Elastomeric barrier coatings for sporting goods. (Tech service).

Elastomers are generally poor diffusion barriers because the constituent polymer chains have a high degree of flexibility above their glass transition temperature. In fact, the permeability of a polymer correlates well with its Tg (figure 1), higher Tg elastomers generally having lower permeability. The requirement of certain large scale applications, such as tires, to maintain flexibility at vein low temperatures, can restrict the choice of materials.

[FIGURE 1 OMITTED]

The permeability of elastomers to gases is of great practical importance in a wide variety of applications, including sports balls, tires, innertubes and diaphragms. Other elastomer applications, such as chemical protective gloves that need elastomeric flexibility for dexterity and tactile sensitivity, also require the material to act as a barrier to molecular diffusion.

Often, the material of choice when both elastomeric flexibility (particularly at low temperatures) and diffusion barrier are required has been butyl rubber. It is used as the innerliner in tires and in chemical protective gloves. On the other hand, it absorbs a lot of energy when strained (i.e., has a high loss tangent), and cannot be used in applications such as tennis balls. It is sometimes used in other balls (such as soccer balls) for better air retention, but usually results in poorer mechanical performance.

Two potential approaches to enhancing the barrier properties of elastomeric articles involve the use of plate-like fillers either in the base elastomer or in a high performance elastomeric coating.

InMat has developed elastomeric nanocomposite coatings with a unique combination of barrier properties and flexibility that can resolve the issues described above. These coatings are produced from a completely aqueous dispersion. The dispersion formulation uses exfoliated vermiculite with very large aspect ratio vermiculite plates mixed with butyl rubber. The barrier properties have been achieved by maintaining the plates in a highly dispersed state whereby they maintain their large aspect ratio and form a tortuous path for any diffusing molecule. Equally important to the barrier properties, however, is the control over the surface interactions of the vermiculite plates which has led to highly filled nanocomposite coatings that can be stretched up to 20% without any damage. In this article, we will refer to these nanocomposites of butyl rubber and vermiculite as elastomeric nanocomposite barrier coatings (ENBC). This technology is currently used commercially in a premium tennis ball produced by Wilson Sporting Goods (ref. 1).

Background

Polymer nanocomposites formed with exfoliated layered silicates have received increasing attention. One simple approach to forming such a composite is to dissolve the polymer in the same media that can also exfoliate the clay. When a coating is made from such a mixture, there is at least the potential of forming a nanocomposite which has large aspect ratio plates dispersed in the polymer phase. Two successful examples of this are the use of water-soluble polymers with clay exfoliated in water (ref. 2) and the mixing of organically functionalized vermiculite dispersed in an organic solution of polycarbonate (ref. 3).

Another successful approach has been to mix an organically functionalized layered silicate with a monomer, and form a nanocomposite during the polymerization. Toyota has demonstrated that significant improvements in the mechanical properties of nylon can be achieved in this manner (ref. 4). This approach, however, is not easily extended to other polymers because of the difficulty in finding suitable monomers (ref. 5) and the need for changes in the polymerization process.

A more attractive approach that has gained significant attention is mixing the organically functionalized clay with a polymer melt. Giannelis (refs. 5-7) has developed an understanding of the thermodynamics of these systems, and some have recently been commercialized (ref. 8).

The use of clay polymer nanocomposite to develop improved barrier materials is well known, the basic idea being that the filler particles force detours in the diffusion path of molecules. The most important factors in determining the degree to which the diffusion is reduced are the aspect ratio of the plate, i.e., the ratio of the lateral dimension of the plate to its thickness and the degree the plates are aligned normal to the diffusion path. The larger the aspect ratio, the longer the distance a diffusing molecule must travel to permeate a polymer film or sheet. Cussler et al (ref. 9) have modeled this process for an idealized structure. The structure and the most important terms in his analysis are shown in figure 2. That analysis shows that the barrier properties will improve with the square of the aspect ratio, as well as with the square of the volume fraction. Thus, it is clear that, for the best diffusion barrier, one should use as much clay with as large an aspect ratio as possible. The limit to this is related to processing and secondary properties such as flexibility.

[FIGURE 2 OMITTED]

One of the major problems with polymer/layered-silicate nanocomposites has been that they are often brittle. This is because of the large aspect ratio, high surface area and high modulus filler that interacts strongly with the polymer (ref. 10).

Most of the work described in the literature has a prime objective to form a nanocomposite that can be melt processed. This means that the nanocomposite will be extruded or injection molded. These processes will often degrade the aspect ratio of the filler because of the high shear conditions typical of melt processing, and thus limit their effectiveness (ref. 3). A coating approach to nanocomposites can avoid this difficulty. Furthermore, the mixing process in thermoplastics does not typically lead to good orientation.

Experimental

Two types of gas permeation measurements were performed. Oxygen transmission was measured using permeation analyzers. These units enable quasi-simultaneous testing of several samples having transmission rates within the range of the instruments. The range of each instrument can be extended to some extent by masking samples to limit the area actually measured. These units can test samples from 0-40[degrees]C, and from 0-95% relative humidity.

The oxygen transmission rate (OTR) measurements are made on both coated mad uncoated substrates. The thickness of the coating on the substrate is determined by the weight change. In some cases, that measurement was confirmed by the use of an optical profilometer. The following equation is then used to determine the oxygen permeability, P, of the coating:

Coating thickness/P = 1/OTR of coated substrate - 1/OTR uncoated substrate

Additional gas permeability (helium or air) measurements were made using an internally developed apparatus utilizing a sample holder that allows the sample to be put under a controlled amount of uniaxial strain. Prior to a measurement, a sample is thoroughly degassed from both sides with a vacuum pump. The system determines gas transmission through sample by exposing one side of the sample to a step change in gas pressure and measuring the rise in pressure in a known volume on the other side of the sample with a capacitance manometer. The pressure vs. time data are acquired and a numerical derivative calculated which is used to determine both the steady-state permeability and the initial kinetics of the gas transmission.

A useful procedure in characterizing the flexibility of an elastomeric coating is to observe a sample with an optical microscope while subjecting it to increasing biaxial strain and noting when the first visible damage occurs. An apparatus was constructed for maintaining a sample in a controlled humidity environment while it is strained and observed with microscope. The distance between two small spots painted on the sample is measured and used to determine the strain The images are captured on video tape and transferred to computer for further analysis. Functional damage is also deretrained by measuring changes in permeability "after submitting samples to strain or prolonged fatigue testing.

Samples

All the formulations in this study are suspensions of butyl rubber and exfoliated vermiculite. Those that were not specially formulated to provide good flexibility trader all environmental conditions were used for the data (refs. 11 and 12) in figure 6 and in the formulation designated B2100. Examples of ENBCs formulated to withstand high strain even at 0% RH are AD2000 and AD5000, as shown in table 1 and figures 7b and 8.

[FIGURE 6 OMITTED]

Technical approach

We have developed a nanocomposite elastomeric barrier coating that is based on completely aqueous dispersions. Our formulations are based on a butyl rubber latex. This initial choice was based on the relatively good gas diffusion barrier properties of butyl rubber as compared with other common elastomers.

Our chosen filler is vermiculite, primarily because of its very high aspect ratio. Vermiculite (figure 3) is one of many layered silicates (or phylosilicate) typically used to form nancomposites. It is a 2:1 layer mineral with a combination of one octahedral sheet and two tetrahedral sheets comprising a layer. When the vermiculite is completely exfoliated, the layers come apart as a single unit, leading to a sheet thickness of about 1 nm.

[FIGURE 3 OMITTED]

The exfoliated vermiculite we use has sheets between 1 and 3 nm thick. The individual sheets are 10-30 microns in lateral extent. The high aspect ratio (of order 10,000) is critical to our ability to obtain good barrier properties.

A unique property of very high aspect ratio plates is a tendency to form self-aligned structures even at quite low concentrations. As the aspect ratio increases, the importance of the edge to face interaction is reduced. In addition, steric considerations dictate that the plates locally align (see figure 4), otherwise these concentrations could not be achieved. Even at concentrations as low as 0.02% (by weight), large (cm) regions of high orientation can easily be induced by small shear forces associated with mild mixing. Figure 4 shows a small glass jar containing a dilute vermiculite dispersion as photographed between crossed polarizers. The light and dark regions are indicative of the birefringence associated with a locally oriented `domain' structure. This strong tendency to align the platelets is a key feature of ENBC technology. When thin coatings are formed from an aqueous suspension, the wet film is typically 300-400 microns thick. The tendency of the plates to form large oriented domains means that there will be excellent alignment of the plates in the direction required for excellent barrier properties.

[FIGURE 4 OMITTED]

When the vermiculite is kept well dispersed, and the coating is formed with little loss in water, the resulting film shows large improvements in battier properties over that of the untilled polymer.

Exfoliated vermiculite has a very large platelet size compared to most other phylosilicates. Other materials commonly used in clay polymer nanocomposites such as montmorrilonites and other bentonites have plate sizes in the range of 0.1-0.5 microns. Although the large plate size of vermiculite provides advantages for the aqueous coatings, it is important to understand why they are not typically used in thermoplastic nanocomposites. Ward et al (ref. 3) demonstrated that when these large plates are processed in an extruder, they typically break up and thus provide little property improvement in the final product.

Results and discussion

The Cussler model for barrier enhancement by plate-like fillers predicts a rather strong dependence of the permeability on the filler concentration. Several formulations containing varying filler concentrations were formulated and applied to rubber substrates. The resulting oxygen permeabilities were determined using the analyzers. Their relative values are shown in figure 5 together with data taken from the literature (refs. 13-16) on several attempts at barrier enhancement using melt processing technology. Interestingly, the data appear to fit quite well to a Cussler model, but one with a substantially smaller aspect ratio than is intrinsic to the vermiculite plates. The reason for this behavior is not clear. Possibilities include plate agglomeration, imperfect plate orientation or plate flexibility.

[FIGURE 5 OMITTED]

Figure 6 shows micrographs taken while biaxially straining coated rubber substrates under desiccating conditions The left side of the figure shows the appearance of B2100, a comparatively brittle formulation which shows obvious damage even at very small strain levels. The right side of the figure is an improved formulation AD2000, which shows no visible damage through 20% strain.

The effect of uniaxial strain on the gas transmission of ENBC AD2000 is shown in figure 7. The relative the transmission increases with strain even for an `ideal' coating because of thinning. The coated rubber substrate shows a slightly increased transmission above the ideal, but nothing that could be considered catastrophic.

[FIGURE 7 OMITTED]

These results clearly show that ENBCs (and especially AD2000) have excellent flexibility despite having very large amounts of well dispersed vermiculite. The only measure we have of the dispersion of the filler is the large reductions we see in the permeability relative to the unfilled polymer. The permeability of the unfilled polymer is ~90 cc-mm/[m.sup.2]-day-atm. Thus, even AD2000 is 30 times less permeable than the unfilled polymer due to the dispersed vermiculite filler.

Comparing samples with the same filler content, it is clear that formulating the coating for flexibility increases the coating permeability. There are many possible reasons for this. At this time, however, we do not know which are correct. Some of the possibilities include: increased permeability of boundary layers, changes in orientation of vermiculite and changes in the vermiculite dispersion in the final coating.

Conclusions and applications

Aqueous coating formulations containing high levels of exfoliated vermiculite that can be used to make ENBCs have been developed, and the properties of the resulting barrier coatings described. These coatings can stretch up to 20% without damage, enabling their use in many products made out of rubber that can benefit from improved barrier performance.

One of these coatings is presently being used commercially to maintain pressure in tennis balls (ref. 1). More advanced formulations have been demonstrated (in laboratory quantities) that show the potential of providing another factor of ten improvement in barrier performance. With this level of barrier, new rubber products could be developed. These might include:

* Tennis balls that keep their pressure 9-12 months without a pressurized package.

* Soccer bails made with natural rubber bladders (latex) that don't need re-inflation for the lifetime of the ball (two years).

* Basketballs that last several years without needing re-inflation.

* Tires that need re-inflation once a year instead of once every three months. This would reduce the number of vehicles that have under-inflated tires, leading to reductions in fuel usage and tire related accidents.

These ENBCs are based on completely aqueous formulations and contain no hazardous materials or volatile organic compounds. The coating thickness is typically 10-30 microns, and thus the weight increase in a product is insignificant, as is any change in mechanical performance. ENBCs are an exciting new class of coatings that will lead to numerous improved rubber products.

References

(1.) Wilson Sporting Goods DoubleCore tennis ball, www. wilsonsports.com\tennis.

(2.) Sakaya, T.T., Kuroda, R.I. and Ogawa, T.T., "Gas barrier resin composition and process for producing the same," EP patent # 0 761 739 A1; December 3, 1997.

(3.) Ward, W.J, Gaines, G.L., Alger, M.M. and Stanley, T.J; "Gas barrier improvement using vermiculite and mica in polymer films," Journal of Membrane Science 55, pp. 173-180, 1991.

(4.) Usuki, A., Fukushima, Y., Fujimoto, M., Fukumori, K., Kojima, Y., Sato, N., Kurauchi, T. and Kamigaito, O., "'Composite material containing a layered silicate," U.S. patent 4,889,885; December 26, 1989.

(5.) Burnside, S.L. and Giannelis, E.P., "Synthesis and properties of new poly9(dimethylsiloxone) nanocomposites,'" Chemistry of Materials, v. 7, #9; September 1995.

(6.) Giannelis, E.P, "Polymer layered silicate nanocomposires," Adv. Materials 8; pp. 29-35; 1996.

(7.) Vula, R.A. and Giannelis, E.P., "Polymer melt intercalation in organically modified layered silicates: Model predictions and experiment," Macromolecules 30: pp. 8,000-8,009; 1997

(8.) Plastics Additives and Compounding, Jan. 2002, p. 30.

(9.) Cussler, E.L., Hughes, S.E., Ward, W.J. and Rutherford, A., "Barrier membranes," Journal of Membrane Science, 38; pp. 161-174: 1988.

(10.) Shia, D., Hui, C.Y., Burnside, S.D. and Giannelis, E.P., "An interface model for the prediction of the Young's modulus of layered silicate--elastomer nanocomposites," Polymer Composites, v. 19, #5; October 1998.

(11.) Feeney, C.A., et al, "Barrier coating of an elastomer and a dispersed layered filled in a liquid carrier." U.S. patent 6,087,016; July 11, 2000. PCT Filed 6/98.

(12.) Feeney, C.A., et al., "Barrier coating of an elastomer and a dispersed layered filled in a liquid carrier and coated articles," U.S. patent 6,232,389; May 15, 2001.

(13.) Elspass, Chester W. et al; "Tactoidal elastomer nanocomposites," U.S. patent 5,807,629; September 15, 1998.

(14.) Bissot, T.C., "Ethylene vinyl alcohol copolymer containing platelet-type mica fillers, processes for preparing same and multilayer containers with layers thereof" U.S. patent 4,818,782; April 4, 1989.

(15.) Blatz, P.S., "Barrier blends based on amorphous polyamide and ethylene/vinyl alcohol unaffected by humidity," U.S. patent 5,110,855; May 5, 1992.

(16.) Xanthos, M. et. al., Intern. Polym. Proc. XIII, pp. 58;1998.

Harris A. Goldberg (president) and Carrie A. Feeney (vice president) co-founded InMat LLC in October 1999. Douglas P. Karim is a senior scientist at InMat, and Michele Farrell is a formulation and process chemist.