Production and Characterization of Recovered Carbon Black (rCB) by Waste Tire Pyrolysis as a Potential Carbon Black (CB) Substitute

3.1. Characterization of Input Material The feed material is rubber granules from scrap tires (a mixture of different car, truck and bicycle tires) that have already been mechanically processed and separated from the steel and fabric fractions. More than 50% of the particles are between 3.15 and 5.0 mm in size. According to [47], this particle size is well suited for obtaining a high combined yield of pyrolysis coke and oil, which is economically relevant for large-scale operation of pyrolysis plants. The detailed sieve analysis, including the particle size distribution curve, can be found in Table S1 and Figure S2. Table 1 summarizes the proximate and elemental analysis of the rubber granules used (as received). It includes the determination of water content, ash content, content of volatile components and content of C, H, N and S. The percentage of oxygen and fixed carbon was determined by calculation. 3.2. Preliminary Study in Laboratory Scale To study the basic degradation of the rubber granules under inert conditions, four TGA curves were recorded at different heating rates (2, 5, 10, 15 K/min). Figure 4 shows the average values for each heating rate. It can be noted that there is almost no mass loss above T > 600 °C. Since the study focuses on rCB, this suggests that the pyrolysis and, thus, the coke formation process is complete. Consistent with previously published experimental results [48,49,50], it is also observed that the degradation reactions shift to higher temperatures as the heating rate increases. The solid pyrolysis residue obtained was 33.84 wt% at a heating rate of 2 K/min, 36.74 wt% at 5 K/min, 32.99 wt% at 10 K/min and 33.79 wt% at 15 K/min. Despite extensive homogenization of the rubber granules, the variations in the solid yields can be attributed to the different compositions of the scrap tires tested, particularly their fillers, and the small amount of TGA sample (few milligrams). DTG curves (Figure 4b) show that mass degradation during scrap tire pyrolysis occurs in a temperature range of 200–600 °C. The bimodal degradation is in good agreement with previously published literature results, although trimodal distributions are also described in some cases, depending on the level of detail [47,48,49,51,52,53,54,55,56]. It can be summarized in two main zones: in the range of about 200–400 °C, mainly oils, plasticizers and natural rubber are de-graded, while in the range of about 400–500 °C, the degradation of synthetic rubber (mainly polybutadiene rubber (BR) and styrene-butadiene rubber (SBR)) takes place. To assess the effect of the pyrolysis temperature on the properties of the focus product, pyrolysis coke, a series of tests were carried out in the laboratory tube furnace (batch operation) over a temperature range of 300–750 °C. The pyrolysis temperature was increased by +50 °C for each batch, and the residue time was set at 60 min. Figure 5a shows the distribution of pyrolysis coke and the sum of oil and gas from the laboratory study where the focus was on pyrolysis coke only and no oil condensation was performed. While 89.38 wt% pyrolysis coke is produced at 300 °C, the yield decreases to 37.73 wt% at 450 °C. Above 450 °C, the coke yield remains almost constant. If scrap tire pyrolysis is operated with a focus on rCB production, however, it is not primarily the quantity but rather the quality of the rCB produced that is important for subsequent reuse as a secondary raw material. A relevant quality feature is the quantity of volatile content. Volatiles are formed during the thermal degradation of the rubber polymers, whereby the organic compounds are adsorbed as impurities on the porous rCB surface [33]. Figure 5b visualizes the quality-reducing contamination of the rCB by volatile components as a function of the pyrolysis temperature for the laboratory study. Obviously, the surface contamination of rCB by volatiles decreases significantly with increasing temperature from 56.03 wt% at 300 °C to 3.44 wt% at 450 °C. Above 450 °C, the reduction in volatiles decreases with a further increase in temperature, with a minimum volatile content of 1.16 wt% at the highest temperature of 750 °C. With respect to the determination of the specific surface area, it should be noted that the rCB samples from the 300 and 350 °C tests were so sticky that they could not be handled in the analyzer. Therefore, the BET surface area could only be determined for all samples produced at a pyrolysis temperature of 400 °C or higher. The surface area distribution of rCB over the pyrolysis temperature (see Figure 5c) shows an interesting profile: up to 500 °C, the BET surface area increases to a maximum of 79.75 m²/g and decreases as the temperature goes up, with the rCB from the 650 °C experiment having a minimum surface area of 59.09 m²/g. Above 650 °C, the BET surface area increases again to 60.39 m²/g at 700 °C and 61.80 m²/g at 750 °C. These values correspond to those of the CB N5xx and N6xx series. In tire manufacturing, these CB grades are mainly used in the lower layers of the tread (base tread) or in the substructure, inner liner and sidewall [57]. Considering the TGA studies (see Figure 4), the observed effect of the BET maximum from 400 to 650 °C can be explained by pyrolysis, which is not yet fully completed. In this temperature range, the material, which is not yet completely degassed, forms a rugged surface structure, which leads to higher BET values. The tendency for the BET surface area to increase from 650 °C can be explained by the reduced formation of carbonaceous deposits on the rCB surface and the lower proportion of volatile components at higher temperatures. These deposits are formed during the pyrolytic decomposition of the long elastomer chains into shorter hydrocarbons, some of which are then deposited on the surface of the rCB and form the carbonaceous deposits (also called “carbonaceous residues”) through polycondensation reactions [58]. This finding is consistent with previously published results [59,60,61]. In general, carbonaceous deposits are undesirable because they negatively affect the surface chemistry and morphology of the rCB. They block or deactivate active surface areas of the rCB [35,62] and smooth the structure of the initially rough surface [63]. This is detrimental to polymer-filler adhesion during reuse in rubber compounds. The interaction between rCB and polymers and thus the use of rCB in new rubber compounds is limited not only by the carbonaceous deposits on the rCB surface, but also by the high ash content compared to conventional industrial CBs [64]. Therefore, the laboratory study also investigated the ash content of rCB as a function of pyrolysis temperature. The results are shown in Figure 5d. Upon completion of pyrolysis, the rCB samples have an ash content >20wt% even when the temperature is further increased to 750 °C. This compares to an ash content of less than 0.5wt% for CBs commonly used in tire production [65]. The high ash content in rCB is due to the inorganic elements, including silicon, zinc and calcium, which act as reinforcing or activating substances for cross-linking in the form of silica, ZnO and CaCO₃ during tire production [62]. Sulfur is also present in rCB, which originally served as a cross-linking agent in the tire compound. The inorganic components in the rubber granules of the scrap tires cannot be removed during the pyrolysis process [66] and therefore remain as impurities in the rCB. Since the exact proportion and composition of the ash depend on the individual formulations of the pyrolyzed scrap tires [64], the ash content in the rCB is somewhat variable. In any case, the ash blocks access to the internal pores, resulting in a low surface area, a reduction in active sites, and consequently impaired interaction in polymer bonding [62]. Evaluation of the initial laboratory test series indicates that a pyrolysis temperature range of 600 to 650 °C is well suited for achieving high coke yield and quality. However, since industrial tire pyrolysis also aims to produce oil as a product, additional experiments were conducted at these two temperatures. In these experiments, the pyrolysis gas was cooled through a condensation unit to recover oil, resulting in a product distribution of 37.52wt% oil and 24.07wt% permanent gas at 600 °C and 37.41wt% oil and 23.90wt% permanent gas at 650 °C. The product distribution from these experiments is visualized in Figure S3. It should be noted that, due to the reactor design, non-isothermal conditions were applied in these tests. Therefore, it is expected that conducting the process under isothermal conditions, where the sample is rapidly heated to the target temperature, would result in slightly higher oil yields. 3.3. Analysis of rCB Produced in Rotary Kiln and Comparison with Other rCBs as Well as Industrial CBs The upscaling experiments to a semi-industrial rotary kiln, based on the laboratory study in Section 3.2, were conducted at 650 °C with a 60 min residue time. With a horizontal inclination of 2°, the kiln rotated at 2.0 rpm and was fed with 8.9 kg/h of rubber granules. In terms of mass balance, the coke yield was 39.11wt%, correlating well with the rCB yield from the laboratory study. Additionally, the oil yield, determined via the bypass condensation unit, was 38.62wt%, resulting in a permanent gas yield of 22.27wt% by difference (see Figure S4). For the gas composition, see Table S3. The measured energy consumption for the kiln operation was 1.3 kWh per kilogram of processed rubber granules, providing a first reference for future environmental investigations such as LCA. All process details for the semi-industrial experiments can be found in Table S2. The analysis of seven industrial CB grades (see Table 2a) serves as a reference for the rCB measurements (see Table 2b). To cover a wide range of carbon blacks, the low reinforcing carbon black N990, three medium reinforcing blacks (N660, N550 and N539) and the higher reinforcing grades N330 and N234 were selected for the investigations. In addition, a high structural CB from the N500 series, designated HS-25, was also tested. As an additional comparison to the rCB1 from the rotary kiln, the measurement results of two other rCB types from industrial waste tire pyrolysis (rCB2 and rCB3) and the experimental materials GP and BG are summarized in Table 2b. It should be noted that the milling process was performed outside of the project study; therefore, the milling type may differ for the rCB types. This has a correspondingly large influence on the particle size distribution obtained and the corresponding specific surface area. The analysis results of the different materials show that all CBs have a high carbon content and low nitrogen, sulfur and hydrogen content compared to rCB. The ash content of the industrial CBs is correspondingly low (<1.0 wt%). In contrast, rCBs have a carbon content of only 57.7–78.7 wt% and a higher content of hydrogen, nitrogen and sulfur. The ash content is also significantly higher in all three rCB samples (between 15.4–21.1 wt%). The reasonable idea of comparing the ash-free pyrolysis coke with the composition of CB shows that the differences are significantly smaller. Although the increased ash content of rCB is not solely responsible for the differences in quality compared to CB, initial approaches to remove the ash [67,68,69,70] appear to be very useful. The importance of this methodology will be discussed again in conclusion. With 33.2 wt% and 34.5 wt%, the experimental materials GP and BG show an even higher ash content than the rCB samples from the scrap tire pyrolysis. Because the GP material is an experimental process development product, its ash content mainly contains iron and copper from the burnt gas pyrolysis electrodes. As far as possible, the characteristics of the GP sample were determined. To protect the analyzer, elemental analysis was not performed on the GP material because of the high content of aromatic compounds and sulfur (yellow discoloration of the sample vessel and strong acrid smell). TGA was used to determine the volatile content at 31.8 wt%. Due to the process-related high ash content and high surface contamination, this batch’s GP sample material is unsuitable as an alternative filler in elastomer products. The same applies to the BG material sample. The mineral content of the rCBs, which severely limits the reinforcing potential compared to the industrial CBs for use in elastomers, was further investigated by SEM and elemental analysis (EDX). Silicon, zinc, sulfur and calcium, as well as small amounts of aluminum, magnesium and potassium, were detected in the three rCB samples (see Figure S5). Their origin, as described above, is due to individual original tire compound formulations. The other metallic components detected are likely to have entered the pyrolysis coke via the scrap tire fractions through contamination or abrasion. Zinc and small amounts of silicon were detected in the BG material sample (see Figure S5). The origin of these trace elements is probably due to the soils from which the sugarcane absorbs them during plant growth. Overall, the EDX analysis of the ashes showed that the mineral content of the rCB is derived from the scrap tire formulations, as expected. However, this analysis also contributes to a better understanding of the ash composition in order to develop separation processes to optimize rCB quality. The carbon black samples were also examined microscopically by SEM at 5000× and 10,000× magnifications (see Figure 6). The spherical primary particles, aggregates and agglomerates of N660 are clearly visible. There is little difference between CB N660 and rCB2 at 5000× magnification. However, coarser particles can be seen in the background of rCB1. At a magnification of 10,000 a corresponding network of “sintered” structures can be seen on the particles. Images of rCB3 are very similar and are not shown here for simplicity. Similar to the rCB sample, the GP material has a coarser appearance than the CB sample and has a platelet-like structure with smaller particles deposited on top. In this comparison, the BG sample shows the largest particles and is clearly distinguishable in appearance from the other materials. In addition to the distinct size difference, the BG surface is relatively smooth and coherent. At 5000× magnification, spherical structures can be seen below the surface, and at 10,000× magnification, individual smaller particles can be identified. As described in Section 2.3, the surface activity of the carbon blacks is determined by BET analysis. When used in an elastomer compound, the specific surface area influences the filler-filler and filler-polymer interactions. This in turn, influences, for example, the tensile strength, elongation at break and abrasion resistance of the resulting elastomer products [71]. The BET surface area of rCB2 (69 m²/g) is comparable to the highly reinforcing CB N330 (76 m²/g) or the N300 series (70–99 m²/g) and rCB1 (65 m²/g) and rCB3 (57.5 m²/g) are similar to the highly reinforcing N400 series (50–69 m²/g). No reliable statement can be made for the BG sample using the BET method. A rough qualitative classification of the BET surface area of the rCBs compared to ASTM D1765-23b standardized CBs can be found in Figure S6. The carbon black structure, which provides information about the arrangement of the primary particles and their aggregate formation, has a significant influence on the dispersion and the stress at 300% elongation when used in new elastomers [71]. In this respect, the rCB samples show values comparable to a medium reinforcing CB N660 with a slight tendency towards a CB N550. The OAN values of the rCBs (rCB1 = 85.5 cm³/100 g; rCB2 = 81.4 cm³/100 g; rCB3 = 92.7 cm³/100 g) are comparable to the OAN of the virgin CB grade N660 with 85.5 cm³/100 g. For the COAN, the values of the rCBs (rCB1 = 76 cm³/100 g; rCB2 = 79.1 cm³/100 g; rCB3 = 81.7 cm³/100 g) are between those of CB N660 with 73.3 cm³/100 g and CB N550 with 86.6 cm³/100 g. This is also true for the VV measurement, which shows the same tendency. Since the torques required by the relevant standards were not reached, the OAN and COAN curves could not be recorded for GB and BG. In this case, the analysis method would have to be adapted. However, the measurements would then be out of the standard and no longer comparable with the data already available. Alternatively, the VV can be specified for these materials. With a VV of 18.3 cm³/100 g for the GP sample and 34.0 cm³/100 g for the BG sample, these materials are more comparable to the less reinforcing carbon black N990. Particle size analysis shows that rCB2 and rCB3 as well as GP and BG, are significantly larger than the CBs. The D97 particle size of the recycled materials is between 11.3–13.6 µm, while that of the industrial CBs is between 1.7–5.2 µm. Since the recycled materials were all provided by industrial partners, there is no precise information on the type of milling. The exception is rCB1, which was processed in a laboratory jet mill. However, experience has shown that milling technology significantly influences particle size and structure and should, therefore, be considered separately in further investigations. Overall, the analyses indicate that the rCB samples have the potential to replace medium reinforcing CBs of the N5xx and 6xx series. This suggests that they could be used in various elastomer applications such as technical rubber goods (e.g., seals, profiles, vibration technology) or in tire production, particularly in the sidewall and inner liner. Future studies should explore this potential in more detail. In subsequent experiments, the rCBs will be incorporated into butyl and nitrile rubber-based elastomer compounds to evaluate their actual in-rubber properties and, thus, their actual potential as a replacement for conventional CBs.