Development, characterization of a water based semi-permanent mold release agent.

Development, characterization of a water based semi-permanent mold release agent

Mold release agents

In any rubber molding process, the easy and complete release of the finished article is essential. To achieve this, a mold release agent or abhesive is employed. Abhesives are designed to prevent or greatly decrease the adhesion of one surface to another surface which is in intimate contact with it, i.e. they reduce the work of separating the finished article from the mold.

Current world sales of release agents are approximately $200-$280 million. Up to 30% of this can be regarded as being used for purposes other than for decreasing polymer-to-mold adhesion. Examples of these products are talc and oil/water emulsions used for the prevention of green rubber to rubber adhesion during storage and mold and die lubricants. The remaining $140-$200 million can be separated into the sales of three distinct types of release agents.

* Sacrificial mold release agents - This type of release agent is applied to the mold surface before each molding cycle and achieves release by its cohesive failure when the molded article is removed, i.e. it acts sacrificially. One of the main criticisms of sacrificial mold release agents is that upon polymer release, a significant quantity of the release agent is transferred to its surface. Thus, if subsequent adhesion processes are involved (e.g. rubber/rubber or rubber/metal, then bond failure such as knit line failure will occur.

Commonly used sacrificial release agents are silicones (usually non-functional polydimethylsiloxanes (PDMS) varying in molecular weight between 6,000 - 80,000) (ref. 1), general oils, greases and waxes, e.g. animal and vegetable fats, talc, mica, soaps and polyvinyl alcohol (PVA).

* Internal mold release agents (IMRA) - Internal mold release agents are generally added to or are present in the unreacted polymer prior to their cure. They work by migrating to the reacting polymer surface during the initial stages of its cure, being pushed towards the surface by their surface active nature. When present at the surface of the polymer, they can then act as conventional sacrificial mold release agents. Unfortunately, as with SMRAs, IMRAs consequently interfere with rubber/rubber and rubber/metal bonding and often provide ineffective release when used on their own. A further disadvantage is that because the IMRA is added to the polymer reactants, it can adversely affect the fully cured polymers' mechanical properties and structures.

Commonly used internal mold release agents are zinc stearate and functional PDMS.

* Semi-permanent mold release agents (SPMRA) - Semi-permanent mold release agents are relatively new to release agent research and development and, at present, are responsible for a small but growing percentage of the total world sales of release agents. Of all SPMRA sales, approximately 25% is to the rubber industry.

SPMRAs are designed to enable more than one release to be obtained per release agent application and provide minimal transfer of the coating to the polymer. These objectives can be achieved by increasing the cohesive strength of the release agent coating. SPMRAs are typically crosslinkable polymers formulated or dissolved in various inert solvents. They are generally applied to the mold surface from solution and by a process of solvent evaporation and solute cure, form a complete, uniform thin film over the entire mold surface. Although the production of a crosslinked film tends to slightly increase the force required for release (compared to a silicone SMRA), it also increases the durability of the film and thus enables many releases to be performed from one application. The increased durability also results in a decrease in the amount of transfer of the release agent to the polymer, often to such an extent that secondary processing e.g. painting or bonding etc., can be performed without a surface preparation stage. This property is of vital importance when in-mold bonding is required, e.g. rubber to metal, and where the removal of possible contaminating release agent is impossible.

An ideal SPMRA overcomes almost all of the disadvantages associated with IMRAs and SMRAs. In formulating such a release agent, numerous factors must be taken into consideration. For example, an ideal SPMRA must:

* be easy to apply to large, small and complex molds, i.e. a one part system of low viscosity;

* wet the mold surface to give a complete and uniform film;

* be non-corrosive to the mold;

* produce a film thin enough to preserve all mold detail;

* solidify (cure) quickly to reduce production (down) time;

* adhere to the mold surface to prevent peeling, flaking or mass transfer;

* withstand processing conditions, e.g. temperature, pressure, shear and other abrasive forces;

* be abrasive resistant (durable) to give multiple release with as little transfer as possible;

* be chemically inert and give easy release to a wide range of polymeric materials;

* have a low coefficient of friction to facilitate easy release from deep or complex molds;

* fulfill a variety of safety specifications, e.g. high flash point and low toxicity;

* have a long shelf life (> 6 months); and

* be cost effective when compared to other types of release agents.

Environmental considerations

Clearly safety and environmental considerations are of utmost importance when formulating a release agent. The two major safety concerns are flammability and toxicity, while the two major environmental considerations are chlorofluorocarbon (CFC) and volatile organic compound (VOC) content.

Unfortunately for solvent-based release agents, fulfilling the safety requirements usually produces a product of either high CFC or high VOC content. For example, trichlorotrifluoroethane (Freon 113) has been used for many years as a non-flammable release agent diluent of low toxicity. It also possesses other properties that make it ideal as a diluent; it has a low boiling point (fast dry time), good solvency, low surface tension (good wetting) and is relatively inexpensive. While trichlorotrifluoroethane is exempt under California Air Resources Board (CARB) regulations for smog forming compounds and is classed as |negligibly reactive' under the Environmental Protection Agency's (EPA) definition of a VOC, it actually destroys ozone in the stratosphere with a depleting potential of 0.8 of trichlorofluoromethane (Freon 11; standard). As a result, trichlorotrifluoroethane is currently taxed in the USA at $1.10/lb. This tax is to be increased at a rate roughly corresponding to the decrease in its rate of production as agreed under the Montreal Protocol of September, 1987.

Virtually all chlorine containing compounds containing less than ten carbon atoms have some ozone depleting potential and thus ultimately face a global ban. This includes compounds such as 1,1,1-trichloroethane (methyl chloroform; which has previously been considered as |safe') and the newly developed HCFCs. Although trichlorotrifluoroethane can be replaced in virtually all release agent formulations with non-chlorine containing solvents, numerous product properties are affected. Resulting formulations show poorer wetting, increased dry times, increased flammability (lower flash points) and increased toxicity. In addition, the alternative solvents, while not destroying ozone in the upper atmosphere, actually cause ozone formation in the lower atmosphere (ozone being a component of air pollution). CARB regulations restrict the use of VOCs; the limit varying with type and classification of product. Typical release agent VOCs vary from 600 to 850 g/L. Under EPA direction, modified versions of the CARB VOC regulations are gradually being enforced in numerous other states, for example, New York and New Jersey.

Compounds facing a global ban under CFC regulations are virtually the only ones exempt from current VOC regulations. Clearly, this is a |Catch 22' situation with the only obvious solution being the development of a water-based product.

This article details the development and characterization of one such water-based product. Its properties (in terms of cure time, release efficiency, slip and degree of transfer) are compared with three commercially available solvent based SPMRAs, each of which fulfill the criteria for a SPMRA listed earlier.

Release agent formulation

Frekote Aqualine R100 (Product A) is a micro-emulsion formed by the high pressure homogenization of a multifunctional polyorganosiloxane with water in the presence of a surfactant blend. The mean particle diameter formed was determined (via a Nicomp Model 270 submicron particle size analyzer, Gaussian analysis) to be between 80-100 nm. The final formulation contains siloxane, ethanol, surfactant and water, resulting in a VOC of 22.4 g/L. The micro-emulsion is non-flammable and is stable to multiple freeze/thaw cycling. Upon application to a warm/hot substrate, the water/ethanol evaporates to leave a film which further condenses under the action of heat to leave an inert, insoluble, crosslinked coating chemically bonded to the mold surface via Si-O-M linkages (similar to a silane coupling agent). The micro-emulsion can be spray applied from 60 [degrees] to 180 [degrees] C using any spray equipment capable of fine atomization.

In order to judge the performance of this water-based product, it was compared to three solvent based SPMRAs (Frekote 700, 800 and HMT; hereafter referred to as Product B, C and D, respectively). Each product is commercially available and meets the criteria for a SPMRA listed earlier. SPMRA formulations B and C contain 25% and 12% w/w trichlorotrifluoroethane, respectively, to help wetting and increase flash point. Formulation D is a non-CFC product. The VOC content of the formulations are 706, 820 and 782 g/L, respectively. Each product contains a moisture curing resin in a solvent blend. Upon evaporation of the solvents and reaction with atmospheric moisture, a cross-linked film remains that is chemically bonded to the mold surface.

Products B, C and D have been formulated specifically for rubber release and are thus designed to be spray applied onto hot (150-180 [degrees] C) molds. Products C and D have been determined experimentally not to interfere with in-mold bonding processes such as rubber/rubber and rubber to metal bonding.

As these systems react with atmospheric moisture, airless spray application is recommended.

Characterization techniques

There are four main properties that are determined when characterizing a SPMRA, namely: cure rate, ease of release, slip and degree of transfer.

Cure rate

The accurate determination of the cure time of a SPMRA is essential for obtaining the minimum mold preparation time (down time), ease of release, lowest degree of transfer and thus overall maximum process efficiency. The use of an uncured film will invariably result in transfer onto the molded article and leave a layer of release agent on the polymer surface. The presence of transfer will decrease the number of releases possible from the coating due to the removal of significant quantities of release agent, principally from the first release. Also, the application of a thermosetting resin or rubber to a coating possessing many potential reaction sites may lead to a chemical reaction between the polymer and the film.

Consequently, this will eliminate any abhesive properties of the release agent.

Clearly, knowledge of the cure time of the release agent is of vital importance for its characterization.

A method has been developed (ref. 2) using a Dynamic Mechanical Thermal Analyzer (DMTA; Polymer Laboratories Ltd., Loughborough, England) which has been found to be accurate and reproducible. The DMTA is an instrument capable of determining the dynamic storage modulus, E [prime] or G', and damping factor (tan [Delta]) of rubbers, polymeric materials, composites, etc. The instrument is of the forced vibration type and its ability to change strain, frequency, temperature and atmosphere, enables it to determine the viscoelastic properties of a sample under a wide range of conditions. The test utilizes a standard woven glass fiber cloth as an inert substrate for the release agent. Controlling the atmosphere surrounding the mechanical head enables the cure rate to be determined at different temperatures and humidities.

Experimental - A glass cloth was clamped to the DMTA in the double cantilever fixture. The temperature cabinet surrounding the mechanical head was brought up to 150 [degrees] C (a typical rubber molding temperature). Release agent solution was applied to the glass cloth and the change in modulus of the resulting composite monitored against time. The cure rate was determined as a percentage of minimum and maximum composite moduli.

Results and discussion - The results (figure 1) show that the water-based release agent (A) has a cure time between 1.3 and 4 times longer than the solvent-based products after a 90% cure. At this point in the cure, each product will exhibit approximately 90% of its final properties, i.e. in terms of durability, transfer, ease of release, etc. Although A takes longer to cure than the solvent-based products, its 100% cure time is still < 5 min. at 150 [degrees] C. Whether this is significant will depend largely on the cycle time of the polymer and the durability of the coating, i.e. the number of releases before re-application is necessary.

Ease of release

To determine the ease of release, a method has been adopted from one originally devised by Mostoroy et al (ref. 3) which was used to measure the fracture energy of bulk epoxy and epoxy-to-metal joints. The test uses a Tapered Double Cantilever Beam (ASTM D3433) (ref. 4) to determine the fracture energy, G, required to separate a polymer from a coated substrate. Unfortunately, the two-dimensional sample configuration leads to very low adhesion values to be obtained from rubber formulations. A more aggressive adhesive was found to be a room temperature curing epoxy (Dexter Hysol RE2039/HD3490) and was thus used to compare the release efficiency of the four SPMRAs. Experimental tests on other molds have shown a close correlation to exist between the relative ease of release of an epoxy with that of a rubber.

The TDCB consists of two identical metal plates. One side of each plate is flat, the other tapered over a distance of 150 mm from a thickness of 10 mm to 4 mm at the opening end. The tapered nature of the beam enables its compliance to increase linearly with distance along the beam. Thus, if the two flat surfaces are adhered and then pulled apart, the energy available to induce and propagate a crack is constant over the length of the beam. This enables the critical energy required for crack propagation, Gc, to be calculated without knowledge of the crack length, a.

The expression that relates the critical crack extension force, Gc, with the applied load, P, and crack length, a, is:

[Mathematical Expression Omitted]

Calibration of each beam pair enables [partial derivative of] C/ [partial derivative of] a to be determined and thus Gc can be calculated. The advantage of using Gc over the actual release force, P, is that Gc is independent of beam construction and thus is a function of release agent and adhesive.

Experimental - The flat surface of each TDCB plate was polished with a sheet of 1200 grade emery paper to obtain a standard roughness, clean surface. The release agent to be examined was applied via an airless spray gun to the flat surface of each TDCB plate (preheated to 150 [degrees] C). Three coats were applied to give a final film thickness of 0.4 - 0.5 [mu] m. The release agent was cured at this temperature for a length of time determined on the DMTA. A small piece (100 x 20 x 0.05 mm) of polyethyleneterephthalate (PET) was placed at the opening end of the lower plate so that its leading 100 mm edge lay over the tapered region of the plate. A PTFE spacer (100 x 10 x 1.5 mm) was placed in the center of the PET film and a second at the opposite (10 mm thick) end of the plate. A second sheet of PET film was then placed on top of the PTFE and directly over the first piece. These PET sheets act as constant pre-crack regions allowing a straight polymer/coating interface to be formed that extends over the full width of the beam and lies perpendicular to the direction of crack propagation. The epoxy resin was carefully poured into the region between the two PTFE spacers on the lower plate. The second plate was then carefully aligned over the first. After allowing the epoxy to cure, the plates were connected to an Instron testing machine (fitted with a 1 kN load cell) and separated at a rate of 5 mm/min. The test was repeated using four sets of TDCB plates; 20 releases being obtained from each set of plates for each of the four release agents. The maximum load, P max, was recorded from each text and the average from the four plates obtained. Knowing the average applied load, P, the specimen width, b, and [partial derivative of] C/ [Partial derivative of] a, enables to be calculated from equation 1. The results are presented in figure 2.

Results and discussion - It has been determined experimentally (ref. 5) that adhesive/release agent combinations that have values for Gc of below 30 J/[m.sup.2], provide what can be classed as an adequate release, i.e. a reasonable force is required (in the case of a steel TDCB combination, up to 250 N) to provide release, but is sufficiently low as to prevent cohesive failure within most polymers. It is below 20 J/[m.sup.2] that the ease of release becomes sufficiently low as to be described as being industrially acceptable. Values for Gc below 10 J/[m.sup.2] can be classed as providing good or easy release and is the region aimed for when formulating a release agent. Above 30 J/[m.sup.2] results in difficult release and one which would not be tolerated in industrial release applications. From the above classification, the values for Gc obtained for each of the release agents (figure 2) illustrate that they all give an easy release to this type of epoxy resin. (In actuality, the adhesive force was too low to be recorded for the first ten releases from the water-based Product A). Experiments using a peroxide cured EPDM or NR on the TDCB apparatus have given much lower adhesion forces; the plates often separating prior to Instron testing.

Clearly, Product A offers excellent release for at least 20 cycles. In fact, over 50 releases can be obtained from a 0.5 [mu] m coating of Product A on this apparatus before the Gc value reaches 20 J/[m.sup.2]; the figure above which the ease of release can be classed as industrially unacceptable. Actual numbers of releases on industrial molds will depend on the molding process, type of polymer, etc.

The TDCB test has proved to be accurate and reproducible. Release parts from this test are suitable for use for the assessment of release agent transfer.

Slip

For easy part removal, not only must a release agent provide easy release, but it must also act as a lubricant (i.e. provide slip). This is especially important in rubber molding where the release agent must reduce the friction between the rubber and the mold to facilitate easy release.

Experimental - Release agent slip was determined on a standard slip tester (Instrumentors, Inc.) using a procedure adapted from ASTM D-1894-78 (ref. 6). The force required to pull an EPDM coated 500g weight over a TDCB plate, coated with a release coating, was determined. The test was used in conjunction with the release test so that the slip after each of the 20 releases could be recorded. The force obtained was compared to the force required to pull the weight over an uncoated TDCB plate. The values given in figure 3 are the percentage reduction in slip resistance compared to the uncoated plate.

Results and discussion - The results presented in figure 3 illustrate that Product A offers higher slip than any of the three solven-based formulations. This slip remains relatively constant throughout the 20 releases illustrating the durability of the coating. The slip of each coating decreases slightly after the first 1-2 releases. Microscopic examination of the film at this point suggests this is to be due to the removal of asperity heights thus leading to increased area of contact on subsequent tests.

Transfer

The detection of release agent transfer depends largely on the sensitivity of the technique. Clearly no release agent (or any other material) can be accurately classed as completely non-transferring. If this were the case, it would never wear out. Surface sensitive analytical techniques such as secondary ion mass spectroscopy, auger electron spectroscopy and x-ray photoelectron spectroscopy will always detect some degree of transfer. The important consideration, therefore, is whether the transfer is contaminating, i.e. will it decrease painting or other adhesion processes to a significant extent? The importance of the term significant is that the degree of acceptable transfer varies from process to process; what may be significant transfer for one process (or set of operating conditions), will be insignificant for another. Products C and D have been industrially accepted as providing release with no contaminating transfer and have been sold on this basis for many years. Product B, however, is known to give a degree of transfer for the first two to five releases, depending on the molding process. For most processes, this transfer is non-contaminating and does not affect subsequent bonding processes. However, as the contamination is largely a result of the transfer of the sol fraction of the coating, it can be easily removed from the surface of the released article by a simple cleaning process (e.g. a solvent degrease).

Experimental - Two experiments were used to compare the water-based release agent A with the three solvent-based products:

* Peel adhesion - Using the release epoxy plaques from the TDCB test, a 180 [degrees] peel test was performed on both epoxy surfaces using a standard pressure sensitive adhesive tape (ASTM D3330 M-83) (ref. 7). Figure 4 shows the change in tape adhesion (transfer) to the released surface for the first 20 releases.

* Paint adhesion - Again using released epoxy plaques from the TDCB test, a standard primer paint was applied to one surface and allowed to dry. Any paint dewetting was noted. Two cuts were made in the film, each about 4 cm. long, intersecting near their middle at an angle of 30 to 45 [degrees]. The x-shaped incision was made with a razor blade, cutting through the paint to the substrate. The adhesion of the paint was determined according to ASTM D3359-87 (ref. 8). The adhesion was rated in accordance with the following scale:

* 5A - No peeling or removal.

* 4A - Trace peeling or removal along incisions.

* 3A - Jagged removal along incisions up to 1.6 mm on either side.

* 2A - Jagged removal along most of incisions up to 3.2 mm on either side.

* 1A - Removal from most of the area of the X.

* 0A - Removal beyond the area of the X.

The results obtained from the paint adhesion to the first 20 TDCB released parts are presented in table 1.

Results and discussions - Both tests show good correlation producing a sensitive and reproducible means for detecting release agent transfer. The water based Product A shows equivalent transfer, according to the two tests used, to Product B; slightly more than for C and D. Consequently, for the majority of processes no part cleaning will be required. It is recommended that a solvent degrease be used for the first 2-5 releases for critical bonding applications.

Conclusions

From the data presented above, the newly developed water-based release agent detailed in this article has much to offer the rubber industry. It has a very low order of toxicity, very low VOC, is non-flammable and actually outperforms existing solvent-based release agents in terms of numbers and ease of release. The only disadvantage, albeit slight, is the transfer observed on the first 2-5 releases. Current research and development is being targeted on this property.

The transition from solvent-based release agents to water-based will be largely driven by increased restrictions on the use of VOCs and CFCs. However, any change will also require a degree of adaptation on the part of the operator and in operating conditions. For example, the wetting and film formation of any water-based formulation, when applied at 150 [degrees]-180 [degrees] C, is negligible due to the instantaneous evaporation of the water at this temperature. A complete uniform film can only be obtained by the application of successive coats (typically 4 to 6) via spray equipment capable of fine atomization. Otherwise, a rough irregular film is formed whose appearance is reproducd on the released part. When applied correctly, water-based release agents can be used as easily and consistently as their solvent-based equivalents. Consequently, this emergent technology will greatly benefit the rubber industry and other areas of polymer processing. [Figures 1 to 4 Omitted] [Tabular Data Omitted]

References

[1]Noll, W., "Chemistry and technology of silicones," Academic Press, 1968. [2]Rigby, M., Ph.D. thesis, UMIST, 1988. [3]Mostoroy, S. and Ripling, E.J., J. Appl. Polym. Sci., 10, 1351, (1966). [4]ASTM D3433-75, "Annual book of ASTM standards," Vol. 15.06, (1985). [5]Clayfield, T., Plastics and Rubber Institute, Adhesion Conference Preprints, Sept. 1987. [6]ASTM D1984-78, "Annual book of ASTM standards," Vol. 08.02, (1986). [7]ASTM D3330 M-83, "Annual book of ASTM standards," Vol. 15.09, (1987). [8]ASTM D3359-87, "Annual book of ASTM standards," Vol. 6.01, (1987).