Performance of recycled aluminosilicate filler in a passenger tire tread compound
The Authors
Christopher Robertson is principal consultant at Polymer Technology Services L.L.C. in Akron, a company that he founded in 2021 to provide technical consulting, expert witness services, and training for the elastomer/rubber industry. He has 25 years of experience as a materials scientist and engineer in the tire, synthetic rubber, rubber additives, and plastics industries.
Bryan Geary is the chief operating officer of Vecor Technologies and president and co-founder (2020) of IntoCeramics in Houston. He held increasingly responsible positions in manufacturing operations and R&D over a 25-year span. Geary’s expertise includes the complex technical aspects of ceramic resources, materials, and manufacturing, as well as a thorough knowledge of traditional ceramic manufacturing techniques, business improvement, plant expansions and re-fits, along with new plant construction.
Ryon Lasiter has over 25 years of experience in manufacturing operations, including the last 12 in the ceramics consulting and manufacturing space. He has held increasingly responsible positions, including production operator, front line supervisor, plant foreman, lab manager, business unit and operations manager, and now serves as technical director at Vecor Technologies and IntoCeramics.
Executive summary
Vecor Technologies has developed a new reinforcing filler for rubber (VC-R) that is composed of 70-percent recycled waste stream aluminosilicate, using additional proprietary processing steps to upgrade the filler for demanding rubber applications like automobile tires.
The VC-R was used to replace 30 percent by volume of precipitated silica in a passenger tire tread formulation in a laboratory mixing and testing study. Tensile strength distribution and Ueshima Lambourn-type wear testing were used to verify that the formulation containing VC-R had comparable tensile strength and wear resistance relative to the control compound with full silica loading.
In addition, there was evidence of increased silanization during mixing for the tread compound with the VC-R filler, which translated to a better dynamic mechanical analysis (DMA) predictor for rolling resistance at constant wet traction predictor.
Three developmental versions of VC-R were evaluated, which enabled an optimized material to be selected and moved forward in the commercialization process.
Introduction
There is rapidly growing interest in replacing traditional raw materials used in manufacturing tires with more sustainable alternatives, and top global tire companies have announced challenging sustainability targets.¹ ²
TECHNICAL NOTEBOOK
Edited by John Dick
Reinforcing fillers—primarily carbon black and precipitated silica—represent about 25 percent of the weight of tires, so the development of recycled and bio-based fillers is an important part of the sustainability push. The main use of precipitated silica is in the tread cap rubber compound of passenger car radial (PCR) tires at a loading of 30 to 35 weight-percent.
Silica from rice husk ash is one alternative that has emerged to replace conventional precipitated silica,³ but other sustainable options are needed to supply the vast demand.
Our research detailed in this article demonstrates the effectiveness of a new filler, VC-R from Vecor Technologies,⁴ for replacing silica in tires. VC-R is composed of 70-percent recycled waste stream aluminosilicate, and we will discuss the performance of this material in a PCR tread compound as a partial (30 vol-percent) replacement for silica.
Experimental details
Rubber mixing was conducted with a 1.6-liter Banbury mixer using a three-pass mixing scheme. The fill factors used were 0.634, 0.628 and 0.613 for the first, second and third passes, respectively.
The first pass used a mixing speed of 30 rpm and included a moderate silanization reaction period of 3 minutes at 140 °C before dropping the batch, with mixing speed adjusted as needed to maintain the temperature. The second pass involved mixing at 30 rpm until the batch reached a drop temperature of 140 °C. The third pass (productive mixing stage with curatives added) involved mixing at 20 rpm until a drop temperature of 110 °C.
A lab two-roll mill was used for subsequent mixing and sheeting of the rubber after each pass.
An Alpha Technologies Moving Die Rheometer (MDR) was used for measuring cure rheology at 160 °C (ASTM D5289). In terms of generating samples for physical testing, curing (vulcanization) was performed in a compression molding process in a heated press for 12 minutes at 160 °C to produce 2-mm thick sheets/slabs and 17 minutes at 160 °C to produce Ueshima wear testing wheels (20 mm thick; 70 mm outer diameter; 30 mm inner diameter).
An AB-2012 FPS (Field Performance Simulation) Wear Testing System manufactured by Ueshima was used for making wear/abrasion measurements at room temperature, using 20 N normal load and sample rotational speed of 150 m/min. The slip ratio was varied in the order of 3 percent, 10 percent and then 5 percent, with weight loss measurements acquired at each slip ratio.
The dynamic mechanical analyzer (DMA) used was Viscoelasticity Analyzer VR7120 by Ueshima. The following DMA testing conditions were used in tension deformation mode: 2-percent static strain; 0.5-percent dynamic strain amplitude; 10 Hz frequency; temperature sweep conducted from -70 to +70 °C in 2 °C increments/steps (equilibration at each temperature).
Dynamic strain sweeps from 0.07-percent to 100-percent strain amplitude were performed using an Alpha Technologies Rubber Process Analyzer (RPA) at 60 °C and 1 Hz after curing the specimen for 12 minutes at 160 °C.
Tensile stress-strain testing and Shore A hardness testing were performed at 23 °C in accordance with ASTM D412 (type C specimen geometry) and ASTM D2240, respectively.
Results and discussion
Three developmental versions of VC-R filler, denoted as VC-R0, VC-R1, and VC-R2, were used as 30 vol-percent replacement of precipitated silica in a lab-scale rubber mixing and testing study.
The filler characteristics are reported in Table 1, where it can be seen that the specific surface area (SSA) ranged from about 80 to 100 m²/g for the VC-R fillers, in contrast to 160 m²/g for the Zeosil 1165MP precipitated silica. The lower SSA for VC-R was the motivation behind using the VC-R as a 30 vol-percent (minor) substitute for silica in the PCR tread.
The potential silica replacement level could be appreciably higher, especially considering that a newer version of VC-R was recently produced with SSA in excess of 150 m²/g. Advanced processing was applied to VC-R1 and VC-R2 to reduce the D50 and D90 particle sizes to about a factor of 10 smaller than the baseline processed filler VC-R0, toward better mechanical strength and wear resistance to be discussed later.
The PCR tread compound formulations are shown in Table 2. Filler reinforcement of rubber is strongly influenced by the volume percent of the filler in the compound. Therefore, to ensure a fair comparison with silica, we increased the parts-per-hundred-rubber (phr) weight loading of VC-R to account for its higher specific gravity relative to silica (2.55 vs. 2.00; see Table 1). This resulted in 30.6 phr of VC-R being substituted for 24 phr of the silica.
A three-stage (three-pass) mixing scheme was used to mix the compounds, as described in the “Experimental details” section. A warm-up compound identical to the control formulation (Table 2) was mixed prior to the control and VC-R compounds. A moderate silanization reaction period of 3 minutes at 140 °C was applied in the first mixing stage.
The time-dependencies of batch temperature and integrated power during the first mixing stage are given in Fig. 1. Relative to the control, compounds with VC-R fillers had slower temperature ramp up to the silanization hold in the first pass due to lower viscosity (see Compound, page 34).
Mooney viscosity results later), which could be easily accommodated for in future mixing by using a higher mixing speed (rpm). The integrated power was accordingly about 10 percent higher on average for the VC-R fillers versus the control. The VC-R fillers did not significantly affect mixing behavior relative to the control in the second and third mixing stages. For all three mixing passes, the compounds containing VC-R were similar to the control in terms of processability on the two-roll mill, general appearance, and qualitative tackiness.
Replacement of 30 vol-percent silica by VC-R fillers in the PCR tread formulation resulted in about a 20-percent decrease in Mooney viscosity (Fig. 2a). Such a viscosity reduction with incorporation of VC-R could be advantageous in tread extrusion, particularly for formulations with very high silica loadings where high viscosities can make processing challenging. The general appearance of the final sheeted-out compounds containing VC-R were qualitatively like the control (Fig. 2c).
The MDR cure rheology curves are plotted in Fig. 2b. The three VC-R compounds had similar cure kinetics, with an average 50 percent cure time (t50) of 3.8 min, which was 18 percent faster than the control compound (t50 = 4.7 min). This faster t50 for VC-R was advantageously paired with a slower average scorch time (ts2; 2 dN-m rise relative to the minimum torque) of 2.4 min (71% increase) relative to the control value of 1.4 min.
Both filler flocculation at very short times and marching modulus near the end of curing were evident for the control, which are hallmarks of silica-silane filler system, particularly for low to moderate extents of silanization (silane reaction with filler surfaces) during mixing. Replacing 30 vol-percent of silica by VC-R caused both of these undesirable phenomena to nearly disappear from the cure rheology responses.
These reductions in filler flocculation and marching modulus are similar to the effects of using intensive silanization reaction conditions during mixing (160–170 °C for 3 min),⁵ which suggests that the VC-R fillers somehow acted to catalyze or accelerate the silanization reaction at the moderate silanization conditions used in our mixing (140 °C for 3 min).
Payne Effect and Dynamic Mechanical Analysis
Significantly reduced filler networking, manifesting as a smaller Payne Effect,⁶ was evident from the RPA strain sweeps for the VC-R compounds compared to the control (Fig. 3). This is entirely consistent with the absence of obvious filler flocculation in the short-time region of the MDR cure rheology for the formulations with VC-R.
This desirable decrease in the Payne Effect suggests that the VC-R incorporation somehow led to more polymer-filler interactions versus filler-filler interactions, possibly by promoting the silane reaction with filler surfaces during mixing as suggested previously when discussing the cure rheology data.
An alternate explanation is that the VC-R filler has intrinsically better interaction with the polymers and/or less tendency to form a filler network. This latter thought is not supported by the fact that the Payne Effect reduction was so large in magnitude (approximately 60 percent reduction in low-strain limit for G′) with only replacing 30 percent of the silica by VC-R. Our previous research on an earlier version of VC-R showed that replacing only 12 percent of the silica had a measurable effect on reducing the Payne Effect.
Dynamic mechanical analysis (DMA) is an important tool for predicting various tire performance characteristics from the temperature dependence of viscoelastic behavior for the tread rubber material.⁷ ⁸ DMA temperature sweep results are shown for the compounds in Fig. 4.
A lower tanδ at 60 °C, which is predictive of better fuel economy (lower rolling resistance), was found for the VC-R formulations relative to the control, and this observation is in agreement with the lower Payne Effect from the RPA strain sweep data. The tanδ at 0 °C is a common approximate predictor of wet traction (higher is better), and the VC-R materials and control had similar values (Fig. 4a).
The reduced Payne Effect for the VC-R compounds, which caused the improvement in rolling resistance predictor, also decreased the dynamic storage modulus (E′) in the rubbery state above the glass transition.
Concerning the implications on predicted tire performance, such a reduction in E′ gave an improvement in winter traction but a decrease in dry handling/cornering, which is indicated in Fig. 4b. This trade-off is the common consequence of lowering the Payne Effect for improving rolling resistance.
Physical Properties
We now consider the physical properties, which are summarized in Table 3. A roughly 10-percent decrease in Shore A hardness was observed for VC-R compounds compared to the control material.
Finite element analysis by Qi, Joyce and Boyce⁹ of the Shore A indenter deforming a material with 60 Shore A hardness showed that the simulated strain ranged from about 10 to 120 percent with an approximate average of 50 percent. Considering the 50-percent strain region in the RPA strain sweep (Fig. 3), it is clear that this decrease in hardness was a consequence of the lower Payne Effect for the VC-R compounds.
The results for the stress at 50-percent strain (M50) from standard tensile testing were consistent with the hardness and RPA results. The stress values at 100 percent (M100), 200 percent (M200), and 300-percent strain (M300) showed that the VC-R materials had very similar stiffness behavior as the control. The ratio of M300/M100 is often used as an indicator of the level of filler reinforcement of rubber, and the VC-R materials had equal or slightly higher M300/M100 than the control.
Because the MDR cure rheology data for the control included the contributions of filler flocculation and marching modulus that were not significantly present in the VC-R compounds (Fig. 2b), we could not use the cure rheology results to make a judgment about crosslink density. However, the similar values of elongation at break from tensile testing (Table 3) provided evidence that the crosslink density did not appreciably change upon substitution of silica by VC-R.
Tensile strength of rubber is sensitive to the presence of crack precursors such as undispersed ingredients or large particles/inclusions.¹⁰ ¹¹ Accordingly, the large particle size for VC-R0 reported in Table 1 (D50 = 5.2 μm; D90 = 13.4 μm) did translate to a reduced tensile strength relative to the control (Table 3). VC-R1 and VC-R2 had significantly smaller particles (sub-micron D50) than VC-R0, which consequently resulted in tensile strengths that were statistically the same as the control strength (within the standard deviations) when these versions of VC-R were incorporated into the tread rubber.
With the tensile testing of several replicates, 10 in our case, a tensile strength distribution can be created to characterize the failure distribution,¹⁰ ¹¹ and such results are given in Fig. 5 for the compounds investigated. The strength deficiency of VC-R0 is quite clear in this analysis, and the improvements in VC-R1 and VC-R2—due to advanced processing to reduce particle size—are evident in strength distributions that overlap substantially with the control’s strength population.
Tire wear can be difficult to accurately predict from lab abrasion testing of a tread compound. Good predictions can come from advanced Lambourn-type wear testing systems like the Ueshima AB-2012 FPS (Field Performance Simulation) when operated with reasonably low slip ratios in the 3-to-10-percent range. Such testing was performed at Ace Laboratories on our materials using 20 N normal load and sample rotational speed of 150 m/min, with data shown in Fig. 6 for slip ratios of 3, 5, and 10 percent.
Examples of the large rubber wheel specimens that were tested are shown in Fig. 7a. The VC-R0 gave significantly reduced wear resistance (faster wear rate) compared to the control at all slip ratios, which is consistent with the reduced tensile strength behavior from larger particles that was already discussed.
The VC-R1 and VC-R2 compounds had statistically equivalent wear resistance to the control, with the exception of a small increase in wear rate that was noted for VC-R2 at the highest slip ratio of 10 percent. Optical microscope images of wear surfaces for VC-R1 and control after testing at 5-percent slip ratio showed similar wear pattern characteristics, as illustrated in Fig. 7b, which confirmed that the partial replacement of silica by VC-R did not change the nature of wear.
These wear results are very encouraging, as achieving suitable wear resistance has been one of the major challenges in replacing traditional reinforcing fillers with more sustainable alternatives in the tire industry.
Summary
An overview is presented in Table 4 of tire performance predictors from the DMA and wear testing results that were discussed. Each lab predictor value was used to calculate a ratio with respect to the comparable lab predictor for the control and multiplied by 100 to form a comparative index.
For a lab predictor where it is better to have a higher value in terms of tire performance, then the value for the control was in the denominator of the ratio. For a lab predictor where it is better to have a lower value in terms of tire performance, then the value for the control was in the numerator of the ratio.
The three VC-R fillers gave improved rolling resistance (lower tanδ at 60 °C) at essentially constant wet traction compared to the control, which was accompanied by a common trade-off in winter traction and dry handling due to the reduction in the Payne Effect (lower E′) for the compounds with VC-R.
The FPS wear results allowed us to discriminate among the three developmental versions of VC-R to select VC-R1 as the best candidate for moving forward in the commercialization process. VC-R1 gave essentially equivalent wear index to the control. The wear index was 54 percent deficient for the compound with VC-R0, which was consistent with its lower tensile strength and large particle size (D50 = 5.2 μm; 10 times larger than D50 for VC-R1 and VC-R2). VC-R2 showed a small reduction in wear resistance, which—in combination with the fact that this filler version had lower recycled content (66.5 percent) than VC-R1 (70 percent)—convinced us to designate VC-R1 as the best filler version.
A radar plot comparison of tire predictors for VC-R1 versus the control in Fig. 8 gives a visual synopsis of our research. The VC-R1 was subsequently produced in 10 kg quantity for ongoing evaluation by a major tire company.
Acknowledgments
Vecor Technologies (Rozelle, NSW, Australia; Houston) is gratefully acknowledged for supporting this research and approving this publication. We thank Ace Laboratories in Ravenna, Ohio, for high-quality services in rubber mixing and testing that were essential for this project. We appreciate helpful technical advice from Donald Picard, technical consultant at Wendacon L.L.C. (Sandown, N.H.), on application development projects in the rubber and plastics areas.
References
- Goodyear Commitments, accessed Sept. 17, 2025.
- Bridgestone Environment & Resources, accessed Sept. 17, 2025.
- W. Waddell, T. Lin, S. Lee, C. Wu, and D. Liao, Sustainable Silica from Rice Husk for the Green Tire Tread, Rubber News, Technical Notebook, Sept. 30, 2024.
- Vecor Technologies, accessed Sept. 17, 2025.
- J. Jin, J.W.M. Noordermeer, W.K. Dierkes, and A. Blume, The Effect of Silanization Temperature and Time on the Marching Modulus of Silica-Filled Tire Tread Compounds, Polymers 12, 209 (2020).
- A.R. Payne, The dynamic properties of carbon black-loaded natural rubber vulcanizates. Part I, J. Appl. Polym. Sci. 6, 57–63 (1962).
- S. Futamura, Effect of material properties on tire performance characteristics—Part II, Tread material, Tire Sci. Technol. 18, 2–12 (1990).
- N. Warasitthinon and C.G. Robertson, Interpretation of the tanδ Peak Height for Particle-Filled Rubber and Polymer Nanocomposites with Relevance to Tire Tread Performance Balance, Rubber Chem. Technol. 91, 577–594 (2018).
- H.J. Qi, K. Joyce, and M.C. Boyce, Durometer Hardness and the Stress-Strain Behavior of Elastomeric Materials, Rubber Chem. Technol. 76, 419–435 (2003).
- C. Robertson, C. Robin, and E. Britton, Using tensile strength distribution to detect undispersed filler and other crack precursors in rubber, Rubber World, Jan. 2024, pp. 32–37.
- D.H.C. Wong, F. Ignatz-Hoover, A. Childress, G.L. Jackson, and A. Padmakumar, Quantifying Sulfur Dispersion Using Population Survival Analysis of Tensile Strength, Rubber Chem. Technol. 96, 214–225 (2023).