence of the viscoelastic properties. Blending polymers that have a rather high glass transition temperature (Tg) is commonly used to get the higher tan [Delta] near 0 [degree] C. However, using the higher Tg polymer allows the tan [Delta] near 60 [degrees] C of the rubber composition to increase at the same time.

The chain end modification techniques solved this dilemma. The number of chain ends of the polymer molecules determines the level of tan [Delta] near 60 [degrees] C. The chain-end modified polymer by introducing atoms or groups, which can interact with the surface of the carbon black, to the active chain end of the anionic living polymer represents the reduced tan [Delta] near 60 [degrees] C (refs. 1-3).

Furthermore, the modification of both ends of the polymer molecules has been recently investigated (ref. 4).

Carbon black (CB) has been widely used as a reinforcing filler in rubber compounds. Improvement of the carbon black has also been attempted for a long time. The carbon black industry has been concentrating on developing a method of controlling the shape of the aggregates for many years. At the same time, some improvement in the characteristics of the surface of the CB particles has been done in recent years (ref. 5). The surface structure of the CB has been gradually clarified by recent analytical approaches (refs. 6 and 7). The chemical analysis of the CB surface has also been made by Lamond et al (ref. 8). Rivlin (ref. 9) performed a quantitative analysis for the chemical groups on the surface. The coupling agent between the CB and the rubber molecules was developed using these chemical groups in order to improve the physical properties of the rubber compound (refs. 10-12). Wolff (ref. 13) investigated the surface of the reinforcing fillers including the CB and the silica. It became clear that there was a big difference between the surface energy of the CB and the silica. The silica has a much lower dispersive component of surface energy and a much higher polar component compared to the carbon black.

The tire tread compounding using the silica has been developed in order to improve the rolling resistance in recent years. The surface of the silica has both isolated and geminal silanol groups (ref. 14). The density of the silanol groups is about 3.5/[nm.sup.2] for a pyrogenic silica and about 8/[nm.sup.2] for a precipitated silica (refs. 15 and 16). Development of the silane-coupling agent for the diene-type rubber compound contributed to the practical use of the silica in fire tread compounds. TESPT(Bis-(3-triethoxysilylpropyl)-tetrasulfane) is the widest used coupling agent in tire tread compounds. The alcoxysilil groups of the agent react with the silanol groups of the silica surface. On the other hand, the tetrasulfane part reacts with the rubber molecules during the vulcanization process. Chemical bonding is formed between the silica particles and the rubber molecules (refs. 17 and 18). This bonding seems to produce a low hysteresis for the rubber composition.

The application of the silica in the tire tread compound is one of the big improvements in technology for reducing the automobile's fuel consumption that originates from the tire. However, the silica as the reinforcing filler for the tire tread compound still has several disadvantages. Controlling the morphology of the silica aggregates is one of the undeveloped areas. Therefore, the selection of a silica that has the appropriate properties for the required characteristics of the rubber compound is difficult at the present time. The high electrical resistance, the lower abrasion resistance and the lower dispersibility are also weak points of the silica. The lower electric conductance of the tire tread compound should be overcome as it has the potential to cause some electrical problems such as generating noise in the radio.

The silica treated carbon black (referred to as silica coated carbon black - SCC) was developed in order to reduce the disadvantages of the silica (ref. 19). The SCC is produced, for example, by the hydrolysis reaction of sodium silicate by adding sulfuric acid to a carbon black slurry (ref. 20). This method is after treatment, so that the SCC inherits the nature of the used carbon black. The same kind of modification for carbon black is proposed, which is obtained by introducing both the feedstock and the silica source into the carbon black reactor (ref. 21).

Several SCCs having different silica content are prepared by changing the amount of the sodium silicate used for the hydrolysis reaction. Their characterization and the performance in the rubber composition are also investigated in this study.

Experimental

Silica treatment

The silica treatment of the carbon black was made in an aqueous system. The CB (N339) slurry (the CB concentration was about 6.6 wt. %) was prepared using a homogenizer. The system was adjusted to pH 10 using an aqueous solution of sulfuric acid and an aqueous solution of sodium hydroxide, and the temperature was controlled at 90 [degrees] C. Sodium silicate was diluted in 1,000 ml of water. The concentration of the sodium silicate was determined by the target content of the Si[O.sub.2] in the SCC. The aqueous solution of sodium silicate was poured into the carbon black slurry at a speed of 20 ml/minute while maintaining the pH and the temperature. The system was then kept at pH 7 for two hours. The system was constantly stirred during these procedures. The SCC slurry obtained from the above procedures was filtered, rinsed and dried to remove the water.

Characterization of the silica treated carbon black

The colloidal characterization of the carbon black and the SCC was measured in conformity with the common method. Test methods of the ash content, the nitrogen surface area, the iodine adsorption and the DBP absorption were done following ASTM D1506, D3037, D1510 and D3037, respectively. The hydrogen fluoride (HF) treatment for the ash of each filler attempted to measure the actual silica content. The silica contents were calculated from the quantity of the substance that was removed from the ash content by the HF treatment.

The HF treatment was also used to eliminate the silica on the surface of the SCC. Four ml of HF were poured into the three grams of the SCC, and then some water was added. The mixture was allowed to stand for about 30 minutes, then was carefully rinsed with water and dried.

The separation of the carbon black and the silica was attempted in order to distinguish the SCC from the mixture of the CB and the silica. A 0.5 gram sample of the filler was put into a test tube, then an adequate quantity of water and toluene was added. At this stage, the water and the toluene formed two phases (the upper phase was toluene and the lower phase was water). The filler in the test tube was dispersed by an ultrasonic generator for 10 minutes. The filler appeared to be uniformly dispersed in both phases at this stage. The test tube was then left to stand for over 12 hours. At this stage, most of the filler was in the bottom of the upper phase (the toulene phase). However, only the silica was in the lower phase (the water phase) because of its strong hydrophilic nature. By measuring the ash content of the filler in the upper phase, the amount of silica that adhered to the carbon black aggregates can be estimated.

A transmission electron microscope was used to observe the aggregates of the SCC. Furthermore, an atomic analysis using an energy dispersion x-ray analyzer was done to confirm the existence of Si atoms on the surface of the SCC aggregates. The electron beam diameter was 30~40 nm for the measurement.

An estimate of the number of silanol groups was also performed. We used the reaction of the trimethylsilyl groups of hexamethyldisilazane (HMDS) with the functional groups, such as silanol groups, on the surface of the filler. The fillers were heated in n-hexane for 12 hours, then dried in a vacuum drier for eight hours at 60 [degrees] C.

Fifty mgms of HMDS were dissolved in 10 ml of n-hexane, and then poured into 500 mgms of the filler in a test tube. A 300 mg sample of HMDS was required only in the case of silica because of the greater number of silanol groups and larger surface area. The sample in the test tube was kept at 80 [degrees] C in a water bath for one hour in order to promote the reaction between HMDS and the filler while refluxing the n-hexane. After cooling, the sample was separated into the supernatant liquid and the filler by a centrifugal separator. The supernatant liquid was diluted to the appropriate concentration and the extinction coefficient was determined in order to know the unreacted HMDS. The quantity of the unreacted HMDS was determined from the intensity of the extinction at 204.5 nm by the calibration curve that was made using the standard HMDS solution. The absorption of HMDS onto the glassware of the instruments was negligible as a result of the careful checking.

Mechanical properties of the compound

The measurement of the mechanical properties was done using the formulation in table 1. The mixing procedure for the robber compounds is shown in table 2.

Table 1 - the formulation of the rubber compounds for the measurements of the mechanical properties

                            CB     Silica    SCC

SBR 1502                    100     100      100
Filler                     40-60   40-60    40-60
Silane coupling agent(1)     0       6        3
DEG(2)                       0       3        0
Zinc oxide                   3       3        3
Stearic acid                 2       2        2
Antioxidant(3)               2       2        2
Sulfur                       2       2        2
TBBS(4)                      1       0.5      1
DPG(5)                       0       1        0

(1.) Bis-(3-triethoxysilylpropyl)-tetrasulfane

(2.) Diethylene glycol

(3.) N-(1,3-dimethyl-butyl)-N'-phenyl-p-phenylenediamine

(4.) N-tert-butyl-2-benzothiazolyl-sulfenamide

(5.) Diphenylguanidine

Table 2 - the mixing procedure for preparing the rubber compositions in table 1

1st step                                        2nd step

Time       Ingredient              Quantity   Curing agents
                                              were mixed in
0'00"      Rubber                  all        this step by a
                                              roll mill.
0'30"      Filler                  1/2
           Silane coupling agent   all
           DEG                     all

2'00"      Filler                  1/2
           Others                  all

3'30"      Sweep run

Rotor: 60 rpm; temperature control: 60 [degrees] C

The viscoelastic properties of the compounds were measured using a strain-type viscoelastic tester. The long rectangular strip rubber specimen was subjected to a 10% static tensile strain.

The sinusoidal strain displacement (20 Hz) was then dynamically imposed on both ends with a maximum strain of 0.5%. The measurements were performed over the temperature range of -100 [degrees] C to 100 [degrees] C.

The measurement of electrical resistance of the compounds was done following ASTM D991 and JIS K6911. JIS K6911 was used for the higher electric resistant samples.

Results and discussion

The investigation of the characterization of the SCC was done to clarify the morphology and toughness of the adhesion of the silica on and the surface of the CB. The typical analytical data of the SCC are shown in table 3 with its base CB (N339). SCC-2, 5 and 10 were obtained following the method listed in the experimental section, and the target value of each SCC was 2, 5 and 10 wt. % of the hole filler weight, respectively. The actual values of the silica content in SCC-2, 5 and 10 were 1.76, 4.83 and 8.92 wt. %, respectively. We can conclude that the method for silica treatment, which was adopted in this study, gives the silica content that is almost equivalent to the target value.

Table 3 - the colloidal characterization of the SCC and their base CB

Samples   Base CB   Ash content  [N.sub.2]SA    IA      DBP
                     weight %    [m.sup.2]/g   mg/g   ml/100g

      CB N339          0.17           97        91      119
SCC-2     N339         1.85           85        80      123
SCC-5     N339         4.90           88        79      120
SCC-10    N339         9.05          104        74      127

The nitrogen-specific area ([N.sub.2]SA) value decreases for a low silica content, and increases with a higher one. However, the changes are not very large. The phased reduction of iodine absorption (IA) with increasing silica content must be noted. Figure 1 shows the plot of [N.sub.2]SA vs. IA of the CB, the silica, the CB/silica blends and SCC. A CB with [N.sub.2]SA = 97[m.sup.2]/g, a precipitated silica with [N.sub.2]SA = 195[m.sup.2]/g and their blends are used as the controls. All of the blends lie along the line between the CB and the silica. The SCCs are in a different area from the line of the CB/silica blends. Furthermore, the IA value of the SCC originally being 40, which included 22 wt. % of the silica, increased to 83 after the HF treatment. This fact suggests that the surface of the SCC is changed to a silica surface because the IA of the silica is almost zero. However, this speculation cannot be fully supported because the line consisting of the CB/silica blends can be changed by using silica having a different [N.sub.2]SA value.

[Figure 1 ILLUSTRATION OMITTED]

Figure 2 is a plot of the relationship between the IA reduction rate and the silica content of the same fillers in figure 1. The reduction rate gives the ratio of the replacement from the CB surface to the silica surface. The IA reduction rates of the SCC are greater than that of the CB/silica blends when the silica content is equal. As the IA value of the silica is zero, this greater reduction at the same silica content might be caused by the elimination of the CB surface in the system. The CB having the same [N.sub.2]SA value is used in both the SCC and the CB/silica blends. As a result, it is concluded the elimination of the CB surface must be the result of surface coverage by the silica.

[Figure 2 ILLUSTRATION OMITTED]

A further analysis for the silica included in the SCC was required in order to confirm the adhesion of the silica on the CB surface. Observations of the SCC's aggregates were made using the TEM and TEM-EDX. Figure 3 is a typical image of the contacting area of the silica and the CB of the SCC-5. The oriented lamella-like structure, which is widely known as a typical structure, is observed in the CB region. The amorphous structure, which can be distinguished from the CB, contacting the CB surface, is also observed. The amorphous region seems to be comprised of silica. Furthermore, the elementary analysis by TEM-EDX was made in order to detect the existence of Si atoms on the aggregate. The result of the analysis for the SCC is shown in figure 4. A fluorescent x-ray emission from the randomly selected points, A to I, on the SCC aggregate was detected. The fluorescent x-ray, Si-K[Alpha]s, was not detected from all points. It is assumed that the silica regions are randomly dispersed in the CB aggregate. The results of these observations also support the fact that the CB surface is covered by silica.

[Figures 3-4 ILLUSTRATION OMITTED]

When using SCC for reinforcing the rubber, the strength of the adhesion between the CB and the silica should be evaluated because the filler particles are generally dispersed into the rubber through a high shear environment during mixing. We attempted the separation of the amorphous silica from the carbon black surface of the SCC by the test method described previously. CB, the hydrophobic filler, is placed in the toluene phase and the silica, the hydrophilic filler, is placed in the water phase. Thus the blends of the silica/CB were separated into both phases by this method. The same test was also done for the various SCC silica contents. SCC-2, 5 and 10 were placed in the bottom of the toluene phase before the ultrasonic treatment. If the adhesion strength between the silica and the CB surface withstands the ultrasonic treatment, most of the SCC is found in the toluene phase and no filler is detected in the water phase.

Furthermore, this test was used for the samples after the pressing procedure. This was done following the test method for the 24M4DBP (ref. 22). According to the test method, 165 MPa, four times the compression is added to the filler samples in a metal cylinder. This simulates the shear stress that the filler aggregates receive during the mixing procedure. In this study we use this procedure to evaluate the adhesion strength of the silica on the CB surface. We only changed the number of compressions to eight times to evaluate the adhesion strength under conditions that are much more severe. The separated fillers in both the toluene and water phases were collected by drying the toluene or water. The ash content of the collected fillers was also measured and taken as the silica content. The separation of the silica and the CB was made for these compressed SCCs and the results are listed in table 4 along with the results for the uncompressed fillers. Almost all of the silica in the CB/silica blends was easily separated by this method and found in the water phase. Little filler was observed in the water phase for the SCC while most of the filler was in the bottom of the toluene phase. The appearance of the interface between the toluene phase and the water phase varied according to the silica content of the SCC.

Table 4 - results of the separation of the silica and the CB by the fractionation using toluence and water

                           Silica content in filler(4)(wt. %)

                                       Filler in    Filler in
             Mechanical               the toluene   the water
            treatment(3)   Original      phase        phase

CB(1)/         none          0.46        0.99%
Silica(2)      none          4.92        1.51%         30%
blend          none          8.04        1.33%         40%

SCC-2          none          1.85        2.31%
               24M8(5)                   1.65%

SCC-5          none          4.90        5.47%
               24M8(5)                   4.64%

SCC-10         none          9.05        7.52%
               24M8(5)

             Mechanical    Appearance of the
            treatment(3)   water phase

CB(1)/         none        No filler was observed
Silica(2)      none        A little filler was observed
blend          none                        "

SCC-2          none        No filler was observed
               24M8(5)                 "

SCC-5          none                    "
               24M8(5)                 "

SCC-10         none                    "
               24M8(5)

(1) - N339

(2) - precipitated silica

(3) - compressed 165 MPa x 8 times

(4) - ash content

(5) - "24M8" meant that the sample was given 165 MPa, 8 times of compression

This meant that the interface (the bottom of the toluene phase) was pushed out to the water phase depending on the hydrophilic nature of the SCC, which is at the interface. The silica content of the SCC in the toluene phase corresponded to the value of the original one within the experimental error. The silica content for the compressed SCC in the toluene phase also corresponded to the original one. As almost all of the filler was founded in the water phase after the ultrasonic treatment for only the compressed SCC-10, separating the silica could not be done. This may be caused by formation of polar functional groups brought about by the crushing of the CB aggregates in addition to its fairly high hydrophilic nature. Either way, the adhesion strength of the silica on the CB surface in SCC seems to be sufficiently high to withstand the shear stress during mixing with the rubber.

It is assumed that one of the characteristics of the SCC treated in the aqueous system is having the same active silica surface as the ordinary precipitated silica. According to the literature (ref. 23), the number of the silanol groups of the silica is varied depending on the method of the silica formation (fumed or precipitated) and the test methods. However, it is widely known that the precipitated silica has a greater number of silanol groups than the fumed silica. The number of silanol groups of the SCCs was measured using the silation agent HMDS. HMDS easily reacts with the silanol groups under mild conditions without significant side reactions (ref. 24). The reaction scheme is as follows:

2 [equivalent] SiOH + [Me.sub.3]Si-NH-Si[Me.sub.3] [right arrow] 2 [equivalent] 2 SiOSi[Me.sub.3] + [NH.sub.3]

However, there is a reaction with the functional groups on the CB surface, such as the hydroxyl group and carbonyl group when the sample is the SCC. Thus, the measurement for the base CB was also made in order to eliminate the effects of the reaction between HMDS and the surface functional groups on the CB. The results of the measurement are listed in table 5. The number of silica silanol groups, 12.6/[nm.sup.2], is higher than the previously reported 4.0 ~ 10.0/[nm.sup.2] value (ref. 16). This may be caused by the mild conditions of the sample pre-drying, namely dried under vacuum condition at 60 [degrees] C for eight hours. The number of functional groups of the base CB was 5.09/[nm.sup.2] and for SCC-10, it was 7.13/[nm.sup.2]. The rate of the surface area occupation by the silica is calculated from the IA reduction rate for the SCCs. The occupation rate for the SCC-10 is 1- ([IA.sub.STCB]/ [IA.sub.baseCB]) = 1 - (74/91) = 0.19. The number of functional groups on the SCC-10 can be estimated from the number of both CB and precipitated silica by assuming two kinds of surfaces on SCC-10, comprising amorphous silica and CB, having an equal surface to the precipitated silica and the base CB, respectively. The estimated number was 6.52/[nm.sup.2] and the agreement with the measured value of 7.13/[nm.sup.2] was relatively good. Based on these results, it is demonstrated that the amorphous silica region of the SCC is the same as the precipitated silica surface.

Table 5 - number of the functional groups react with HMDS

                       Number of functional
           Reacted    groups react with HMDS(*)
            HMDS

Sample     mol/g        /g      /[nm.sup.2]

CB N339   4.10.E-04   4.94E+20      5.09
SCC-10    6.16.E-04   7.41E+20      7.13
Silica    2.25.E-03   2.70E+21      12.6

(*) The functional groups for the CB N339 are hydroxyl and

It is also expected that the rubber compound containing the SCC has some unique properties. Figure 5 is a comparison of the temperature dependence curves of tan [Delta] for the silica, the CB and the SCC-5. The most significant difference of the SCC-5 is the decreasing tan [Delta] in the higher temperature region similar to the silica compound. The tan [Delta] in the high temperature region of the rubber compound corresponds to the rolling resistance of the tire when the compound is used as a cap or a side tread. The increase in the tan [Delta] in the lower temperature region was also observed for the SCC-5 compound and the silica compound. This region of tan [Delta] corresponds to the wet traction of the tire. Therefore, the SCC has both good wet traction and low rolling resistance compared with the CB.

[Figure 5 ILLUSTRATION OMITTED]

The electric resistances of the compounds including the various fillers were measured and the results were plotted with respect to the filler loading in figure 6. SCCs with silica contents of 2 wt. % and 22 wt. % were used here. The results for the silica and the CB are also plotted in figure 6. The volume resistivity of the CB compound was on the order of [10.sup.5] [Omega] [multiplied by] m at the loading level of 60 phr and [10.sup.7] [Omega] [multiplied by] m at 40 phr. The volume resistivity of the silica compound, on the other hand, was on the order of [10.sup.15] regardless of its loading. The volume resistivity of the SSC compound was dramatically changed according to the silica content in the filler and the loading level. When the silica content was 22 wt.%, the volume resistivity was altered from 2.25 x [10.sup.5] [Omega] [multiplied by] m (40phr) to 8.77 X [10.sup.6] [multiplied by] m (60 phr). When the silica content was 2 wt. %, their value was almost the same as that of the CB. It was demonstrated that the silica content of the SCC significantly affected the electric resistance of the compound and the value can decrease to the same level as the CB by reducing the silica content to about 2 wt. %.

[Figure 6 ILLUSTRATION OMITTED]

Conclusion

We devised the novel filler, SCC, that has advantages over the silica in order to overcome the defects when using silica in the tire compound. It was demonstrated that the SCC was made by the hydrolysis reaction of sodium silicate in a CB slurry. We could obtain the SCC having a silica content that was equivalent to the target value by this method.

Several kinds of analyses were performed in order to clarify the characterization of the SCC. The effects of the silica treatment were revealed especially in the iodine absorbed number among the general colloidal characteristics. The SCC apparently showed a different behavior from that of the CB/silica blends in the plot of [N.sub.2]SA vs. IA or the silica content vs. IA, namely showing that the silica covered the surface of the CB. The result of the TEM observation also supported the fact that the amorphous silica adhered to the CB surface. Furthermore, Si-K[Alpha] radiation from Si atoms, which seemed to originate from the amorphous silica, was detected. They were dispersed everywhere on the SCC aggregates based on the TEM-EDX measurements.

It was confirmed that the silica that adhered to the surface of the CB still remained on the CB surface after 165 MPa x 8 times compression procedure by the fractionation using toluene and water. This seems to mean that the adhesive strength between the silica and the CB surface withstands the shear stress during the mixing procedure.

The results of the measurement for the number of silanol groups representing the silica regions of the SCC have as many silanol groups as the precipitated silica.

The compound that included SCC had superior properties as a tire compound. The compound had a lower tan [Delta] in the higher temperature region and higher tan [Delta] in the lower temperature region similar to the silica compound. The electric resistance of the rubber compound also decreased by reducing the silica content in the SCC to an appropriate value.

References

(1.) F. Tsutsumi, M. Sakakibara and N. Oshima, Rubber Chem. Technol., 63,8 (1990).

(2.) K. Noguchi, A. Yoshioka, K. Komuro and A. Ueda, ACS Rubber Div. 129th Meeting, New York, 1986.

(3.) N. Nagata, T. Kobatake, H. Watanabe, A. Ueda and A. Yoshioka, Rubber Chem. Technol., 60, 837 (1987).

(4.) D.F. Lawson, M.L. Stayer, and T.A. Antkowiak, ACS Rubber Div. 153rd Meeting, Indianapolis, May (1998).

(5.) F. Forster, and B. Frend, IRC, Kobe, Japan, 27A-7 (1995).

(6.) J.B. Donnet, and T.K. Wang, Proceeding of IRC'95, Kobe, 451 (1995).

(7.) D. Goritz, P.G. Maier, Proceeding of IRC'95, Kobe, 455 (1995).

(8.) T.G. Lamond and C.R. Price, Rubber Journal, 152, 49 (1970).

(9.) D. Rivlin, Rubber Chem. Technol., 44, 307 (1971).

(10.) L.W. Howland, U.S. Patent 2,315,855 (1940).

(11.) A.R. Payne, P.M. Swift and MA. Wheelans, J. Rubb. Res. Inst. Malaysia, 22,275 (1969).

(12.) T. Yamaguchi, I. Kurimoto, K. Ohashi and T. Okita, Kautschuk Gummi. Kunst., 42,403 (1989).

(13.) S. Wolff, Rubber Chem. Technol., 69, 325 (1996).

(14.) R.K. Iler, "The chemistry of silica," John Wiley & Sons, New York, 1979.

(15.) J.H. de Boar, M.B.A. Hermans and J.M. Vleeskens, Proc. K. Ned. Akad. Wet. Ser. B50, 45, 54 (1957).

(16.) J.B. Peri and A.L. Hensley, Jr., J. Phys. Chem., 72, 2926 (1968).

(17.) S.N. Chakravarty, A.L. Kapur, S.N. Sau and M. Mithal, Rubber India, 33, 11 (1981).

(18.) P.K. Pal, S.N. Chakravarty and S.K. De, J. Appl. Polym. Sci. 28, 659 (1983).

(19.) T. Kawazura, H. Kaido, K. Ikai, F. Yatsuyanagi and M. Kawazoe, U.S. Patent 5,679,728 (1997)

(20.) T. Anzai, H. Murakami, H. Ito and K. Maeda, Japanese Unexamined Patent Publication (Kokai) No. 63-63755.

(21.) K. Mahmud, M.J. Wang and R.A. Fransis, U.S. Patent 5,830,930 (1998).

(22.) ASTM D 3493.

(23.) R.K.Iler, "The chemistry of silica," Chapter 6, John Wiley & Sons, New York, 1979.

(24.) J. Chmielowiec and B.A. Morrow, J. Colloid and Interface Science. 94, 319 (1983).3