Keywords

1 Introduction

End-of-life tires (ELTs) are a waste stream of high relevance in terms of quantity. Today, only a very small portion is recycled into new tires. The average share of ELT recyclates across all components of a tire is about 3% [1, 2]. In addition to being used to modify bitumen for road construction or thermal recovery, ELTs are granulated and pressed with adhesives to form molded parts. Polyurethane is often used as a binder. Processing is relatively simple, but such materials achieve only moderate mechanical properties. The highest-quality form of recycling appears to be the admixture of finely ground ELT recyclate to rubber compounds. Here, it is often possible to achieve characteristic properties equal to those of primary raw materials. However, this route is rarely used today, especially because a number of interactions occur in the process. The diffusion of chemicals between the recyclate and the rubber compound is the major factor to be mentioned here. This influences the crosslinking reaction and thus both the processing behavior and the crosslinking density of the resulting materials. These systemic interactions are often unpredictable even for experts in the field and pose high hurdles for the utilization of ELT recyclates in the rubber industry. The research question of the work presented here was the qualitative and quantitative influence of ELT recyclates on rubber compounds. In this paper, these effects are presented and quantified using exemplary natural rubber compounds crosslinked with α-sulfur and a sulfenamide accelerator. The intent is to demonstrate the unusual effects to experts in compound development and process engineering and to facilitate targeted introduction of recycling applications.

2 General Influence on Material Behavior

First, an overview of the basic influence of the ELT material on the material behavior of rubber compounds will be given. Different amounts of rubber powder as well as different grain sizes were admixed to a rubber compound and the material behavior was characterized by means of crosslinking curves.

The ELT material was provided by Mülsener Rohstoff- und Handelsgesellschaft MRH (Table 1). It origins from truck tires, produced by ambient (suffix “W”) respectively cryogenic grinding (“K”). The material differs by its particle size (mesh size). The ratio of surface area to mass is calculated based on an idealized spherical shape and a density of 1.15 g/cm3. It is only a rough approximation, because ELT material typically shows fractal surfaces and ambient and cryogenic ground materials differ in their specific surface area, which is usually larger for ambient ground material [3].

Table 1. Overview of the ELT materials used
Full size table

The basic rubber compound into which the ELT material was incorporated is a sulfur-cured, sulfenamide-accelerated natural rubber (Table 2). Soluble α-sulfur was applied for all experiments. Tests with insoluble µ-sulfur did not yield to any significant differences. The mixtures were named according to the mass content wP of ELTs. The preparation of mixtures was carried out with a laboratory kneader Brabender Plasticorder (rotor N50, temperature 50 ℃, rotor speed 50 rpm, filling degree 0.75). The crosslinking was measured approximately 21 h after mixture preparation with a moving die rheometer (MDR, MonTech D-RPA 3000) at a temperature of 150 ℃, a frequency of 1.67 Hz and an amplitude of 0.5° .

Table 2. Recipe of the rubber mixtures. Indication in phr (per hundred rubber)
Full size table

Figure 1, left, shows crosslinking curves at various content of cryogenic ground ELTs (K0204). Three characteristic changes in the rubber compounds result, which are more pronounced with increasing ELT content:

  1. 1.

    The minimum elastic torque S'min increases, thus the viscosity rises.

  2. 2.

    The crosslinking reaction starts earlier without significantly affecting the crosslinking time until the maximum torque S'max.

  3. 3.

    The maximum elastic torque S'max decreases. This is due to a lower elastic shear modulus of the rubber and, above all, a result of a lower crosslink density.

Figure 1, right, shows the influence of different mesh sizes of the ELT recyclates at a mass fraction wP of 40%. There is a tendency for both the minimum (1) and the maximum torque (3) to rise slightly with increasing mesh size.

Fig. 1.
figure 1

Crosslinking behavior as a function of ELT content (K0204, 174 cm2/g) (left) and particle size at an ELT content of 40% (right)

Full size image

3 Increase of Viscosity

As shown by (1) of Fig. 1, left, the viscosity of rubber compounds increases when ELT particles are admixed. This is particularly relevant for the processability with calendering, injection molding or extrusion. A sharp increase in viscosity, for example, often leads to high, pulsating pressure and temperature peaks during extrusion.

In order to exemplarily quantify the influence of ELT material on the flow behavior, the shear rate γ̇ dependent viscosity η was measured. Therefore, different proportions of ELT material were admixed to an unfilled natural rubber (type “first latex crepe” provided by Wagu Gummitechnik GmbH). The ELT material was K0204 according to Table 1 and was admixed in proportions of 10 to 90% to the matrix. The measurement of the true dynamic viscosity η was carried out with a high-pressure capillary viscometer (HPCV) GÖTTFERT Rheograph RG25 at a temperature of 80 ℃. The measurements were adjusted by the Bagley and the Weißenberg-Rabinowitsch correction.

Figure 2 illustrates the influence of ELT powder (K0204, 200–400 µm) on the true dynamic viscosity η. On the left, the viscosity is shown as a function of shear rate for various ELT contents up to 90%. On the right, viscosity is shown as a function of ELT content for selected shear rates, relevant for calendering, extrusion and injection molding. The viscosity rises with increasing recyclate content, which is more pronounced at higher shear rates (the divergence of the 10% curve was confirmed by a double measurement). According to Fig. 2, right, the increase as a function of ELT content is approximately exponential. In practice, the increase in viscosity is frequently already so pronounced at 30% recycled content that problems or productivity reductions can occur during calendering or extrusion. However, the compound retains some flowability up to and including 90% ELTs.

Fig. 2.
figure 2

True dynamic viscosity of ELT/caoutchouc compounds at 80 ℃ as a function of shear rate for variable ELT content wP (left) and ELT content for selected shear rates (right)

Full size image

4 Accelerated Scorch and Reduction of Crosslink Density

4.1 Preliminary Consideration

As shown in (2) and (3) of Fig. 1, left, with increasing ELT content, the maximum torque decreases and the crosslinking reaction starts earlier. Both phenomena are most likely due to a mutual diffusion of chemicals. It can be assumed that these effects do not necessarily occur in every case, but depend on the formulation of the rubber compound and in particular on the crosslinking system used. However, the effects described were observed in varying degrees in many of the author's tests to date.

4.2 Decrease of Maximum Torque Due to Sulfur Migration

Once ELT material is admixed into a rubber compound, sulfur diffuses from the rubber compound into the ELT particles [4]. Basically, a diffusion of sulfur from higher to lower concentration takes place, whereby the sulfur bound in crosslink bridges in the ELT material no longer seems to be relevant at that point. Furthermore, in the case of the presence of different caoutchouc types, the chemicals migrate more quickly to the phase with the lower glass transition temperature. This diffusion leads to a depletion of sulfur in the rubber compound and a corresponding reduction in the crosslink density. As a result, the vulcanizate shows a lower elastic shear modulus, which is reflected in the form of a lower maximum of the elastic torque (S'max).

Compensation of Sulfur Migration.

To counteract sulfur migration, the content of sulfur (S) and accelerator (CBS) in the base rubber compound should be increased [5]. Figure 3 shows the crosslinking and deformation behavior of rubber compounds with 40% ELT material at different levels of S and CBS. When ELT material is admixed to the base compound without any change in its S and CBS content (1), the elastic shear stiffness (maximum torque S'max) drops. This is also associated with a decrease in SHORE hardness from 65 to 57 SHORE A.

In this context, it is occasionally recommended in literature to take the caoutchouc content of the ELT material into account when dimensioning the crosslinking system [6]. The average caoutchouc content of ELTs is about 45%. Therefore, with 100 phr caoutchouc in the base compound and an admixture of 100 phr ELT, S and CBS would have to be increased to 1.45 times. This is the case for mixture (4) of Table 3. For mixture (2), only S was increased accordingly, and for (3), only CBS. For (5), S and A were increased only by half of that, thus to 1.23 times.

Figure 3 shows the influence of increasing the S and CBS content. On the left side crosslinking curves are shown, which were determined analogously to Chapter 2. On the right side, the stiffness (secant modulus) of selected samples is displayed (measured according to DIN 53504, S2, 200 mm/min). It is found that both S'max and SHORE hardness can be raised back to the level of the initial mixture (base) in this way.

Table 3. Recipe of the rubber mixtures. Indication in phr (per hundred rubber)
Full size table

However, looking at mixture (4), the following can be observed: This mixture has a slightly higher SHORE hardness and S'max than (base). According to Fig. 3, right, the material nevertheless shows a lower stiffness in tensile test when compared to mixture (base). In consequence, this implies that even this compensation does not necessarily achieve the identical material behavior of the initial compound. A comprehensive empirical optimization is usually required to bring ELT/rubber compounds into the target window, and it is necessary to focus on one specific parameter, such as SHORE hardness or stiffness. Furthermore, even with identical formulations of the base compound and the recyclates, it must be assumed that, as a result of sulfur migration and the potential additional crosslinking of the recyclates, such compounds might always possess two phases with different stiffness characteristics [7].

Fig. 3.
figure 3

Influence of the sulfur and accelerator content for 40% ELT content on the crosslinking behavior (left) and the secant modulus (DIN 53504, S2, 200 mm/min) (right)

Full size image

4.3 Accelerated Scorch

The induction phase is the time prior to actual crosslinking and is therefore crucial for the processing time at a specific temperature. Before crosslinking begins, the complexes important for sulfur transfer are formed during the induction time (scorch phase). Sulfenamide accelerators react with zinc oxide and sulfur to form an active sulfur-accelerator complex, which determines the duration of the induction phase. This is followed by the transfer of the sulfur to the rubber. In this context, it seems plausible that such sulfur-accelerator complexes diffuse from the ELT material to the rubber matrix and lead to a premature start of crosslinking [7].

Figure 4, left, shows the influence of the amount of rubber powder (K0204) on the accelerated scorch of the mixtures according to Table 2. The right side shows the influence of different particle sizes at a filling ratio wP of 40%. The reduction in induction time occurs immediately after mixing, and there is an additional time-dependent decrease. This time dependent decline can also be observed for the compound (base). However, for the ELT compounds, the ratio of the time t90, required to reach 90% of S'max, over the induction time ti deteriorates significantly (see Fig. 5, right). The crosslinking reaction thus starts earlier, but without significantly influencing the overall crosslinking time. The processing window thus becomes smaller without significantly reducing the total cycle time. An influence of the particle size is present at the beginning (Fig. 4, right), but equalizes after about 100 h. This supports the hypothesis that these are diffusion processes which require a prolonged time for longer diffusion paths or low specific surface area. Finally, Fig. 5, left, shows the induction time 21 h after mixing as a function of wP.

Fig. 4.
figure 4

Induction time at 150 ℃ as a function of storage time after mixing with various ELT content (K0204) (left) and various particle size for an ELT content of 40% (right)

Full size image
Fig. 5.
figure 5

Induction time of rubber compounds at 150 ℃, 21 h after mixing, as a function of ELT content (K0204) (left) and ratio of the time t90 required to reach 90% of S'max over the induction time ti as a function of the time elapsed after mixing (right)

Full size image

4.4 Summary

A number of physical and chemical interactions occur when end-of-life tire recyclates are admixed to sulfur-cured rubber compounds.

In practice, the increase in viscosity often leads to reduced productivity in extrusion. The financial benefits of ELT recyclates, which would be the main consideration in many applications, are therefore often outweighed. Additional flow additives therefore have to be used in many cases, which also has a negative impact on costs.

Sulfur migration has to be counteracted at higher ELT contents by increasing the sulfur and accelerator content in the base compound. However, this results in a considerable increase in the total sulfur content of the elastomers. Furthermore, materials are produced with two phases of potentially different stiffness and with a partially altered deformation behavior, which is increasingly determined by the properties of the tire compounds as the ELT content increases.

If accelerated scorch is problematic in the given manufacturing process, it can be counteracted by lowering the crosslinking temperature. However, this results in increasing cycle times. Even the use of retarders can only compensate for this effect to a very limited extent. Initial preliminary tests showed very limited efficiency and were not considered in this paper.

5 Conclusions

The addition of recyclate to rubber compounds leads to atypical changes in processing characteristics, often surprising even to the expert. Their systemic correlations pose a challenge for material development and involve a considerable amount of work. In order to expand the material recycling of elastomers, it therefore appears appropriate to study the effects described here in depth and to systematize them in order to ultimately facilitate the implementation of recycling in existing rubber compounds and to initiate the development of new compounds that are unaffected by such effects.