Epoxidized Natural Rubber with Surface-modified Silica as a Filler

The discussion about the environmental impact of micro plastics is not new, it has been going on since 2008 the latest. However, since the Fraunhofer Institute published its study in June 2018 [1], identifying tire-road wear as the biggest source of micro plastics release to the environment, tire components using synthetic materials are in discussion. Most tire compounds are blends of synthetic rubber (e.g. BR, SSBR, ESBR), sometimes with Natural Rubber. Further synthetic components are the reinforcements, which however usually are not part of tire wear and thus are not in focus of the micro plastics discussion. However, their carbon footprint is quite high.
Engineers and chemists involved in tire compound development have been looking for alternatives to synthetic rubbers. Natural Rubber, though available in many grades is not a general alternative as its glass transition temperature Tg is low (-73°C). For properties like wet and dry braking rubber grades with a high glass transition temperature are needed. This paper reports investigations involving epoxidized Natural Rubber, a sustainable rubber with negative carbon footprint and surface modified silica fillers, which are designed to offer efficient filler to polymer bonding.

Epoxidized Natural Rubber and bonding mechanisms to fillers

Epoxidized Natural Rubber (ENR) is a slightly chemically modified Natural Rubber. A part of its double bonds is converted into epoxy groups, usually between 20 and 50%. With the content of epoxy groups, the polarity of the polymer rises and thus the glass transition temperature is increased. ENR 25
(with 25% of all double bonds converted to epoxy groups) has a Tg of -43°C, ENR50 -27°C.
ENR itself or in combination with NR could be used to produce tire compounds completely made from renewable resources. The carbon footprint of such compounds would be very low, if not even negative.
However, to gain wide acceptance in the rubber industry, ENR needs to prove its performance is similar to or better than conventional systems. Therefore, an effective bonding between the polymer and the filler is essential. Generally, for linking ENR to fillers, mainly silica, there are two options:

Fig. 1: Bonding mechanisms between Silica and ENR. Copyrigth: [5] Anke Blume et.al

• Direct links of ENR epoxy groups to filler surface silanols (A)
• Sulfur silanes (tetrasulfane, disulfane, mercapto) to link to the ENR double bonds (B)
• Amino- or epoxy-functional linkers to react with the ENR epoxy groups (C) (Fig. 1)

Direct links have been evaluated by several investigations [2]/[3]/[4]/[5] which found that only about 15% of all epoxy groups covalently link to the silica surface. This is good enough to get good rolling resistance and wet skid properties. (Fig. 2)
However, if any reported, results were not good enough to get good abrasion resistance. Even a compound adding about 1% sulfur silane as coupling agent [2] according to mechanism (B) did not show sufficient abrasion resistance for a tire tread compound.

Fig. 2: Wet skid / Rolling resistance performance of ENR compounds.

Since its introduction to the market in the early 1990’s ENR has rarely been used for commercial applications. Difficulties in compounding ENR with silica and silanes have been observed during the years and contributed to the low usage. When mixing ENR and silica, usually viscosities rise so much, that compounds get brittle and smooth processing in downstream equipment fails. The use of silanes helps, but risk of pre-scorch is always high. Literature interpretation about these phenomena include hydrogen bonding between epoxy groups and silica silanol groups, ring-opening reactions of the epoxy groups, silane-silane side reactions, direct reactions and networking of epoxy groups within the polymer chain etc [2]/[5]/[7].
One way out of these problems could be to force the filler-to-polymer-bonding reaction into the right direction by immobilizing the coupling agent on the filler surface before mixing the compound. In the best case the coupling agent has finished its covalent reaction with the silanol groups of white fillers already in a separate reactor, so that in the mixer only distribution and dispersion take place. That way side reactions of the silane are suppressed and by proper process management compounding should be easy. The coupling between ENR and functionalized filler surface should take place at vulcanization temperatures only.
As a side effect mixing cycles for silica compounds could be as short as by mixing carbon black. No reaction time during mixing needs to be foreseen to ensure proper reaction of the silane coupling agent to the filler surface. No energy should be wasted to heat up and mix out the silica water shell and alcohol side products generated by the surface reaction.
A rubber compound test program was laid out to check for the mixing behavior and compound properties of ENR 25 with surface-modified silica grades. Focus of this study has been to TESPT-modified silica grades. A further study using other functionalities according to mechanism (C) has been conducted and can be accessed through the author.

Compound properties – Formulation

Fig. 3: Compound compositions. Copyright: Behn Meyer

To secure comparable results all compounds where mixed including the same raw materials (Fig. 3). Group 0 is a reference group based on Natural Rubber SMR10 and a blend of silica with BET 140 and carbon black N774. This group is using silane TESPT as a coupling agent in an appropriate dosage of 8% (w/w) based on the silica weight. Silanization is accomplished during mixing. Group 1 contains ENR25 (Ekoprena from FGV, Malaysia) and is according to literature [2] as a reference compound. Here a low silane dosage is dispersed and reacted during mixing. Group 2 is equivalent to Group 1 with higher silane dosage, 8% (w/w) like Group 0.
Group 3 is using a surface-modified silica. A BET 140 silica equivalent to the one in Group 0 was pre-reacted with 8% (w/w) TESPT in a reactor and added to the mix. The dosage has taken into account that volumes of water and ethanol are not present in that product anymore, but need to be mixed out of Groups 0 to 2. Group 4 is nearly equivalent to Group 3, just the BET surface of the base silica has been 185 m^2/g and silane dosage is 10% (w/w).

Compound properties – Mixing cycle

Fig. 4: Mixing procedure. Copyright: Behn Meyer

The masterbatch mixing cycle (Fig. 4) displayed was used as Group 0 to 2 need reaction time at elevated temperatures (maximum 150°C measured in the compound) to properly react the silane at the filler surface. Groups 3 and 4 would be possible to mix at a total mixing time of 4 – 5 minutes. The Final mix has been conducted on a two roll mill with a total mixing time of 3 minutes and a maximum temperature of 100°C.

Mixing results and cure characteristics

All groups showed a smooth surface after masterbatch mixing, no porosity and no brittleness. Mooney Viscosity of all FM groups varies between 59 and 77 MU, which is well processable. Groups 3 and 4 don’t show very high viscosities (>130 MU) after masterbatch mixing like reported for similar compounds in [7]. This supports the filler surface is modified in a way the silanol groups are not accessible to form hydrogen bonds anymore.

Fig. 5: Rheometer data. Copyright: Behn Meyer

The reference Group 0 with NR shows the highest modulus (Fig. 5), whereas all ENR25 groups reach at best 66% of it. As the MH level is a result of several interactions and forces, this might be due to a variety of reasons, e.g. better compatibility of silica in ENR and thus lower filler-filler interactions as well as a higher filler-polymer bonding in NR. NR offers 100% of its double-bonds for bonding, whereas in ENR25 only 75% double-bonds remain after epoxidation. The low MH of Group 1 might hint to the influence of the silane concentration. Group 2 shows the typical “marching modulus” of a highly silane coupled dry mix compound and the high T90 time accompanied with that. Groups 3 and 4 show save TS2 scorch times at low T90 times, at high MH values and no “marching modulus”.

Tensile properties

Fig. 6: Tensile test data. Copyright: Behn Meyer

The NR reference Group 0 yields the highest tensile strength (TS) and elongation at break (EAB) (Fig. 6). A modulus at 500% elongation just can be measured and the reinforcement factor M300/M100 is quite high. This is a picture the compounder is used to for Natural Rubber, which is the material of choice when high tensile and elongation values are needed. The special “strain strengthening” effect, caused by natural components like phospholipids and proteins, add to the performance of Natural Rubber. ENR as a modified Natural Rubber has lost these components during modification and is a high Tg polymer as described before. Such polymers show lower TS and EAB values usually, and these vary in dependence of the network density of the cured compound.
Keeping in mind these effects, the data obtained for the ENR groups can be interpreted as follows. Group 1 is compounded by the lowest concentration of silane, which results in good reinforcement factor but low M300. Group 2 with 4 times higher silane concentration shows a higher network density, represented by the lower EAB and higher M100 and M300. The hardness increases a bit for the same reason.
At the same dosage of coupling agent, Groups 3 and 4 optimize TS and EAB, as the full dosage of coupling agent is already allocated at the filler surface and nearly all of it is available for polymer-filler bonding. In those groups which are mixed conventionally, up to 10% of the silane dosage is lost due to silane-silane reaction according to experience of the author. A possible additional effect, which however could not be quantified properly here, is the interaction of ENR polymer main chains via polar interactions or even covalent bonding by their epoxy groups.

Compression set and rebound

Fig. 7: Compression test and Rebound. Copyright: Behn Meyer

Compression set data for the ENR groups are remarkably low, indicating a strong network (Fig. 7). At high temperatures the low compression set of NR is gone, obviously some of the bonds present at low temperatures get loose. This might be caused by a secondary network built up by proteins and phospholipids.
Rebound values are lower for ENR25 compounds than for the NR compound. This is caused by the high Tg of the ENR compounds and indicates the potential for a good dry and wet skid. In tendency Groups 3 and 4 show the lowest rebound, thus potentially the highest dry and wet skid.

Abrasion resistance

Fig. 8: DIN Abrasion Index. Copyright: Behn Meyer

The test compounds have been tested for DIN Abrasion (Fig. 8). This quick and easy test is good as a first indication of compounds based on the same or similar polymers. Its significance for tire wear is generally low as it is a very high severity abrasion test, which is only valid for very harsh conditions during driving on the road. However, big differences give a hint on effects which must be verified by tire test programs usually.
As expected, the NR reference compound shows a very high abrasion index value, corresponding to a very low volume wear. This is the reason why NR compounds are usually selected for tires and non-tire products under high abrasion conditions.
Not surprisingly the ENR compound Group 1 is poor in abrasion, as its Tg is high and its network density relatively low as previously discussed. Group 2 with its higher network density is considerably better and is even topped by Group 3 with identical composition, but the use of surface-modified silica.
Surprisingly, Group 4 improves abrasion resistance again. Probably the higher silane dosage rather than the higher BET surface shifts the index near to that of Natural Rubber. This program doesn’t compare the performance of this compound to any high Tg synthetic rubber, but assumption is that the abrasion performance of a pure SSBR should be lower. Results obtained in our lab are usually around an index of 110, best-case using HD silica with extremely high BET and CTAB values around an abrasion index of 120.

Conclusion

The test program conducted shows that it is possible to mix ENR compounds without much hassle but yielding good to excellent compound performance. Especially those compounds where surface-modified silica has been used, show a very interesting set of physical data, which would qualify them for use in tire tread compounds and a variety of non-tire compounds.
It would be wishful to test the dynamic-mechanical properties of such compounds and to test them on the tire for final judgement. However, this compounding route promises to solve the basic wear problem of ENR compounds and to help exploring the full potential of this sustainable rubber type. It could help to ease the pressure on rubber products as a source for microplastic.The producer of Ekoprena ENR, Felda Global Ventures Holding Bhd (Malaysia) and Behn Meyer Europe recently entered into an agreement to cooperate on the commercialization of both ENR and Surface-modified silica to tap its full potential.

Tire Technology Expo 2020 Stand C645

Literature:

[1] J. Bertling, R. Bertling, L. Hamann: Kunststoffe in der Umwelt: Mikro- und Makroplastik. Fraunhofer UMSICHT, Juni 2018
[2] Siti Salina Sarkawi et.al.: „Improving Cure Reversion Resistance and Physical Properties of Silica Filled Epoxidised Natural Rubber Compound for Tyre Treads”, Australian Journal of Basic and Applied Science, October 2014
[3] A K Che Aziz et.al., “Silica-Reinforced ENR tires”, Tire Technology International 2018, Annual Review 2018
[4] Paul Brown (TARRC): „Epoxidised Natural Rubber in Retreaded Tires“, Tire Technology Expo and Conference, Hannover 2016
[5] Anke Blume et.al.: „Interactions between Silica and Epoxidized Natural Rubber with and without Silane”, Tire Technology Expo and Conference, Hannover 2016
[6] Product information Malaysian Rubber Board / Felda Global Ventures on Ekoprena, 2016
[7] Siti Salina Sarkawi et.al.: “Properties of Epoxidized Natural Rubber Tread Compound: The Hybrid Reinforcing Effect of Silica and Silane System”, Polymers & Polymer Composites, Vol. 24, No. 9, 2016.