Characterization of waste tire rubber devulcanized in twin-screw extruder with thermoplastics

Abstract

Waste tire rubber (WTR) supplied by a truck tire retreader were processed in an intermeshing co-rotating twin-screw extruder (ICTSE). The extrusion process evaluated the efficiency of the thermomechanical recycling in the devulcanization of WTR rubbers. Samples were prepared by varying the process parameters, the particles sizes and thermoplastics, and the latter was used as devulcanization auxiliary agents. After extrusion, samples were subjected to solvent extraction to determine the soluble fraction (SF). Subsequently, these SF were characterized by Fourier transform infrared (FTIR) spectroscopy, thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC). The highest SF (29 wt%) was obtained with higher screw rotating speed and with smaller particle size. Higher SF indicated a higher degree of devulcanization. The FTIR and DSC analyses showed that natural rubber was the main rubber extracted from extruded samples. In addition, polypropylene was more effective than low-density polyethylene in the devulcanization process, promoting higher SF.

Keywords

Devulcanization twin-screw extruder recycling tires natural rubber

Introduction

The twentieth century was marked by tires manufacture, whose popularity around the world expanded the consumption of its main tread constituent, the natural rubber (NR) or 1,4-cis-polyisoprene.^1,2 The tires tread should have great wear resistance, good tensile properties, low rolling resistance, durability and low noise properties. In addition to NR, the tread may contain polybutadiene (BR) and styrene-butadiene rubber (SBR).³ North America and Europe are the major world tire consumers, importing almost 100% of their NR domestic need. Thus, associating high demand for NR with the recycling techniques would be an interesting alternative to solve the disposal problem of post-consumer tires.

There are some technologies for recycling vulcanized rubbers. Chemical processes using specific reagents with high temperature and pressure is one of them; however, this route is slow and creates environmental problems.⁴ The pyrolysis of vulcanized rubber is a thermal decomposition process, which occurs in reducing atmosphere and high temperature creating carbon black (CB), oil and gas for the petrochemical industry and other applications.^4,5 There are continuous techniques such as microwave⁶ and ultrasound,⁴ which are environmentally friendly processes, but still have low efficiency. The rubber recycling through devulcanization is the conversion of a three-dimensional into a two-dimensional structure, by selective cleavage of crosslinks through physical or chemical energy supply. Recycling through devulcanization would reduce the pressure on the latex production and the occupied volume in landfills, impacting on lower production costs and less environmental pollution.

A devulcanization technique recently researched is the continuous thermomechanical devulcanization through twin-screw extruder. The principle of this method is the shear stress application on the vulcanized rubber during extrusion, which may result in selective break of crosslinks due to the strength and directionality of employed tension.^7

-11 The crosslinks break occurs when the shear stress reaches a certain level in which the shear force overcomes the energy force of the sulphur chemical bonds C–S–, –S–S– and –S–Sx–S–. However, the breaking of main chain bonds (C–C) may occur, being an undesirable effect which cause degradation and consequent loss of properties.

The energy sources of the extrusion process are heat and high shear. The heat is generated by an external source (heating resistances) and by adiabatic heating of the polymer. However, there was a limitation related to polymer degradation temperature. The shear stress is responsible for the tensioning and disentanglement of the polymeric chains, stretching the main bonds and crosslinks. These energy sources are also important to reduce the viscosity and mix polymers with additives such as fillers, reinforcements and blends. In researches involving devulcanization, it is common to add additives that help the selective rupture of crosslinks. These additives or devulcanization auxiliary agents (DAAs) can be antioxidants, as calcium stearate, thermoplastics or miscellaneous additives.^12
-14 Its use is reported in several devulcanization studies; however, this issue was not extensively analysed considering thermoplastics as process auxiliary.

The aim of this study was to evaluate the devulcanization process of waste tire rubber (WTR), obtained from truck tires tread, through extrusion process in an ICTSE. In addition, the influence of process parameters, compositions and additives incorporated during process were analysed, including promising use of thermoplastics as devulcanization auxiliary.

Experimental

Materials

WTR was supplied by BrumaPneus (Brazil) as shavings of retreaded truck tires, containing unknown and varied composition. WTR varied greatly in size, with length ranging from 1 cm to 3 cm and width and thickness between 0.1 cm and 0.5 cm. The shavings were washed to remove impurities, milled cryogenically in a knife mill and granulometrically separated to obtain a homogeneous mixture.

Two polypropylenes (PPs), one supplied by SABIC, grade PP5005 P with MFI = 2.00 g 10 min⁻¹, and the other commercially known as PP H201 from Braskem (Brazil), with MFI = 20.00 g 10 min⁻¹, were added together with the WTR during extrusion. Moreover, a Braskem low-density polyethylene (LDPE), known as EB 853 with MFI = 2.70 g 10 min⁻¹, were extruded with WTR instead of PP. The thermoplastics should act as swelling agents or DAAs. They increase the chain breakage selectivity and work as a fluid, transferring shear tension to WTR particles during extrusion.

The antioxidant recipe was a mixture composed of the elements Pentaerythritol tetrakis(3-(3,5-di-tert-butyl-4-hydroxyphenyl)propionate) and tris(2,4-di-tert-butylphenyl)phosphite, commercialized, respectively, as Irganox 1010® and Irgafos 168® and manufactured by BASF (Germany).

Processing of WTR in a twin-screw extruder

WTR were extruded in an intermeshing co-rotating twin-screw extruder (ICTSE), Coperion, diameter of 35 mm and L/D = 44, and this profile is shown in Figure 1. It was investigated the effects of process parameters, such as screw rotating speed, temperature profile and feeding rate, of the thermoplastics and antioxidants (75 wt% Irganox 1010; 25 wt% Irgafos 168) and of composite granulometry, on the WTR devulcanization.

The design of experiment (DOE) utilized is shown in Table 1. The DOE is a “once a factor at a time (OFAT)” method, where sample WTR02 was defined as reference. OFAT was utilized as the experiment possess a considerable number of parameters with unpredicted responses. Primary parameters varied in OFAT were extrusion temperature, screw rotation speed, thermoplastics and granulometry, whereas secondary parameters were feed rate and antioxidant concentration. For posterior analyses, neat thermoplastics were extruded in the same conditions that WTR02 sample.

Table 1.

Design of experiment adopted to analyse WTR devulcanization in an ICTSE.^a

Samples	Temperature profile (°C)	Screw speed (rpm)	Feeding rate (kg h⁻¹)	Thermoplastic (wt%)	Antioxidant (phr)	Particle Size (mesh)
WTR01	270	150	10.0	15.0 (PP-MFI = 2.00)	0.50	6–30
WTR02	270	350	10.0	15.0 (PP-MFI = 2.00)	0.50	6–30
WTR03	270	550	10.0	15.0 (PP-MFI = 2.00)	0.50	6–30
WTR04	270	350	5.0	15.0 (PP-MFI = 2.00)	0.50	6–30
WTR05	220	350	10.0	15.0 (PP-MFI = 2.00)	0.50	6–30
WTR06	270	350	10.0	25.0 (PP-MFI = 2.00)	0.50	6–30
WTR07	220	350	10.0	15.0 (LDPE-MFI = 2.7)	0.50	6–30
WTR08	270	350	10.0	15.0 (PP-MFI = 20.00)	0.50	6–30
WTR09	220	350	10.0	15.0 (PP-MFI = 2.00)	1.00	6–30
WTR10	270	350	10.0	15.0 (PP-MFI = 2.00)	0.50	14–30
WTR11	270	350	10.0	15.0 (PP-MFI = 2.00)	0.50	6–14

WTR: waste tire rubber; ICTSE: intermeshing co-rotating twin-screw extruder; MFI: melt flow index (g/10 min); PP: polypropylene; LDPE: low-density polyethylene.

^a Extrusion die temperature and the temperature of the last six zones of an extruder with 10 zones.

Soluble fraction by solvent extraction

To evaluate the efficiency of WTR devulcanization by ICTSE, soluble fractions (SFs) of each sample (Table 1) were obtained by solvent extraction, using xylene. The solubility parameters of xylene and rubbers, as NR, are similar, then, xylene was adopted as the extraction solvent.¹⁵ First, particles of extruded samples with size between 20 and 30 mesh were dried at 70°C for 4 h. Approximately, 700 mg of these particles were inserted into a steel wire envelope with 120 mesh of aperture. The envelope was sealed, to prevent loss of solid particles, and weighted. Each sample was tested in triplicate. The envelopes were immersed in a beaker with 200 ml of xylene and extracted at 70°C for 8 h. Afterwards, the envelopes were washed in xylene and dried at 70°C for 4 h. After complete xylene evaporation, the envelopes were weighted again. The SF is calculated as the ratio between the weight difference before and after immersion in xylene and the sample weight before immersion. Furthermore, all the solvent inside the beaker was evaporated, leaving the SF as precipitate. Then, the initial (beaker) and final (beaker + SF) weigh of each beaker were used for verification of the SF calculated.

Characterizations of WTR and SF

WTR before extrusion and their SF were analysed by thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC). TGA evaluated the thermal stability of the materials and the final residues. It was performed in a TA Instruments model Q500 using performed in a Mettler Toledo calorimeter, model 822e, showed and Thermo Nicolet infrared equipment, model Nexus 4700, heating times of 10°C min⁻¹, from 25°C min⁻¹ to 600°C min⁻¹. It was used nitrogen atmosphere until 500°C and an oxidant atmosphere with O₂ was adopted after this temperature. DSC analysis, performed in Mettler Toledo model 822e instrument, showed the transitions temperatures of the main polymers. This test used hermetically sealed aluminium pans, nitrogen atmosphere with 50 ml min⁻¹ flow rate and heating rate of 20°C min⁻¹. Ash content of the WTR shavings before extrusion was determined according to ASTM D4574. The main radicals and chemical bonds of the SF before and after extrusion were identified by Fourier transform infrared (FTIR) spectroscopy. It was carried out in a Thermo Nicolet equipment, model Nexus 4700, with 32 scans ranging from 4000 cm⁻¹ to 400 cm⁻¹.

Results and discussion

Characterization of WTR shavings before extrusion

WTR shavings contained different kinds of rubbers, once they come from several suppliers. Also, truck tires are harder than ordinary tires and therefore they can have NR, BR and SBR in their compositions. The characterization of WTR shavings evaluated the presence of polymers and filler content in the composite.

Figure 2 shows FTIR spectrum of the SF from WTR shavings used to identify the possible rubbers and additives of truck tires. Bands 1496, 1376 and 699 cm⁻¹ are related to SBR, BR and NR, respectively; however, the other bands do not have an absorption pattern that can be directly related to these three rubbers. This was expected once the SF from WTR before extrusion should not present significant portions of thermoplastic rubber and/or devulcanized rubber. The WTR shavings had different types of fillers and additives and were submitted to many types of environments. Thus, rubber degradation due to the previous thermal history, mechanical stress and aging during usage time possible occurred.

The FTIR spectrum of the NR presents characteristics bands such as 2962, 2928 and 2855 cm⁻¹, which are due to C–H asymmetric stretching of –CH₃, of –CH₂– and to –C–H symmetric stretching of the same chemical groups. However, these groups are also present in another polymers and rubbers, not being decisive to identify NR. According to ASTM D3677, bands 1665, 1370, 885 and 833 cm⁻¹ belong to NR spectrum. Bands around 1664, 1450, 1378 and 838 cm⁻¹ can also identify NR.¹⁶ Wanga et al.¹⁷ associated the band 1662 cm⁻¹ to a protein presence in NR, band 1447 cm⁻¹ to vibration of –CH2– and bands 1375 and 839 cm⁻¹¹ to vibrations of methyl radical and hydrogen, respectively, both linked to carbon in the NR double bond. According to Le et al.,¹⁸ bands in 1376 and 888 cm⁻¹ can identify NR, and high absorption intensity at 1376 cm⁻¹ is created by methyl from 1,4-cis-polyisoprene (IR). Band 888 cm⁻¹ is assigned to out of plane bonds vibration from CH₂ in the IR. However, this band has low intensity, making the NR identification impractical.

Truck tires can be made of blends of NR with SBR and BR, which have characteristic bands at 990, 962, 909 and 741 cm⁻¹.^19,20 According to ASTM D3677, bands from 980 cm⁻¹ to 965 cm⁻¹ are assigned to vibration of C–H link from carbons of butadiene double bond. The characteristic bands of a SBR spectrum are 1490, 962, 909, 758 and 699 cm⁻¹. Thus, FTIR analysis of the SF from WTR shavings indicated the presence of NR, BR and SBR. However, it is not possible to affirm certainly this statement, once these rubbers are still vulcanized and, therefore, they have not been solubilized. Additionally, presence of low molecular weight additives such as waxes, paraffin or oils found in tires formulations and soluble in xylene could mask the results.

Figure 3 presents the evaluation by TGAs, of the thermal stability and composition of WTR before extrusion. In Figure 3, there were several weight reductions over the tested temperature range. First, a reduction of 5.4 wt% at 310°C occurred due to decomposition of stearic acid, antioxidants, accelerators and oils.²¹ Subsequently, there were four different weight losses between 310°C and 545°C. The reductions of 25.5, 16.8 and 10.3 wt% were assigned to rubbers that make up the tires tread band. From 310°C to 415°C occurred a weight loss of 25.5 wt% that may be related to NR. Weight reduction of 16.8% between 415°C and 460°C could represent BR degradation.²¹ At 500°C occurred a reduction of 10.3 wt%, which may be related to the formation of benzene and styrene, whose formation is favoured in this temperature range. The weight loss of 25 wt% between 500°C and 545°C, when the inert atmosphere of N₂ was replaced by an oxidative atmosphere, is usually attributed to the oxidation of carbon black. Automotive tire tread typically have 20–35 wt% of carbon black for 60 wt% and 65 wt% of rubber.²⁰ Over 650°C, there is a residue of 17% assigned to inorganic components or ashes. As TG just analyses 100 mg of sample, it cannot represent the entire sampling of WTR. Ash content analysis was applied using 5.00 g of material for each one of the three replicates and showed a residue of 3.5 wt%.

Figure 4 shows the TGA of SF from WTR before extrusion, where a behaviour similar to that from WTR shavings (Figure 3) can be observed.

According to Figure 4, there was no significant weight loss until 315°C. Between 315°C and 420°C, there was weight reduction of 35%, with DTG peak at 383°C. The NR literature reports the beginning of structural degradation at 300°C. There was an continuum weight decrease until 475°C, with reductions of 16 wt%. An intense reaction occurred from 475°C to 500°C, with maximum decomposition rate at 489°C, resulting in a weight loss of 30%. Between 500°C and 575°C almost 11 wt% was lost. The atmosphere transiton from nitrogen to oxygen happened at 550°C, causing CB thermal oxidation. According to literature, the convertion of NR in smaller hydrocarbons, especially aromatic compounds occur in this temperature range.²⁰

The physical and chemical changes occuring during heating of a material can be analysed by DSC. In rubber studies, DSC is useful to determine the glass transition temperature (T_g), where the mechanical behaviour changes from vitreous to rubbery.

Figures 5 and 6 show DSC curves from WTR shavings before extrusion and from this WTR SF, respectively. WTR shavings present an inflexion at −59.8°C while SF had an inflexion at −50.8°C. Several studies show the T_g of NR between −72°C and −49°C.^22
-24 According to Fainleib et al.,²⁵ the degradation of NR can create complex structures containing epoxy rings and polar end groups like –OH. These hydroxyls can form hydrogen bonds, decreasing the mobility of olygomeric/polymeric chains and, by consequence, increasing the T_g of the SF. Figure 6 shows an endothermic peak around 100°C that could be the melting of the ethylene vinyl acetate (EVA), which is usually added in tires composites.

SF of extruded WTR through solvent extraction

WTR before extrusion, neat PP and LDPE were submitted to xylene extraction and results are shown in Table 2.

Table 2.

SFs of WTR before extrusion and of the neat thermoplastics after extrusion.

Sample	SF (wt%)
WTR before extrusion	11.08
Extruded PP-MFI = 2.00 g 10 min⁻¹	0.56
Extruded PP-MFI = 20.0 g 10 min⁻¹	2.10
Extruded LDPE-MFI = 2.70 g 10 min⁻¹	97.84

SF: soluble fraction; WTR: waste tire rubber; PP: polypropylene; LDPE: low-density polyethylene.

WTR has SF of 11% that is attributed to soluble materials as oils, paraffin, flow aids, thermoplastics and thermoplastic/degraded rubbers. PP with higher MFI is more soluble due to a broader molecular weight distribution and lower molecular weight. LDPE has a higher SF (97.84 wt%) than PP once its melting point is lower than the melting point of PP. Therefore, at extraction temperature (70°C), LDPE chains have higher mobility than PP chains, favouring the solubilization of small crystallites.

Table 3 shows the total and relative the SF from extruded WTR. Total SF considers the devulcanized rubber released during extrusion, the soluble portion already present in the WTR before extrusion and the thermoplastics added during extrusion. The SF of the WTR polymeric network generated on extrusion, named ‘Relative SF’, was estimated by subtracting the SF before extrusion and the fraction regarding the thermoplastic, all shown in Table 2, from the total SF. Relative SF was graphically represented in Figure 7. The influence of the screw rotating speed is observed by comparing the SF of samples WTR01, WTR02 and WTR03. The SF increased with screw rotating speed indicating that a higher shear rate provided by extrusion significantly contributed to break the chemical bonds C–S, S–S and possibly C–C from the main chain. The increase of screw rotating speed caused higher viscous heating and shear stress, contributing to break chemical bonds.

Table 3.

SF of the samples of WTR.

Samples	Total SF^a (wt%)	WTR SF^b (wt%)	Thermoplastic SF (wt%)	Relative SF (wt%)
WTR01	21.32	9.42	0.084	11.82
WTR 02	25.27	9.42	0.084	15.77
WTR 03	29.38	9.42	0.084	19.88
WTR 04	26.74	9.42	0.084	17.24
WTR 05	26.15	9.42	0.084	16.65
WTR06	21.85	8.31	0.140	13.40
WTR07	27.75	9.42	14.67	3.65
WTR08	25.47	9.42	0.315	15.74
WTR09	24.90	9.42	0.084	15.40
WTR10	28.25	9.42	0.084	18.75
WTR11	25.81	9.42	0.084	16.31

SF: soluble fraction; WTR: waste tire rubber.

^a After extrusion.

^b Before extrusion.

In Figure 7, the process response to variability in the dimensions of feed particles can be observed, because the difference between the samples WTR10 (14-30 mesh) and WTR11 (6-14 mesh) regard only this factor. The SF of WTR10 samples, extruded with grains of smaller granulometry, was bigger than the SF from sample WTR11, which has larger particles. The higher superficial area of particles from sample WTR10 can favour the abrasion contact between them during extrusion, promoting better chain scission.

The effect of feeding rate on the devulcanization process by extrusion can be seen by comparing WTR02 (10 kg h⁻¹) and WTR04 (5 kg h⁻¹). The sample with lower feeding rate had higher SF than sample obtained with higher feeding rate. Lower feeding rates cause a longer residence time inside the extruder and, by consequence, the rubber receive more heat and shear, contributing to devulcanization.²⁶

The influence of the type of thermoplastic adopted as DAA can be analysed by comparing SF from samples WTR02 (15 wt% of PP with MFI = 2.00 g 10 min⁻¹) and WTR07 (15 wt% of LDPE with MFI = 2.70 g 10 min⁻¹). Samples extruded with LDPE have lower SF than samples processed with PP.

Characterization of SF from extruded WTR

Figure 8 shows FTIR spectra of some extruded WTR SFs compared to the spectra of the SF from WTR before extrusion. The bands occurring in WTR before extrusion frequency range are intensified in the spectra of SF from extruded WTR. In addition, new bands are present in the spectra of SF from extruded WTR. Although the WTR formulations were unknown and heterogenic, a pattern in the spectra of FS from extruded WTR was observed.

According to Agostini et al.,²⁷ bands at 1375 and 888 cm⁻¹ can identify NR. The intensity increase of band at 1375 cm⁻¹ for WTR extruded is due to the absorption of methyl radical in the 1,4-cis-polyisoprene (IR). The absorption at 888 cm⁻¹ is caused by out-of-plane vibration of CH2 from C=CH2 of 3, 4 units of IR.

Bands observed in the WTR due to the extrusion were 1665, 1240, 1129, 1082, 888, 838, 745, 699 and 573 cm⁻¹. The bands 1665, 1370, 885 and 833 cm⁻¹ in the extruded WTR are related to absorption bands from NR, according to ASTM D3677. Bands at 1315 and 870 cm⁻¹ are also assigned to chemical bonds present in the NR. Bands 1664, 1450, 1378 and 838 cm⁻¹ can also identify the presence of NR.¹⁶ Bands at 1375 and 839 cm⁻¹ are due to methyl and hydrogen vibrations, respectively, which are attached to the doubled bonded carbon from NR.¹⁷ Among the bands observed in the SF from extruded WTR, just band 969 cm⁻¹ can be assigned to butadiene, when BR is present. The SBR can be identified by band 699 cm⁻¹, related to the absorption of the aromatic ring from SBR structure.²⁸

There were no bands related to disulphide bond (S–S), usually found between 500 and 400 cm⁻¹. Bands related to sulphur absorption, such as 1415 and 1055 cm⁻¹ also were not identified. FTIR analysis indicated the presence of three rubbers in the extruded WTR.²⁸

TGA evaluated the thermal stability of SF from extruded WTR, as can be shown in Figure 9. The thermal behaviour of WTR shavings before extrusions was added for comparison. Figure 9 shows that after 250°C, the samples begin to show a different thermal behaviour and seem evident that WTR shavings before extrusion present a higher thermal stability. At 400°C, the SF of extruded WTR lost approximately 65% from its initial weight, while WTR shavings before extrusion lost just 25%. This was expected because WTR shavings are formulations that contain all ingredients of tires for cargo trucks, the rubbers are vulcanized and therefore have higher thermal stability than the thermoplastic rubber SF of extruded WTR.

The thermal behaviour of all SF from extruded WTR was similar. Just sample WTR07, which contained 15 wt% of LDPE, had a different decomposition behaviour. As showed before, LDPE was almost completely soluble in xylene; therefore, SF from WTR07 is very rich in LDPE. As this thermoplastic has higher thermal stability than devulcanized rubbers (thermoplastic), it could endure high temperature for a longer time. This enhances that the possibility of the LDPE demonstrates less efficiency than PP in the devulcanization of the rubbers in the present WTR. With respect to SF from others extruded WTR, the thermal behaviours exhibited good similarity, except as WTR03, extruded at 550 rpm, and WTR10, extruded with smaller particle size, whose showed higher thermal stability at 450°C. These two samples were those that showed higher FS and may present the greatest amount of rubber.

Figure 10 shows DSC curves of SF from extruded WTR and WTR shavings before extrusion. These curves represent the second heating after a temperature controlled first heating and cooling. The range between −72°C and −49°C shows the T_g of NR.²³ Most curves shown a pronounced inflexion, except samples of WTR before extrusion and of WTR07 SF, which contain 15 wt% of LDPE instead of PP. Samples with a more pronounced T_g, around −60°C, indicate a higher concentration of devulcanized NR (thermoplastic). DSC curve of WTR07 presents a melting peak at 108°C attributed to LDPE, whereas other samples show a melting peak at 157°C that corresponds to PP. All DSC curves were built in the same scale, so the more intense endothermic heat flow of WTR07 indicates the presence of a higher amount of LDPE in this sample than PP in the others. This could indicate that PP was more effective in the devulcanization of rubbery network. Finally, there is a low intensity melting peak at 95°C which can indicate other kinds of thermoplastics as LDPE or EVA, which could be present in the WTR original formulation.

Therefore, despite the unknown sources of the WTR shavings, after the cleaning, homogenization and extrusion process, they presented a similar behaviour, comparing the SF. This occurred by analysing only 2.00 g of each extruded sample, with three specimens of 0.650 g of WTR shavings from trucks, which were used for a long time and, therefore, were possibly hardened, increasing the crosslink density. After extrusion, it was observed that the WTR particles were not completely softened but joined to each other by surface. This indicates that devulcanization occurred on the particles surface; however, there is a possibility of these particles are again superficially revulcanized with sulphur. Thus, very heterogeneous WTR can be recycled, turning them into new products for various applications.

Conclusions

Important aspects about the WTR composition were possible to be understood due to the carried characterization. Thermogravimetric analysis of the WTR shavings presented four significant weight losses assigned to the presence of rubbers, thermoplastics, additives (antioxidant and oil) and carbon black. DSC curve showed one inflexion at −60°C due to NR glass transition. Another less pronounced inflexion occurred at 90°C, which can represent the melting point or even the T_g of some thermoplastics present in the WTR formulation. SF of the WTR shavings extracted in xylene was approximately 11 wt%. The FTIR spectra of this SF showed the presence of NR among others rubbers as SBR and BR. DSC analysis showed a slight inflection around −50°C and a pronounced peak approximately at 110°C related to LDPE melting, which is soluble in xylene.

The FTIR spectra of SF from extruded WTR contained all NR characteristic bands, although there were evidences of BR and SBR presence as well. TGA showed that SF from extruded WTR is less thermally stable than the original WTR shavings. In addition, SF from WTR extruded with LDPE was the most thermally stable sample, indicating high amount of this thermoplastic available in the SF. Finally, the DSC curves of SF had a distinct inflexion around −60°C assigned to the T_g of NR. DSC analysis also confirmed the high LDPE content in the sample extruded with this thermoplastic, indicating that the devulcanized NR content in that SF is lower than the content in the SF from samples extruded with PP.

Processing conditions in twin-screw extruder, addition of thermoplastics and the WTR particle size affected the amount of generated SF, reflecting in the devulcanization process occurrence and in rubber partial degradation. Higher screw rotating speed and lower particle size increased the SF, reaching values around 29 wt%. Also, the SF of WTR extruded with PP was considerably higher than SF from samples extruded with LDPE. These results promoted a better understanding of thermomechanical devulcanization, supporting future researches and developments.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Rodgers

. Rubber compounding: chemistry and applications. New York: Marcel Dekker Inc, 2004.

Hofmann

. Rubber technology handbook. Munich: Hanser Publishers, 1989.

Imbernon

Norvez

. From landfilling to vitrimer chemistry in rubber life cycle. Eur Polym J 2016; 82: 347–376.

Adhikari

Maiti

. Reclamation and recycling of waste rubber. Prog Polym Sci 2000; 25(7): 909–948.

Isayev

Yushanov

Chen

. Ultrasonic devulcanization of rubber vulcanizates. I. Process model. J Appl Polym Sci 1996; 59(5): 803–813.

Garcia

Sousa

FDB

Lima

. Devulcanization of ground tire rubber: physical and chemical changes after different microwave exposure times. Exp Polym Lett 2015; 9(11): 1015–1026.

Formela

Cysewska

Haponiuk

. The influence of screw configuration and screw speed of co-rotating twin screw extruder on the properties of products obtained by thermomechanical reclaiming of ground tire rubber. Polimery 2014; 59(2): 170–177.

Jalilvand

Ghasemi

Karrabi

. A study of EPDM devulcanization in a co-rotating twin-screw extruder. Iran Polym J 2007; 16(5): 327–335.

Edwards

Danon

van der Gryp

. Quantifying and comparing the selectivity for crosslink scission in mechanical and mechanochemical devulcanization processes. J Appl Polym Sci 2016; 133(37): 43932.

10.

Barbosa

Nunes

Ambrósio

. Devulcanization of natural rubber in composites with distinct crosslink densities by twin-screw extruder. Mater Res 2017; 20(2): 77–83.

11.

Chen

Zhang

. Effects of high shear stress on the devulcanization of ground tire rubber in a twin-screw extruder. J Appl Polym Sci 2013; 128: 2307–2318.

12.

Matsushita

Mouri

Okamoto

. Rubber composition and method for producing the same. Patent 6,316,508, USA, 2001.

13.

Shi

Jiang

Ren

. Structure and performance of reclaimed rubber obtained by different methods. J Appl Polym Sci 2013; 129(3): 999–1007.

14.

Yazdani

Ghasemi

Karrabi

. Continuous devulcanization of waste tires by using a co-rotating twin screw extruder: effects of screw configuration, temperature profile, and devulcanization agent concentration. J Vinyl Addi Technol 2013; 19(1): 65–72.

15.

Flory

. Introduction to physical polymer science. New Jersey: John Wiley & Sons, 2006.

16.

Nora

Ebdonb

. Ozonolysis of natural rubber in chloroform solution. Part 1. A study by GPC and FTIR spectroscopy. Polymer 2000; 41: 2359–2365.

17.

Wanga

Peng

. The impact of esterification on the properties of starch/natural rubber composite. Compos Sci Technol 2009; 69: 1797–1803.

18.

Abhijeet

Ilisch

. The role of linked phospholipids in the rubber-filler interaction in carbon nanotube (CNT) filled natural rubber (NR) composites. Polymer 2014; 55(18): 4738–4747.

19.

Datta

Stocek

Kuritka

. Determination of compounding formulation of cured rubber by reverse engineering. Polym Eng Sci 2015; 55(6): 1450–1458.

20.

Martínez

Puy

Murillo

. Waste tire pyrolysis – A review. Renew Sustain Energy Rev 2013; 23: 179–213.

21.

Ramesan

. Thermogravimetric analysis, flammability and oil resistance properties in natural rubber and dichlorocarbene modified styrene butadiene rubber blends. React Funct Polym 2004; 59(3): 267–274.

22.

Martins

Pessoa

JDC

Gonçalves

. Thermal and mechanical properties of the açaí fiber/natural rubber composites. J Mater Sci 2008; 43: 6531–6538.

23.

Colom

Faliq

Formela

. FTIR spectroscopic and thermogravimetric characterization of ground tyre rubber devulcanized by microwave treatment. Polym Test 2016; 52: 200–208.

24.

Garcia

de Sousa

FDB

de Lima

. Devulcanization of ground tire rubber: physical and chemical changes after different microwave exposure times. Exp Polym Lett 2015; 9(11): 1015–1026.

25.

Fainleib

Pires

Lucas

. Degradation of Non-vulcanized Natural Rubber – Renewable Resource for Fine Chemicals Used in Polymer Synthesis. Polím Ciência e Tecnol 2013; 23: 441–450.

26.

Ambrósio

Larocca

Pessan

. Influence of the process parameters of an intermeshing co-rotating twin screw extruder on the morphology and notched Izod impact strength of PBT/ABS/MGE blends. Polym Eng Sci 2010; 50: 2382–2391.

27.

Agostini

DLS

Constantino

CJL

Job

. Thermal degradation of both latex and latex cast films forming membranes. J Therm Anal Calorim 2008; 91(3): 703–707.

28.

Silverstein

Webster

Kiemle

. Spectrometric identification of organic compounds. New Jersey: John Wiley and Sons, 2005.