SILICA GRAFTED WITH EPOXIDIZED LIQUID POLYBUTADIENES: ITS BEHAVIOR AS FILLER FOR TIRE TREAD COMPOUNDS
ABSTRACT
Achieving high polymer-filler interaction and lowering the energy consumption required to disperse precipitated silica in a rubber matrix have been the main motivations behind the recent interest in silica-coating processes. A simple, low-cost, and environmentally friendly process is available to graft epoxidized liquid polybutadienes onto a silica surface. The polymer-grafted silicas are applied as reinforcing fillers in typical car tire tread compounds. The processability, cure kinetics, and properties of the vulcanizates are greatly affected by the degree of epoxidation, molar mass, and microstructure of the epoxidized polybutadienes used for coating the silica surface. The dynamic-mechanical behavior is superior to that of the reference compound, using bis-[triethoxysilylpropyl] tetrasulfide as a coupling agent, in stiffness and hysteresis at low temperatures, which is indicative of superior performance in wet grip and emergency maneuvers (hard-handling). Thus, the use of this reinforcement system for high-performance car tires, for which safety features should be prioritized, is promising.
INTRODUCTION
The tire industry invests a good deal of research effort on continuously improving their products for comfort, safety, durability, and fuel economy. Searching U.S. patents applied to tires during the past 4 decades (Figure 1), we observe that the durability requirement (abrasion resistance) was receiving special attention from manufacturers and researchers by the end of the 1970s. However, the number of patents related to the rolling-resistance issue (fuel economy) has assumed a prominent position in the past decade, reflecting the global concern about the level of atmospheric emissions in large urban centers and the threat of depletion of nonrenewable energy resources. Safety (skid resistance, wet grip), despite being allegedly primordial,1 has received less attention than other performance criteria.
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
The critical performance characteristics of tires depend on their constructive aspects, road conditions, and the dynamic-mechanical properties of the tread compound. “Green tire” technology, introduced at the beginning of the 1990s, allowed for the production of tire treads with durability comparable to that obtained with carbon black–filled compounds, but with lower rolling resistance and better wet traction. The green tire tread is mainly characterized by the use of silica as the main reinforcing filler and the use of an organosilane as a polymer–silica coupling agent.2
Alternatively, the literature indicates that epoxidized rubbers may exhibit efficient interaction with silica and could advantageously replace silanes. Rocha et al.3 have shown that the graphitization of silica with SBR 7 mol% epoxy, probably occurring during vulcanization, resulted in a significant reduction in filler networking (Payne effect). Later, Jacobi et al.4 described the reaction of the hydroxyl groups present on the silica surface and the lightly epoxidized polybutadiene during the mixing process. It was shown that the reaction can be easily controlled by manipulating process parameters being monitored by torque evolution. The compounds obtained by this method exhibited significant improvement in polymer–filler interaction, a 40% lower Payne effect, and a higher hysteresis at low temperatures than the compound silanized with bis-[triethoxysilylpropyl] tetrasulfide (TESPT), behavior that can result in tire treads with lower rolling resistance, greater skid resistance, and better wet grip. One drawback was that the compounds obtained had high viscosities and were considered difficult for the industry to process.
More recently, Kaewsakul et al.5 proposed the combined use of TESPT and epoxidized natural rubber (ENR) in silica-filled compounds, which can produce properties similar to those of the NR–silica–TESPT system, enabling a silane reduction of at least 50%. However, the extensibility and flexibility of the vulcanizates were significantly affected. In another approach, Saramolee et al.6 proposed the use of epoxidized low–molecular weight natural rubber (ELMWNR) as a compatibilizer in NR–silica compounds to improve their processability. When sufficient amounts of ELMWNR (28 and 51 mol% epoxy) were used, the viscosity and Payne effect of the compounds reached levels similar to those of TESPT-silanized compounds, but the mechanical properties were worse and the hysteresis in the service temperature range of a tire was higher. In the last two works mentioned, the interactions between epoxy groups and silica are mainly attributed to hydrogen bonds. However, in a recent study, Braum and Jacobi7 provided evidence that the reaction between epoxidized liquid-hydroxylated polybutadiene (ELHPB) and silica occurs through the formation of covalent bonds, the kinetic parameters of the reaction having been determined.
The aim of this work is to present a method of producing polymer-grafted silicas using epoxidized liquid polybutadienes and to investigate the effects of microstructure, molar mass, and the degree of oligomer epoxidation on the properties of the compounds, mainly for application in treads of high-performance tires.
EXPERIMENTAL
Materials
Liquid-hydroxylated polybutadiene ([LHPB], Petroflex, now Lanxess, Duque de Caxias, Rio de Janeiro, Brazil), liquid polybutadienes ([LBR-300, LBR-305, and LBR-307], Kuraray, Kurashiki, Japan), surfactant Tween 20 (Synth, Diadema, Brazil), formic acid 98% (Casa da Química Indústria e Comércio, Diadema, São Paulo, Brazil), toluene, and hydrogen peroxide (Reagen, Colombo, Paraná, Brazil) were used in the preparation of the epoxidized liquid rubbers. The highly dispersible silica (Zeosil® 1165MP, Rhodia, Lyon, France), used for surface modification and rubber compounding, presented a total specific surface of 139 m2/g (Brunauer–Emmett–Teller method8) and an external specific surface of 113 m2/g (t-plot method, using the Lippens–de Boer reference isotherm9). The other ingredients used for rubber compounding were solution styrene–butadiene rubber (SSBR) with styrenic units of 25% and 1,2-vinyl of 63%, extended with 37.5 parts of treated distillate aromatic extract (TDAE) oil (SPRINTAN® SLR-4630, Styron, Schkopau, Germany), high-cis polybutadiene rubber (cis-BR, Neodene 40, Karbochem, Sasolburg, South Africa), TESPT (Evonik, Westphalia, Germany), zinc oxide (Companhia Mineira de Metais, Três Marias, Minas Gerais, Brazil), stearic acid (Braswey, São Paulo, Brazil), polymerized 2,2,4-trimethyl-1,2-dihydroquinoline (TMQ, Nord Chemie, Cotia, Brazil), sulfur (Intercuf Industria, Campinas, Brazil), N-(1,3-dimethyl-buthyl)-N′-phenyl-p-phenylenediamine (6PPD, Flexsys, Ann Arbor, MI, USA) and N-cyclohexyl-2-benzothiazole-sulfenamide (CBS, Flexsys).
epoxidized liquid rubbers
Preparation
ELHPB and normal (nonhydroxylated) ELPB were produced in our laboratory from commercial LHPB and liquid polybutadiene (LPB), respectively, according to procedures similar to those described elsewhere10: The liquid rubbers were epoxidized in a 44% (w/w) toluene solution at 50 °C for 6 h via the performic acid method generated in situ using the stoichiometric molar ratio of reactants (H2O2:C=C:HCOOH).
Characterization
The number–average molar mass (Mn) and dispersity (Mw/Mn) of LHPB and ELHPB samples were determined via gel permeation chromatography (Viscotek TDAmax, Malvern Instruments, Malvern, UK) using tetrahydrofuran as the solvent, conventional calibration using polystyrene, and detection by refractive index. The microstructure of all liquid polymers was determined via Fourier transform infrared analysis by liquid film on KBr.11 The epoxy content was determined by 1H NMR (Inova 300 MHz, Varian, Palo Alto, CA, USA) at room temperature, using deuterated chloroform as a solvent for sample preparation. Chemical shifts are reported in parts per million (δ) relative to tetramethylsilane (Sigma-Aldrich, St. Louis, MO, USA). The epoxide group content, %Epoxy, was calculated with a peak integral at 2.7 and 2.9 ppm assigned to hydrogens in epoxy groups, and the peak integral assigned to hydrogens in alkene groups at 4.9 ppm (1,2-vinyl) and 5.4 ppm (1,4-cis and trans), following Eq. 1:
The characteristics of the commercial and modified liquid polymers are given in Table I.
Preparation of the surface-modified silicas
Silica and epoxidized liquid rubber were mixed in a ball mill (CT-242, Servitech, Tubarão, Brazil) for 5 min. The amount of oligomer added, in parts per hundred of silica (phs), was varied in such a way that the relative disponibility of epoxy groups per unit surface of silica Q was maintained at around 8 × 10−6 mol/m2. The polymer-impregnated silica thus obtained was preheated under vacuum to 105 °C and maintained for 120 min, cooled at room temperature, and then heated again under vacuum to 170 °C and maintained for 8 min. This procedure aims to promote the partial reaction between silica and epoxidized rubber (precure) and was defined based on previous kinetic study.7
Preparation of rubber compounds
A typical car tire tread formulation, based on a polymer blend of solution SBR and cis-BR, was used to evaluate the effect of the surface-modified silicas on the mechanical and dynamical properties of the vulcanizates. The formulations containing unmodified silica (sample designed as blank), silica silanized with TESPT (sample designed as reference), and polymer-grafted silicas are given in Table II. To make the rubber formulations comparable from the standpoint of vulcanization kinetics, use of low–molecular weight polybutadienes was compensated for with a proportional reduction in the amount of high–molecular weight rubbers, all while maintaining the total polymer content in the compound constant. However, given that the epoxidized polymer is selectively immobilized on the silica surface, the S-SBR/cis-BR ratio must be kept constant, so the rubber matrix composition is not significantly affected.
The compounds were mixed following a three-stage mixing procedure. In the first and second stages, an internal mixer (HAAKE PolyLab Rheomix 600, Thermo Fisher Scientific, Karlsruhe, Germany) was used. The second stage was employed only for the heat treatment of the master batch (to complete the reaction of salinization or the grafting reaction between epoxidized polymer and silica), according to the procedure described in Table III. The curatives were added in the third stage using a two-roll mill (MAC 350 2V, Novo Hamburgo, Rio Grande do Sul, Brazil) at 60 °C for 6 min.
Properties of rubber compounds and vulcanizates
Mooney Viscosity
The viscosities of the filled compounds were determined using a Mooney viscometer (MV 2000, Alpha Technologies, Akron, OH, USA) at 100 °C, using a large type rotor (L), a preheating time of 1 min, and a moving rotor time of 4 min, according to ASTM Standard D 1646.
Cure Characteristics
The cure characteristics were determined using an MDR rheometer (MDR 2000, Alpha Technologies) at 170 °C with ±0.5° of arc, according to the test procedure described in ASTM Standard D 5289.
Payne Effect
Strain sweep experiments on the uncured compounds were carried out using a Rubber Process Analyzer (RPA 2000, Alpha Technologies) at 100 °C with strain amplitudes ranging from 1.4 to 1255.5% at a constant frequency of 0.100 Hz. Using the same equipment, the samples were vulcanized under no strain at 170 °C for a time equivalent to the optimum cure state (t90) obtained from MDR rheometry measurements. Thereafter, the temperature was reduced and maintained at 100 °C for 6 min before starting the strain sweep from 0.28 to 150% at 0.100 Hz.
Mechanical Properties
The compounds were vulcanized in the form of discs (4 mm thickness × 70 mm diameter), using a hydraulic press (PH C355-4, Copè, Brazil), at optimum cure time (t90) at 170 °C. The mechanical properties of the vulcanizates were measured in an EMIC tensile tester (model DL 2000; EMIC, São José dos Pinhais, Paraná, Brazil) at a crosshead speed of 500 mm/min using ring test specimens prepared according to standard UNI 6065.
Abrasion Resistance
The abrasion resistance of the vulcanizates was determined using a DIN abrasion tester (Maqtest, Franca, Brazil) according to ISO 4649.
Dynamic Mechanical Properties
The dynamic-mechanical properties of the vulcanizates were measured using a DMA/DMTS (Eplexor® 150 N; Netzsch Gabo Instruments, Ahlden, Germany) in elongation mode, using specimens with dimensions of 2 mm × 5 mm × 30 mm. Temperature sweep experiments were run from −80 to +80 °C at a heating rate of 2 °C/min, adopting static and dynamic strains of 2.0 and 0.1%, respectively, at a frequency of 1 Hz. Strain sweep experiments at constant temperatures of 0, 23, and 70 °C were run from 0.05 to 5% strain at 10 Hz, maintaining static strain at 15%.
RESULTS AND DISCUSSION
Epoxidation of the liquid rubbers and preparation of the coated silicas
The two-step epoxidation reaction of the liquid rubbers is shown in Scheme 1. The formation of performic acid occurs in the first step (a), which is slow and endothermic. In sequence, the performic acid reacts quickly and exothermically with the double bonds of the polymer resulting in oxirane rings (b). One proposed mechanism involves the formation of a polar cyclic intermediate via intramolecular transfer of the proton to the carbonyl oxygen of the peracid and the simultaneous attack on the π bonds of the polybutadiene.12–15 As evidenced in Table I, it appears that only the 1,4-isomeric units suffered epoxidation. Previous studies of polybutadiene rubbers with different microstructures show that the reactivity of the 1,4-isomeric units is greater than the 1,2-units.16
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
In a recent work,7 the kinetic behavior of the chemical reaction occurring between ELHPB and silica was studied using differential scanning calorimetry (DSC), and features of an autocatalytic process were observed. The proposed mechanism involves the nucleophilic attack on a carbon of the oxirane ring by a silanol group. The reaction between silica and epoxidized polybutadiene is shown in Scheme 2.
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
Mixing rheology
The mixing of rubber compounds is an energy-intensive process. Hence, alternatives that enable energy savings in the mixing room are always welcome. The evolution of torque as a function of time from the first mixing stage for compounds without a surface modification agent (blank), containing TESPT (reference), and filled with ELHPB–g–silica is shown in Figure 2a. Figure 2b shows the energy required for the first stage of the mixing process as a function of the degree of epoxidation, in the case of ELHPB, and as a function of molar mass of ELPB used as silica surface modifying agents. As shown, the energy consumed in the first stage increases linearly with the degree of epoxidation of the ELHPB used in the silica treatment. The content of the epoxy groups is kept approximately constant (8 × 10−6 mol/m2). This increase cannot be satisfactorily explained solely by the decrease in the content of the liquid polymer because formulations C (with LHPB) and D (with ELHPB08) have the same composition. On the other hand, with the increase in the number of epoxy groups per molecule, it is more likely that a given polymer chain will be linked to the silica surface after precuring. Thus, with an increase in the number of chains attached to the surface, a larger hydrodynamic effect of the filler can be expected during mixing, resulting in greater torque. In summary, if the improvement in polymer–filler interaction leads to an increase in the mixing energy required, the consequent reduction in silica–silica interaction should have the opposite effect; the observed phenomenon should be the result of both contributions.
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
The effect of the molar mass of the ELPB samples on the mixing energy during the first stage is also shown in Figure 2b. As shown in Tables I and II, the degree of epoxidation of the polymers is similar (8 ± 1 mol%), and the proportions between liquid polymer and silica are identical (60 phs). Initially, there is a large increase in energy (16 kJ/g) when the molar mass is increased by 18 kDa (from 8 to 26 kDa), but the following increase of 18 kDa (from 26 to 44 kDa) causes a further increase of only ∼3 kJ/g. One might also think that the increase in hydrodynamic effect with the increase in chain length is responsible for the observed phenomenon because after a certain molar mass the number of entanglements is much greater and can result in additional resistance to the flow. Nevertheless, the amount of energy required during the first step of mixing when a polymer of greater molar mass was used (ELPB44) was about 7% less than the reference compound using TESPT (68 ± 2 kJ/g).
It is evident in Figure 3 that the previous reaction between silica and ELHPB was not complete because, after ∼5 min of processing, the torque again rose, making it necessary to extend the mixing cycle another 4 min. Similar behavior has been observed in previous work4 and was expected because the curing conditions employed (8 min at 170 °C) would not be sufficient to ensure a high degree of conversion during preparation of the coated silicas.7 According to previous kinetic studies7 performed using DSC, the reaction is essentially complete in approximately 16 min at 170 °C, but longer times may be required for the complete curing of large quantities of the sample in a vacuum oven, given the difficulties of thermal diffusion with this type of process. However, preliminary tests showed some thermal degradation of the polymer when the material was subjected to curing times greater than 10 min. It is thought that it is possible to essentially complete the grafting reaction before rubber mixing, if a more suitable process is used, such as one that uses a fluidized zone mixer, for instance.
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
Similar to what has already been demonstrated earlier,4 it is very likely that the time required to complete the reaction between the epoxy groups and silanols in the second stage of mixing can be greatly shortened through process optimization. Such adjustments may include providing additional thermal energy directly or indirectly through process conditions that favor the generation of heat during mixing, such as increasing the shear rate by increasing the rotor speed and/or by changing the chamber fill factor.
Rheological properties of unvulcanized compounds
Strain Sweep Experiments (RPA 2000)
The dependence of the elastic shear modulus (G′) on dynamic strain—a phenomenon known as the Payne effect—is generally used as a measure of filler networking, mainly controlled by the quality of dispersion and the polymer–filler interaction.17 Figure 4 shows the storage modulus G′ as a function of the dynamic strain at 100 °C and 0.1 Hz for unvulcanized compounds filled with silica grafted with ELHPB (a) and ELPB (b).
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
As shown, the compound containing LHPB (Figure 4a) exhibits a high Payne effect (ΔG′ = 0.98 MPa) similar to that of the blank compound, demonstrating that the unmodified liquid polymer does not provide good polymer–filler interaction. Thus, the low mixing energy observed in the first and second stages is merely the result of a plasticizer effect, which also reduces the stability of the filler in the matrix and contributes to an increase in the flocculation rate (reagglomeration of silica). As shown, if the silica is impregnated with ELHPB08 and is not precured and the respective compound is not subject to the second mixing stage, the Payne effect is much less (ΔG′ = 0.43 MPa) than that of the compound using LHPB. This is an indication of the effectiveness of the physical interactions between the epoxy groups and the silanols in preventing filler reagglomeration. However, ΔG′ is still much greater than that observed for the compound that used ELHPB08–g–silica, precured, and subjected to the remill stage (ΔG′ = 0.27 MPa), which reaffirms the importance of these treatments on the stability of aggregates dispersed in the matrix.
Although proving very effective in suppressing filler networking, the performance of ELHPB as a silica surface modifying agent was slightly lower than that of TESPT. The increase in the degree of epoxidation does not seem to significantly affect the Payne effect, probably because the content of epoxy groups in the ELHPB–g–silica compounds was kept constant, resulting in no significant change in the polarity of the polymer coating around filler aggregates. Regarding the surface modification of silica with ELPB, whose results are shown in Figure 4b, there is also a significant reduction in the Payne effect. However, the ability of the coating to prevent filler agglomeration decreases when the molar mass of the polymer is increased. This result was unexpected because the increase in the viscosity of the matrix should help reduce the flocculation rate. Thus, the low mobility of the higher molar mass chains possibly hindered their intrusion into the pores and impaired the proper coating of silica aggregates. On the other hand, ELPB08 was found to be even more efficient than TESPT in reducing the Payne effect. Compared with the compound containing silica treated with ELHPB08, with an identical degree of epoxidation and an even lower molar mass, the possible effect of the polymer microstructure on its performance in surface shielding is evident: The higher chain mobility of ELPB08, resulting from the lower content of vinyl and 1,4-trans units, may facilitate the diffusion of chains throughout the reactive sites and thus favor the reaction of silica graphitization.
Mooney Viscosity
As shown in Figure 5, the viscosity of the final compound does not appear to be significantly affected by the degree of epoxidation of ELHPB. The viscosity is similar to the reference for the compound filled with silica modified with ELPB26, and it increases with the molar mass of the polymer. In the case of the compound containing ELPB08–g–silica, the viscosity is much less than the reference and that of the compound containing ELHPB08–g–silica.
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
Kinetics of vulcanization
When sulfenamide type accelerators (such as CBS) are used in rubber compounds, the general desire is to obtain a high cure rate along with high process safety. Furthermore, changes in process parameters and compound formulations, which are capable of minimizing the optimum cure time (t90) and simultaneously maximizing the incubation time and the maximum cure rate, are always desirable.
It is evident from Figure 6a that the optimum cure times (t90) for all compounds filled with ELHPB– and ELPB–g–silicas are similar to or less than that of the reference compound. Furthermore, Figure 6 shows that the total torque variation, MH-ML (b); the incubation time (c); and the maximum cure rate (d) increase with the degree of epoxidation of the coating polymer, which means that there is a general improvement in curing characteristics. The same, except for the incubation time, is valid for an increase in the molar mass of the ELPB.
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
In a previous study,4 it was observed that the increase in the content of epoxy groups caused decreases with incubation time and MH-ML, effects which were attributed to the possible interaction between epoxy groups and CBS.12 More recently, Sengloyluan et al.18 reported that the use of ENR-51 as a compatibilizer in a silica-filled NR compound resulted in longer scorch and cure times than the analogous silanized compound because of the adsorption of polar curatives by the remaining free silanols and epoxy groups. The opposite effect observed here suggests that the precuring of polymer-coated silica and heat treatment performed in the second mixing stage ensured a high degree of conversion of the epoxy–silanol reaction, significantly reducing the concentration of free epoxide. Corroborating this assumption is the observation that the sample containing ELHPB08–g–silica, for which such heat treatments were not performed (open triangle), showed a much higher t90 and a much lower cure rate.
Additionally, the increase in the degree of epoxidation improves the ELHPB–silica interaction and can result in more efficient surface shielding, preventing the adsorption of curatives, accelerators, and protectives to silica micropores. This could explain the corresponding increases in incubation time, maximum cure rate and, partially, in MH-ML, with an increase in the degree of epoxidation of ELHPB.
As shown in Figure 6, the use of unmodified LHPB adversely affects the vulcanization kinetics of the compound because it causes a reduction in process safety and an increase in curing time. In addition, the additive appears to have affected the crosslink density of the compound because the MH-ML was much lower than observed for the other compounds. Thus, LHPB may have merely acted as a reactive plasticizer, consuming curing agents to form elastically ineffective crosslinking points.
To explain the effects on MH-ML, however, it is also necessary to take into account the process of filler flocculation, which occurs at the beginning of vulcanization19,20 and is faster at higher temperatures.21 Figure 7 shows the curves of the derivative of torque with respect to time for some compounds; the higher the value of the derivative below the 1-min analysis, the higher the flocculation rate. It is observed that the flocculation rates for all compounds containing silica modified with epoxidized polymers were superior to that of the reference, indicating that filler networking may contribute importantly to the higher values of MH-ML seen. However, there is no clear relationship between the maximum value of the derivative below 1 min and the degree of epoxidation or molar mass of the polymer.
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
In general, compounds containing silica modified with ELPB of higher molar mass (26 and 44 kDa) demonstrated kinetic behavior in line with, or better than, that of the reference compound. Moreover, the compound containing ELPB08–g–silica exhibited very low values of MH-ML and a maximum cure rate. Because this sample exhibited a t90 and an incubation time only slightly higher than those of the compounds filled with ELPB26– and ELPB44–g–silica, the lower value found for the maximum cure rate is probably a consequence of the low value of the MH-ML. This sample also exhibited lower MH-ML than the compound containing ELHPB08–g–silica, whose polymer has an identical degree of epoxidation and a molar mass still lower than ELPB08. Such a difference can be attributed to the lower flocculation experienced by the compound filled with ELPB08–g–silica.
Mechanical properties
The results of stress–strain tests for compounds vulcanized at 170 °C are shown in Figures 8 and 9. As would be expected based on the results already observed, the compound containing LHPB exhibited the worst mechanical properties, demonstrating low polymer–filler interaction. In addition, analogous to previously discussed aspects, the precuring of the ELHPB-impregnated silica and the total cure in the second mixing stage proved essential to bettering the mechanical reinforcement as evidenced by increases in tensile strength (TS) and moduli at 100 (M100) and 300% (M300) elongation. There is no well-defined relationship between the degree of epoxidation of ELHPB and M100 or M300: Although M100 varied around the values of the blank and reference compounds, the value of M300 was far below the other. On the other hand, the TS of the compounds filled with ELHPB17– and ELHPB26–g–silica were close to that of the reference compound, whereas the elongation at break (EB) was 30 to 45% higher. Probably, the mechanical properties of these vulcanizates can be improved significantly by sulfur compensation relative to the reference compound.18
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
The improvement in reinforcement with an increase in the degree of epoxidation of ELHPB is more evident in the analysis of the variation in the reinforcement index (M300/M100). In Figure 9, it is observed that the ratio M300/M100 increases proportionally to the degree of epoxidation but remains below the value obtained for the reference compound (4.6), even for the highest degree of modification studied. The low reinforcement value in relation to that of the reference could be explained by a lower interaction at the interface coated–silica/matrix compared with that of the silanized–silica/matrix system because the silane used may establish chemical bonds with the rubber matrix. Moreover, the epoxidized rubber coating presents a reduced availability of unsaturation, leading to a lesser probability of crosslinking at the interface than in the matrix. This hypothesis explains the low reinforcement of the compounds filled with ELHPB–g–silica compared with that of the reference compound, but it cannot explain the improvement in reinforcement when the degree of epoxidation of ELHPB is increased.
The phenomenon may be explained by the formation of a polymer coating consisting of chains with some segments chemically bonded to the silica surface, highly immobilized, and unbound segments forming entanglements with the polymer chains of the matrix. A morphological model for explaining the reinforcement produced by the chemical immobilization of epoxidized polymer chains on the silica surface is shown in Figure 10, being analogous to that proposed by Litvinov et al.22 to explain the reinforcement of EPDM with carbon black. The following types of network junctions could be present in these vulcanizates: chemical bonds between epoxidized segments and silica (A), chemical crosslinks between coating and matrix, and entanglements between coating and matrix (B). The proposed model may explain the increase in reinforcement with a higher degree of epoxidation because the increase in the number of epoxidized units per chain could also result in an increase in the number of loops and a reduction in their heights, increasing the probability of entanglements with the matrix chains and/or reducing their mobility (entanglements would be closer to the silica surface).
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
Regarding the compounds filled with ELPB–g–silica, Figure 8a,b shows strong increases in M100 and M300 with the increasing molar mass of coating polymer. Therefore, however, reductions are observed in EB and TS. The reinforcement index also increases with increases in the molar mass of ELPB but at a lower rate than seen when the degree of epoxidation of ELHPB is increased. When there is no good polymer–filler interaction, the flocculation process can result in the occlusion of rubber in the agglomerates formed, resulting in an increase in the effective volume of filler and, consequently, in M100 because, under such deformation, these associations are not completely broken.23 However, the Payne effect of unvulcanized compounds containing ELPB26– and ELPB44–g–silica was much lower than that of the compound without a coupling agent (blank), but the values of M100 were 60 to 75% higher. Furthermore, at 300% elongation, the effect of interaggregate association should disappear, making the role of the polymer–filler interaction in reinforcement more evident.24 Because the values of M300 for vulcanizates containing ELPB26– and ELPB44–g–silica were also high, it can be concluded that the increase in polymer–filler interaction should have been the main cause of the increase in modulus. When the increase in M100 and M300 is a result of the increase in the crosslink density of the matrix, a reduction in EB can be expected and, above a given value of crosslink density, TS can also decrease. However, when the modulus increases because of reinforcement by particulate fillers (hydrodynamic effect), paradoxically, there is a simultaneous increase in EB.24 Because the increase in the molar mass of the epoxidized polymer caused an increase in the modulus and a simultaneous decrease in TS and EB, one can imagine that, during the preparation of the surface-modified silicas, some chains are long enough to bond chemically to the surface of more than one aggregate to form a network of “giant multifunctional crosslinks.”24 Such strong interconnections would help to maintain reduced interaggregate distances, also explaining the largest Payne effect observed for these samples in comparison with compounds containing silicas grafted with highly epoxidized polymers of low molar mass, for which one would expect greater filler–filler interaction because of the higher polarity of the coating.
Dynamical properties
Strain Sweep Experiments (RPA)
The agglomeration of filler particles at high temperatures (during vulcanization) has an important influence on the dynamic properties of the vulcanizates because filler networking results in nonlinear rheological properties and thus in increased energy dissipation under dynamic conditions of service.17
The results of the strain sweep experiments for vulcanized samples are shown in Figure 11. Before the experiment the samples were subjected to a much higher temperature (170 °C) than that of the experiment in the crude state (100 °C), thus different trends could be expected because flocculation can be facilitated at higher temperatures. However, comparing the graphs in Figures 4 and 11, similar trends are observed. Therefore, the following hypotheses cannot be ruled out: (1) an apparent increase in the Payne effect with the increase in the degree of epoxidation of ELHPB could be a result of the increased polarity of the coating formed around the aggregates, and/or the increased thickness of the rubber shell; (2) the lower mobility of the chain segments of higher molecular weight of ELPB may have hampered their intrusion into silica pores, impairing the proper shielding of aggregates; and (3) the strong interconnections between aggregates, formed by shared chains, could be responsible for maintaining reduced interaggregate distances, increasing the Payne effect.
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
Strain Sweep Experiments (Eplexor)
Strain sweep experiments were also performed in elongational mode at 23 and 70 °C, from 0.05 to 5% of dynamic strain, at a frequency of 10 Hz, and a static strain of 15%. The results of such experiments are shown in Figure 12. In general, the behavior of the compounds containing ELHPB– and ELPB–g–silica at low temperatures are, in theory, superior to the reference compound because the vulcanizates exhibit higher hysteresis at 23 °C and, therefore, should provide better tread adhesion on a wet track surface.25,26 The higher values of E′ exhibited by the compounds containing the polymer-grafted silicas are also indicative of better performance in emergency maneuvers (hard handling) because stiffer tread tends to improve cornering stiffness.27 On the other hand, the results of tanδ at 70 °C, which correlate more closely with the rolling resistance offered by the tread, are superior to that of the reference compound, and so, an inferior behavior with respect to heat generation and fuel consumption can be expected.
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
Such high hysteresis in the silanized compound may be due to a combination of factors: (1) the higher Payne effect, discussed above; (2) energy dissipation because of the immobilized polymer layer on the silica surface; and (3), energy loss associated with the free ends of the oligomeric chains grafted onto the silica surface. Reducing the amount of plasticizer (S-SBR is extended in 37.5 phr TDAE oil) and/or adding extra sulfur18 could partially compensate for these effects.
Temperature Sweep Experiments (Eplexor)
Figures 13 and 14 show the results of the temperature sweep experiments at low strain (±0.1%) for a series of vulcanized compounds containing ELHPB– and ELPB–g–silica, respectively. In the samples filled with ELHPB–g–silica, the modulus in the rubbery plateau increases with the increase in the degree of epoxidation, whereas the maximum tanδ decreases. On the other hand, E′ is higher, and the maximum tanδ is lower for the vulcanizates containing silicas grafted with the polymers of higher molar masses (ELPB26 and ELPB44). The reduction in tanδ and the increase in E′ are indicative of an increase in the effective volume of filler because of the increase in polymer immobilization on the silica surface (the rubber shell).25 In the rubbery plateau of all compounds containing polymer-grafted silica, the temperature dependencies of E′ and tanδ are greater than those observed for the blank and reference compounds.
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
These results can be interpreted in light of the glassy layer model28,29: Because of the reduced mobility of the chain segments bonded to the filler surface, a quasi-vitreous layer of polymer is formed around the aggregates, as suggested by Figure 10. When the aggregates percolate to form a filler network, the interfaces between the clusters will consist of a polymer layer of even less mobility. Thus, Arrhenius-like temperature dependence can be expected for the stiffness and strength of these hinge-like filler–filler bonds, and a straight line of slope Ea/R can be obtained by plotting lnE′(T) against 1/T well above the bulk glass–rubber transition region of the vulcanized compound, where Ea represents activation energy, and R represents the gas constant.
The results on the variation of Ea with the silica surface-treatment type are shown in Table IV. The activation energies were estimated from the slopes of the linear relations obtained within the temperature range of 25 to 50 °C. All compounds exhibited activation energies on the order of ∼10 kJ/mol, which is within the range of van der Waals interactions and is consistent with the values reported in the literature.30,31 All polymer–g–silica–filled compounds exhibited higher Ea values than did the compounds containing untreated (blank) or silanized silica (reference), reflecting better interfacial adhesion of the epoxidized polymer–silica system.32 In the case of the reference compound, strong polymer–filler interaction is achieved via a bifunctional organosilane (TESPT), which means that the silica–matrix coupling consists of individual contact points corresponding to the spatial location of the silane bridges.31 On the other hand, in the case of the polymer-grafted silicas, many chain segments are bonded to the filler surface resulting more energy being required to thermally activate the immobilized polymer layer. This may also explain why the silanized compounds exhibited much lower dynamic moduli, that is, they were dynamically softer. On the other hand, carbon black–filled compounds should be dynamically harder than polymer-grafted silicas because the Ea for such compounds is still higher (∼13 kJ/mol).29
Activation energy does not seem to depend on the degree of modification of ELHPB, keeping the amount of epoxy groups per unit surface of silica constant. Thus, the increasing immobilization of chain segments may be offset by a decrease in coating thickness because the amount of epoxidized polymer decreases with an increase in the degree of epoxidation.
The behavior exhibited by the vulcanizates containing ELPB–g–silica indicates a probable reduction in polymer–filler interaction with an increase in polymer molar mass because the Ea decreased. Moreover, the difference between the Ea values for ELPB08– and ELHPB08–g–silica indicates that the microstructure of the polymer coating may influence the energy required to thermally activate the immobilized polymer layer.
Hybrid compounds
Hybrid compounds were also prepared by blending second-stage masters containing silanized silica and polymer-grafted silicas in varying proportions.As shown in Figures 15–17, many of the properties evaluated exhibited synergistic behavior, and none of the trends approached addictive behavior (along the dashed line). With respect to mechanical properties (Figure 15), it is observed that M300 increases as the proportion of silanized silica is increased. Moreover, both series exhibited a significant increase in tensile strength when the proportion of ELHPB–g–silica was 67%. Figure 16a shows that the Payne effect exhibited by hybrids containing 33% ELHPB–g–silica is less than that of pure compounds (100% silanized silica or polymer-grafted silicas).
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
Citation: Rubber Chemistry and Technology 90, 1; 10.5254/rct.16.83760
In Figure 16b, the DIN abrasion test results for the vulcanizates filled only with polymer-grafted silicas are worse than those for the reference, but they improved as the proportion of silanized silica was increased. However, as already observed by Martin et al.,33 the correlation between the results of the abrasion tests in the laboratory and the on-the-road tire wear was poor and can even show a reversal of ranking, depending on test conditions. The DIN abrasion test was much more severe than normal tire wear on the road.
Some dynamic properties of the hybrid compounds are shown in Figure 17. The results were obtained by running strain sweep experiments at 0.05 to 5% of dynamic strain, a frequency of 10 Hz, and a static strain of 15%. The vulcanizate containing 67% ELHPB13–g–silica exhibited higher storage moduli at 23 and 70 °C and tanδ at 23 °C than do the pure compounds, whereas tanδ at 70 °C is intermediate. Similar behavior was found for the series containing higher epoxidized polymer (ELHPB26), with the difference that the moduli at 23 and 70 °C were intermediate compared with those of the pure compounds.
Therefore, the results reveal that the use of such hybrid reinforcement systems may be advantageous insofar as improvements in dynamic behaviors associated with the properties of traction and hard handling can be expected, minimizing losses in reinforcement, rolling resistance and abrasion in comparison with silanized compounds.
CONCLUSIONS
The energy required to disperse the polymer-grafted silicas was 7 to 44% lower than that required for the silanized compound, depending on the degree of modification, molar mass, and microstructure of the epoxidized liquid polymer used. The Mooney viscosity of the compounds filled with polymer-grafted silicas increases with the increasing molar mass of the epoxidized polymer, but it does not appear to be significantly affected by the degree of epoxidation. Regarding the kinetics of vulcanization, a general improvement was observed compared with that of the silanized compound, particularly for compounds containing silica coated with liquid polymers that are highly epoxidized or have higher molar mass.
Important improvements were observed in the reinforcement and abrasion resistance of vulcanizates with an increasing degree of epoxidation and higher molar mass of the polymer coating. Furthermore, there is evidence that the polymer microstructure can also considerably affect such properties. The dynamic–mechanical results indicate that the hysteresis and stiffness of the vulcanizates filled with polymer-grafted silica should be intermediate with respect to those of the compounds filled with silanized silica and carbon black, potentially creating a good balance between traction properties and hard handling, without greatly increasing rolling resistance compared with the silanized compound. The performance of compounds filled with polymer-grafted silicas was worse than that of vulcanizates filled with silanized silica with respect to abrasion resistance. However, it is believed that both abrasion resistance and hysteresis at 70 °C can be further improved by adjusting formulation, mainly by reducing the amount of TDAE oil and/or adding extra sulfur.18
The production of compounds using a hybrid reinforcement system (silanized silica plus polymer-grafted silica) is also feasible and constitutes an alternative means of optimizing the mechanical and dynamical behavior of vulcanizates.
Number of U.S. patents registered per decade with categories of “skid resistance,” “wear resistance,” and “rolling resistance,” as applied to tires. Source: www.freepatentsonline.com.
Two-step “in situ” epoxidation reaction of the liquid polybutadienes: (a) formation of performic acid, and (b) the addition of oxygen to the double bonds with regeneration of formic acid.
Schematic representation of the reaction between silica and epoxidized polybutadiene.
(a) Evolution of torque as a function of time from the first mixing stage for compounds without a surface modification agent (blank), containing TESPT (reference) and filled with ELHPB–g–silica (coating polymer with different degrees of epoxidation). (b) Energy required in the first stage of the mixing process as a function of the degree of epoxidation of ELHPB (epoxidized liquid hydroxylated polybutadiene) and molar mass of ELPB (normal epoxidized liquid polybutadiene) used as silica-surface modifiers.
Torque curves from the second mixing stage for compounds without a surface modification agent (blank), containing TESPT (reference) and filled with ELHPB–g–silica (coating polymer with different degrees of epoxidation).
Dependence of storage modulus (G′) on the percentage of dynamic strain at 100 °C and 0.1 Hz for unvulcanized compounds filled with (a) ELHPB–g–silica (coating polymer with different degrees of epoxidation) and (b) ELPB–g–silica (coating polymer with different Mn). The open, down-facing triangle series refers to the ELHPB08-impregnated silica, which was not precured, and the respective compound, which was not subjected to the remill mixing stage.
Mooney viscosity of final compounds as a function of the degree of epoxidation of ELHPB and molar mass of ELPB used as silica surface modifying agents.
Effect of the degree of epoxidation of ELHPB and molar mass of ELPB on the rheometric characteristics of the rubber compounds filled with the respective surface-modified silicas. The open triangle refers to the ELHPB08-impregnated silica, which was not precured, and the respective compound, which was not subjected to the remill mixing stage.
Derivative of torque as a function of time for compounds filled with (a) ELHPB–g–silica (coating polymer with different degrees of epoxidation) and (b) ELPB–g–silica (coating polymer with different Mn).
Effect of the degree of epoxidation of ELHPB and molar mass of ELPB on the mechanical properties of the rubber compounds filled with the respective surface-modified silicas. The open triangle refers to the ELHPB08-impregnated silica, which was not precured, and the respective compound, which was not subjected to the remill mixing stage.
Effect of the degree of epoxidation of ELHPB and molar mass of ELPB on the reinforcement index (M300/M100) of the vulcanizates filled with the respective surface-modified silicas.
Morphological model explaining the reinforcement produced by the chemical immobilization of epoxidized polymer chains on the silica surface, where the chemical bonds between epoxidized segments and silica (A) and the entanglements between coating and matrix (B) are marked.
Dependence of storage modulus (G′) on the percentage of dynamic strain at 100 °C and 0.1 Hz for vulcanized compounds filled with (a) ELHPB–g–silica (coating polymer with different degrees of epoxidation) and (b) ELPB–g–silica (coating polymer with different Mn).
Effect of the degree of epoxidation of ELHPB and molar mass of ELPB on the dynamical properties of vulcanizates filled with the respective surface-modified silicas at 23 and 70 °C.
Dependence of E′ (a) and tanδ (b) on temperature for vulcanizates filled with ELHPB–g–silica under ±0.1% of dynamic strain at 1.0 Hz.
Dependence of E′ (a) and tanδ (b) on temperature for vulcanizates filled with ELPB–g–silica under ±0.1% of dynamic strain at 1.0 Hz.
Mechanical properties of the hybrid compounds containing silanized silica and polymer-grafted silicas in varying proportions.
(a) Difference between the storage moduli at 1255.5 and 1.4% of strain (ΔG′) for the unvulcanized compounds, and (b) DIN abrasion results for hybrid compounds containing silanized silica and polymer-grafted silicas in varying proportions.
Dynamical properties of the hybrid compounds containing silanized silica and polymer-grafted silicas in varying proportions.
Contributor Notes