Liquefaction of ground tire rubber at low temperature

Waste Management xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Waste Management journal homepage: www.elsevier.com/locate/wasman

Liquefaction of ground tire rubber at low temperature Xiangyun Cheng, Pan Song, Xinyu Zhao, Zonglin Peng, Shifeng Wang ⇑ Department of Polymer Science and Engineering, Shanghai Key Lab. of Electrical Insulation and Thermal Aging, Shanghai Jiao Tong University, Shanghai 200240, China

a r t i c l e

i n f o

Article history: Received 18 May 2017 Revised 31 August 2017 Accepted 4 October 2017 Available online xxxx Keywords: Ground tire rubber Recycling Liquefaction Swelling Carbon black

a b s t r a c t Low-temperature liquefaction has been investigated as a novel method for recycling ground tire rubber (GTR) into liquid using an environmentally benign process. The liquefaction was carried out at different temperatures (140, 160 and 180 °C) over variable time ranges (2–24 h) by blending the GTR with aromatic oil in a range from 0 to 100 parts per hundred rubber (phr). The liquefied GTR was separated into sol (the soluble fraction of rubber which can be extracted with toluene) and gel fractions (the solid fraction obtained after extraction) to evaluate the reclaiming efficiency. It was discovered that the percentage of the sol fraction increased with time, swelling ratio and temperature. Liquefied rubber was obtained with a high sol fraction (68.34 wt%) at 140 °C. Simultaneously, separation of nano-sized carbon black from the rubber networks occurred. The separation of carbon black from the network is the result of significant damage to the cross-linked-network that occurs throughout the liquefaction process. During liquefaction, a competitive reaction between main chain scission and cross-link bond breakage takes place. Ó 2017 Published by Elsevier Ltd.

1. Introduction Vulcanized rubber was developed in the 19th century and has become one of the most widely used polymers across the world. It has been utilized in numerous industrial products, especially in tires (Xiang et al., 2015). Estimated data suggests that the growth in worldwide demand for tires, increasing approximately 4.3% per year, reached nearly 1 billion units in 2015 and 2.9 billion units in 2017 (Machin et al., 2017). Concurrently, the generation of the amount of waste tires has increased significantly, with approximately 17 million tons generated annually around the world (Czajczyn´ska et al., 2017; Luo and Feng, 2017). Thus, the increasingly rubber consumption pose a serious threat to the environment because of the limited number of efficient processes for the disposal of scrap material. Waste rubber recycling is the preferred method for disposal of waste rubber and is supported both legislatively and economically. GTR has the potential to be used as an alternative to rubber (Saiwari et al., 2012) and rubberized asphalt for road construction (Yan et al., 2016). A number of studies have reported the use of GTR in rubber stocks, including natural rubber (NR), butadiene (BR), styrene/butadiene rubber (SBR), etc. The use of 20% reclaimed GTR has been shown to improve the mechanical performance of BR/GTR and SBR/GTR, for example, the model tire compound (BR/ IR/SBR) has been prepared with less than 20 wt% GTR content ⇑ Corresponding author. E-mail address: [email protected] (S. Wang).

(Grigoryeva et al., 2004). Research concerning the incorporation of GTR in NR demonstrated that good tensile strength and abrasion resistance were maintained, especially when smaller particles within a 10 phr loading were used (Li et al., 2004). In order to avoid degrading the basic properties of the materials, it has been suggested that GTR should be subjected to devulcanization or reclamation to enhance the GTR loading content; otherwise, the amount of GTR incorporated into fresh rubber will not surpass 10 phr (Karger-Kocsis et al., 2013). It has also been widely reported that rubberized asphalt significantly improves the mechanical properties of asphalt pavement, leading to high performance in terms of resistance to rutting, cracking, water and fatigue damage. In addition, rubberized asphalt pavement also assists in improving traffic safety and reducing traffic noise (Ameli et al., 2016; Lo Presti et al., 2012; Nazzal et al., 2016). As rubber chains have a chemically crosslinked-network structure, endowed by sulfur cross-linking and reinforced by carbon black, their devulcanization and swelling in asphalt is limited. The lack of interaction between the asphalt and rubber particles causes the rubber to act as only dispersed filler within the asphalt (Gawel et al., 2006; Ghavibazoo and Abdelrahman, 2013) rather than as an effective network-forming modifier. Furthermore, the density of rubber is greater than that of asphalt, which often leads to the poor stability of rubberized asphalt (Lo Presti, 2013). In order to improve the compatibility between asphalt and GTR, pre-devulcanized or reclaimed rubber can be used to prepare rubberized asphalt with better stability and less pollution (Yao et al., 2015).

https://doi.org/10.1016/j.wasman.2017.10.004 0956-053X/Ó 2017 Published by Elsevier Ltd.

Please cite this article in press as: Cheng, X., et al. Liquefaction of ground tire rubber at low temperature. Waste Management (2017), https://doi.org/ 10.1016/j.wasman.2017.10.004

2

X. Cheng et al. / Waste Management xxx (2017) xxx–xxx

The reclamation of GTR involves the conversion of waste vulcanized rubber from the thermoset state to the plastic and viscous states, allowing for the rubber to be more easily processed and revulcanized. Therefore, the reclamation of waste rubber can be considered to a process in which the cross-linked network structure of vulcanized rubber is damaged or degraded due to the effects of heat, oxygen, mechanical force and reclaiming agents. Nowadays, the various methods used to reclaim rubber can be broadly clasified into three groups: (1) chemical reclamation; (2) physical reclamation; and (3) biological reclamation (Saiwari et al., 2012; Shi et al., 2013; Verbruggen et al., 1999). The majority of previous studies have focused on reclaimed rubber consisting of approximately 20% sol fraction (Formela et al., 2014, 2016a). This type of reclaimed rubber has mostly been used in tires in the form of physical filling with carbon black bound in the rubber’s crosslinked network. However, several research groups have started to investigate the use of reclaimed rubber that possesses a high sol fraction content. Such rubber can be applied as a plasticizer in tire rubber and as an extender in rubberized asphalt (highly liquefied rubber is defined as liquefied rubber with possessing over 50% sol fraction content). A co-rotating twin-screw extruder has been used to obtain liquefied rubber at 300 °C consisting of 73.5% sol fraction of 73.5% (Shi et al., 2014). Wu et al. conducted a liquefaction using a temperature-controlled hydraulic press and obtained rubber with over 80 wt% soluble fraction at a temperature of 280 °C (Wu et al., 2017, 2016). Common processing conditions for rubber liquefaction rely on fairly harsh conditions, such as a high temperature (220～300 °C), a high pressure and a strong shearing force. High temperatures usually give rise to the generation of malodorous gases consisting of some components of the rubber or rubber additives volatizing and acting as secondary pollutants. Volatile substances released during the liquefaction process include benzene, toluene, ethylbenzene, xylene and other benzene compounds (BTEX) and sulfides, which are harmful to people’s health (Formela et al., 2016b; Martínez et al., 2014; Rajan et al., 2006). Accordingly, a low cost, highly efficient, easy to operate and environmentally benign GTR liquefaction process is desired. Plasticizers such as aromatic oil and pine oil are added to GTR before devulcanization. For general purposes, these oils are used to improve the diffusion and dispersion of the reclaiming agents. Swelling of GTR with oil containing the chemicals facilitates the penetration of the reactive chemicals into the rubber matrix (Formela et al., 2016b). Additionally, it has been reported to accelerate the oxidation of the rubber and to prevent the formation of gel that acts as a radical acceptor (Rajan et al., 2006). It was found that the sol fraction content of reclaimed rubbers progressively increases with following increase in oil content (in range: 0–20 phr) (Xu et al., 2014). In this study, a greater quantity of oil has been added to the reaction system in order to achieve liquefaction of waste rubber. It is envisaged that the liquefaction of rubber at low temperatures and with high oil content can be achieved due to promotion of liquefaction caused by the pre-swelling of GTR by the oil. Thereafter, the highly liquefied rubber can be used as a plasticizer alternative for tire rubber and asphalt. An additional part of this work focuses on the presence of carbon black in the liquefied rubber. Carbon black is added to rubber for reinforcing the three-dimensional cross-link rubber network because of the strong physical and chemical reaction between them. It is generally believed that the rubber that is present outside of the carbon black exists in three different states: primary bound rubber, loosely bound rubber, and the free rubber chains (Leblanc, 2002; Li et al., 2016a). Bound rubber is formed through interactions involving physical bonding via vander Waals forces, and polymer-polymer and carbon black-polymer chemical covalent bonding during the compounding process; it cannot be dis-

solved by the solvent that is used for rubber (Gabriel et al., 2016; Litvinov et al., 2011). The interactions between rubber and carbon black influence the separation of rubber from carbon black during the liquefaction of GTR. 2. Experimental 2.1. Materials GTR (500–600 mm) was obtained by shredding and grinding scrap rubber at ambient temperature by Jiangsu Baoli International Investment Co., Ltd (Jiangyin, China). The GTR consists of 8.24 wt% soluble material, 41.46 wt% natural rubber (NR), 12.09 wt% synthetic rubber (SR), 31.59 wt% carbon black and 6.62 wt% inorganic filler. The aromatic oil used for the experiments was furfural extract oil of commercial grade and was supplied by Shanghai Gaoqiao Petrochemical Co., Ltd (Shanghai, China), and the percentage of aromatic hydrocarbon compounds was 65%. 2.2. Liquefying process of GTR The GTR (mG) and aromatic oil (ma) were weighed and mixed at room temperature in certain proportions (1/1 signifies that the ratio of GTR to oil is 1:1, 2/1 is 2:1, 3/1 is 3:1, etc.). The blends were first stirred for approximately 3 min, then placed in an oven with hot air at constant temperature for liquefaction. The liquefaction of GTR was conducted at low temperature (140–180 °C), for a short period of time (2–24 h) and with a variable oil content (0–100 phr). Traditionally, the content of oil used in the devulcanization of GTR is less than 20 phr. To enhance the liquefaction effect and reduce the liquefaction temperature, a high oil content was selected. The experimental conditions for the samples are presented in Table 1. 2.3. Sol-gel analysis Both the content of sol fraction and the gel fraction of the liquefied rubbers were determined in order to carry out a comprehensive investigation of the relationship between liquefaction conditions and the degradation efficiency of GTR. The blends of GTR and aromatic oil were enclosed in filter paper and weighed (m0) before liquefaction. Then liquefied samples were subjected to Soxhlet extraction in toluene for 72 h to remove the majority of the soluble low molecular weight compounds and to separate the sol fraction from the products. The residues were dried in a vacuum oven at 60 °C to constant weight and weighed again (m1), and were taken to be the gel fractions. It is worth noting that the aromatic oil (ma) and soluble material (mb) of the original GTR, liquefied rubber were both extracted with toluene, thus they were deducted when calculating the content. In addition, the mass of rubber (mr) in GTR was not equal to that of mG, this can be Table 1 Samples and corresponding experimental conditions. Sample

Oil content (phr)

Temperature (°C)

Time (h)

1/1 1/1-140-2 1/1-140-4 1/1-140-6 1/1-140-8 1/1-140-12 1/1-140-24 1/1-160-4 1/1-180-4 2/1-140-4 3/1-140-4

100 100 100 100 100 100 100 100 100 50 33.3

/ 140 140 140 140 140 140 160 180 140 140

/ 2 4 6 8 12 24 4 4 4 4

Please cite this article in press as: Cheng, X., et al. Liquefaction of ground tire rubber at low temperature. Waste Management (2017), https://doi.org/ 10.1016/j.wasman.2017.10.004

X. Cheng et al. / Waste Management xxx (2017) xxx–xxx

3

explained by material composition. The sol fraction was calculated using Eq. (1):

Sol fraction ¼

m0  m1  ma  mb  100% mr

ð1Þ

After drying, the gel samples were shaped rapidly at ambient temperature on a two-roll mill. Subsequently, the sheets were cut into circles with a diameter of 2.2 cm. These small circular samples were then placed in toluene at room temperature for 72 h to ensure uniform swelling. The swollen samples were removed from the toluene, gently dabbed with filter paper to remove excess solvent and then weighed before being dried in a vacuum oven overnight at 60 °C to constant weight (Wu et al., 2017). The cross-link density of the rubber network was calculated by the swelling method according to the Flory-Rehner Eqs. (2) and (3) given by Horikx (1956):

Ve ¼

tr þ vt2r þ lnð1  tr Þ ts ð0:5tr  t1=3 r Þ

ð2Þ

tr ¼

mr mr þ ms ðqr =qs Þ

ð3Þ

where Ve is the cross-link density per unit volume (mol/cm3), ʋr is the volume fraction of the polymer in the swollen gel phase of the sample, ʋs is the solvent molar volume (106.4 cm3/mol), mr is the mass of the rubber (g), ms is the weight of solvent in the sample at equilibrium swelling (g), qr is the density of the GTR (1.15 g/ cm3), qs is the density of the solvent (0.866 g/cm3) and v is the Flory-Huggins polymer-solvent interaction parameter (0.43 for the GTR/toluene) (Xu et al., 2014). In consideration of experimental error, both the sol fraction and the cross-link density measurements of the gel were conducted in triplicate and are indicated with error bars in the following figures. 2.4. Efficiency of liquefaction of GTR A theoretical relationship between the sol fraction and the decreasing cross-link density of the gel fraction has been suggested by Horikx. Horikx’s equations describe two limiting cases of network breakdown: main-chain scission and cross-link scission, both of which can be used to discuss rubber degradation. In the case of only main-chain scission occurring, the relationship between the sol fraction and the relative decrease in cross-link density can be described by Eq. (4):

2 3 1=2 2 ð1  Sf Þ Vf 4 5 1 ¼1 2 Vi ð1  S1=2 Þ

ð4Þ

i

In the case of only cross-link scission taking place, the relationship between the sol fraction and the relative decrease in cross-link density can be described by Eq. (5):

2 3 2 cf ð1  S1=2 Þ Vf f 5 1 ¼14 2 Vi c ð1  S1=2 Þ i

ð5Þ

i

where Si and Sf are the sol fraction of the rubber network before and after liquefaction, respectively. Vi and Vf are the cross-link density of the network before and after the treatment and liquefaction, respectively. ci and cf are the average number of cross-link sites per chain in the insoluble network before and after liquefaction respectively, as described by Verbruggen et al. (1999). Fig. 1 was drawn according to Eqs. (4) and (5). The solid curve in Fig. 1 represents the condition in which only main chains are broken and the dashed curve corresponds to only cross-link scission. By applying experimental data to the model, the sites at which rubber molecular

Fig. 1. Random main-chain scission and cross-link scission curves as proposed by the Horikx theory.

chain scission and cross-link breakage take place could be estimated. 2.5. Characterization and measurements The molecular weight of the sol fraction after toluene extraction was characterized by gel-permeation chromatography (GPC) (HLC8320GPC, Tosoh Co., Japan). GPC analyses were performed with a tetrahydrofuran (THF) flow rate of 0.35 mL min1 at 40 °C, and polystyrene was used as the standard. TGA (Q5000IR, TA Instruments, USA) was used to study the thermal degradation behavior of the gel in the samples under different experimental conditions (Datta et al., 2017; Garcia et al., 2015). Samples of approximately 4 mg were heated from room temperature to 550 °C under a nitrogen flow and at a heating rate of 20 °C min1, and were then heated from 550 to 700 °C under an oxygen flow with a heating rate of 20 °C min1. An elemental analyzer (EA) (Vario EL Cube, Elementar, Germany) was used to determine sulfur content in the gel fraction of the liquefied samples. Rheological characterization was performed using a rubber processing analyzer (RPA) (Alpha technologies, USA). Frequency sweep tests were carried out at a temperature of 45 °C, at a strain amplitude of 7% and within a frequency range between 0.1 Hz and 30 Hz (Mangili et al., 2015). The carbon black sample was separated by centrifugation to remove the aromatic oil and the sol fraction from the toluene extract. The morphology of carbon black that was separated by liquefaction was observed with SEM using a Nova NanoSEM 450 (FEI Company, USA) and TEM were carried out with a JEM-2100 (JEOL Ltd., Japan). To prepare the suspension (0.02 g L1), carbon black particles were dispersed into toluene via ultrasonication for 30 min. The well-dispersed carbon black samples in toluene were dropped onto small pieces of silicon sheets or carbon support films and were dried under vacuum at 60 °C for 24 h (Li et al., 2016b). 3. Results and discussion 3.1. Effect of reaction time on liquefaction of GTR The effect of time on the liquefaction of GTR was studied by preparing swollen GTR possessing 100 phr oil was prepared at different reaction times. During the liquefaction process, part of the gel fraction in the rubber is converted to a sol fraction that can be extracted with solvent. As shown in Fig. 2(A), the amount of

Please cite this article in press as: Cheng, X., et al. Liquefaction of ground tire rubber at low temperature. Waste Management (2017), https://doi.org/ 10.1016/j.wasman.2017.10.004

4

X. Cheng et al. / Waste Management xxx (2017) xxx–xxx

Fig. 2. Effect of reaction time on liquefaction of GTR (A) Percentage of sol fraction content of GTR after liquefaction, (B) TGA results for the samples, (C) Cross-link density of GTR after liquefaction, (D) Horikx plot of the data from liquefaction, (E) S content in the gel fraction, (F) Storage modulus of the liquefied samples (with 100 phr aromatic oil at 140 °C and with different reaction times).

sol fraction increased sharply within 4 h, followed by a slow increase from 4 h to 8 h until reaching a maximum value (68.34 wt%) at 8 h. The amount of sol fraction then decreases slightly following an increase in time. This result is consistent with changes in the rubber content in the gel fraction of GTR, as shown in Fig. 2(B). Obviously, liquefaction proceeds with less degradation of the GTR

when compared to conventional thermodynamic liquefaction methods carried out using a co-rotating twin screw extruder at 280 °C (Yazdani et al., 2011). The cross-link density of the samples was analyzed using the swelling method and relevant experimental results are presented in Fig. 2(C); these results show an upward trend with increasing

Please cite this article in press as: Cheng, X., et al. Liquefaction of ground tire rubber at low temperature. Waste Management (2017), https://doi.org/ 10.1016/j.wasman.2017.10.004

5

X. Cheng et al. / Waste Management xxx (2017) xxx–xxx

time (Edwards et al., 2016). The relationship between the amount of sol content and the decrease in cross-link density was established using the Horikx theory, as shown in Fig. 2(D). The data in the initial stage (within 2 h) is located on the line of main chain scission. However, data obtained from liquefaction at a later stage is distributed above the curve related to main chain scission and can be attributed to heterogenous liquefaction process. In this case, part of the cross-link network in GTR is damaged resulting in an increase in the percentage of more sol fraction content. The remaining crosslink network undergoes reorganization increasing the crosslink density (Saiwari et al., 2012). Fig. 2(E) shows that the sulfur content in the gel fraction of the samples first decreases and then increases, indicating that there is competition between main chain scission and cross-link bond breakage during the liquefaction process. The sol content in 1/1-140-8 is the highest, while the sulfur content in the gel fraction is the lowest. This indicates greater crosslink bond breakdown, so that the sulfur element migrates from the gel fraction of GTR to the sol fraction. As shown in Fig. 2(F), the storage modulus (G0 ) of sample 1/1 is the largest before liquefaction. When liquefaction was conducted for 8 h, the G’ of the sample reaches a minimum value and then increases with time. The G’ represents the ability of the material to store elastic deformation energy. Compared to the original blend of GTR and aromatic oil, the G’ of the experimental samples is greatly reduced after liquefaction. During the liquefaction process, a portion of the gel fraction is converted to the sol fraction with a low molecular weight, contributing to a significant decrease in the elasticity of the samples. Liquefaction of GTR over 8 h produces the highest content of sol fraction and so sample1/1-140-8 has the lowest storage modulus. The number-average molecular weight (Mn) of the sol fraction following toluene extraction of the samples is shown in Table 2. The Mn of each sample has two peaks; the lower peak is considered to be the Mn of the aromatic oil (extracted by toluene during the extraction process), and the higher peak is the Mn of sol fraction. The Mn and Mw of the sol fraction gradually increase with time and this increase is related to recombination of the active chain segments. Two types of reactions most likely occur during liquefaction because of the complicated composition of GTR. GTR consists of both NR and SR, reinforced with carbon black and cross-linked with sulfur. Different rubbers exhibit different structural changes during thermal degradation. The degradation of NR containing isoprene units occurs due to molecular chain fracture; conversely the degradation of the SR (possessing butadiene structural units) cross-link arises predominantly due to cross-linking reactions (Madorsky, 1964; Saiwari et al., 2014). Therefore, in addition to

the competitive reaction between main chain scission and crosslink breakage, destruction of the cross-link network and recombination of the active chain segments, occur simultaneously (Wu et al., 2017). As shown in Table 3, the SR content in the gel fraction of GTR increases after 8 h thus proving that a cross-linking reaction is present in the liquefaction process. However, during the liquefaction process, recombination of the active chain segments gradually becomes the dominant mechanism as the percentage of the sol fraction first increases and then decreases, while the Mn of the sol fraction increases. Within an 8 h period, breakage of the cross-linked network is the dominant reaction and this leads to an increase in the amount of sol fraction content. The reorganization of chain segments intensifies after 8 h leading to a decrease in the amount of sol fraction content. In general, reaction time is an important factor that influences the liquefaction process, and the desired reaction product can be obtained by selecting an appropriate reaction time. 3.2. Effect of temperature on the liquefaction of GTR Temperature is one of the most important physical factors that influence the degradation of rubber. To determine the effect of temperature on the liquefaction process, we varied the temperature from 140 to 180 °C. The percentage of sol fraction of the liquefied samples at different temperature is shown in Fig. 3(A). TGA results of the gel fraction are shown in Fig. 3(B). The sol fraction content of 1/1-180-4 reaches 77.58 wt%. To some extent, the transformation of the gel fraction into the sol fraction indicates severe network degradation (Jiang et al., 2013). On the other hand, the cross-linked density of the sample increases with increasing temperature; this increase gradually slows at higher temperatures, as shown in Fig. 3 (C). These phenomena are due to the various reactions that occur during the reclaiming process i.e. damage to the rubber network and the generation of reactive radicals. The former increases the amount of the sol fraction in the liquefied rubber, while the latter increases the

Table 3 Percentage of different components in the gel fraction of the samples. Sample

Rubber (wt%)

NR (wt%)

SR (wt%)

GTR 1/1-140-2 1/1-140-4 1/1-140-6 1/1-140-8 1/1-140-12 1/1-140-24

53.55 52.39 43.71 41.13 36.37 37.10 40.91

41.46 40.51 29.65 29.42 26.87 25.97 26.31

12.09 11.88 14.06 11.71 9.50 11.13 14.60

Table 2 Mn of the sol fraction in different samples. Sample name

Aromatic oil 1/1-140-2 1/1-140-4 1/1-140-6 1/1-140-8 1/1-140-12 1/1-140-24 1/1-160-4 1/1-180-4 2/1-140-4 3/1-140-4

M n (104 gmol1)

M w (104 gmol1)

Mw =Mn

Peak1

Peak2

Peak1

Peak2

Peak1

Peak2

— 6.13 9.57 9.51 10.95 13.19 15.12 12.20 12.57 7.78 7.46

0.59 0.50 0.48 0.46 0.74 0.69 0.65 0.82 0.86 0.49 0.50

— 6.18 10.62 10.61 12.51 15.57 18.23 14.12 14.54 9.01 8.57

1.28 1.30 0.94 0.95 1.17 1.19 1.36 1.28 1.37 0.98 0.86

— 1.01 1.11 1.12 1.14 1.18 1.21 1.16 1.16 1.16 1.15

2.18 2.59 1.94 2.07 1.59 1.72 2.11 1.57 1.59 2.00 1.73

Please cite this article in press as: Cheng, X., et al. Liquefaction of ground tire rubber at low temperature. Waste Management (2017), https://doi.org/ 10.1016/j.wasman.2017.10.004

6

X. Cheng et al. / Waste Management xxx (2017) xxx–xxx

Fig. 3. Effect of temperature on the liquefaction of GTR (A) Sol fraction of GTR after liquefaction, (B) TGA results of the samples, (C) Cross-link density of GTR after liquefaction, (D) Horikx plot of the data from liquefaction, (E) S content in the gel fraction of the samples, (F) Storage modulus of the samples (with 100 phr aromatic oil in 4 h at different temperatures).

cross-link density. A reduction in the gel content of the liquefied rubber limits the increase in cross-link density. The relative data points (1/1-140-4, 1/1-160-4 and 1/1-180-4) have been analyzed by applying Horikx theory. Fig. 3 (D) shows that the data point of samples 1/1-160-4 and 1/1-180-4 fall outside the interface, while the data points for sample 1/1-140-4 are

located above the curve representing main chain scission (this is where the sol fraction content is high and the decrease in crosslink density is minimal). This signifies that a portion of the rubber network is seriously damaged and is transformed into sol fraction while other parts of the rubber network remain highly crosslinked. This indicates that the heterogenous liquefaction of sample

Please cite this article in press as: Cheng, X., et al. Liquefaction of ground tire rubber at low temperature. Waste Management (2017), https://doi.org/ 10.1016/j.wasman.2017.10.004

X. Cheng et al. / Waste Management xxx (2017) xxx–xxx

1/1-140-4 is a result of an extensive recombination of the active chain fragment, which becomes more pronounced at higher temperatures. However, due to the limitation of the volume space of the rubber network, the cross-link density of the samples cannot infinitely increase. In addition, the Mn of the sol fraction also increases when the temperature rises, which indicates that a high temperature promotes the recombination of small molecular chains in the sol fraction.

7

Changes in the sulfur content of the gel fraction are shown in Fig. 3(E), the percentage of sulfur content decreases as the temperature increases. Theoretically, the selective chain-breaking mechanism of vulcanized rubber is based on two factors: the dissociation energy of chemical bonds and the difference in elasticity of chemical bonds (Dluzneski, 2001; Fukumori and Matsushlta, 2003). In general, the thermal stability of the chemical bond depends on the value of the dissociation energy. The higher the dissociation

Fig. 4. Effect of oil content on GTR liquefaction (A) Sol fraction of GTR after liquefaction, (B) Cross-link density of GTR after liquefaction, (C) TGA results of the gel fraction in liquefied rubber, (D) Horikx plot of data from liquefaction, (E) S content in the gel of the samples, (F) Storage modulus of the samples as a function of frequency (with different oil content at 140 °C in 4 h).

Please cite this article in press as: Cheng, X., et al. Liquefaction of ground tire rubber at low temperature. Waste Management (2017), https://doi.org/ 10.1016/j.wasman.2017.10.004

8

X. Cheng et al. / Waste Management xxx (2017) xxx–xxx

energy, the better the thermal stability. Since the dissociation energy of the SAS/CAS bond is smaller than that of the CAC bond, sulfur is more readily removed from the rubber network so that more sulfur elements migrate from the gel fraction to the sol fraction. As shown in Fig. 3(F), the elasticity of the samples is significantly decreased at high temperature, and that the storage modulus values of 1/1-160-4 and 1/1-180-4 are very closely related because of their similar sol fraction contents. From these results, it can be seen that temperature plays a crucial role in the liquefaction process and that temperatures in the range of 140–180 °C facilitate the degradation of GTR. Although the reorganization of the network and the formation of active chain segments takes place simultaneously at high temperature, network breakdown remains the dominant process in the reaction system. 3.3. Effect of oil content on GTR liquefaction Liquefaction was carried out using aromatic oil and the effect of the oil content was investigated. Samples 3/1-140-4, 2/1-140-4 and 1/1-140-4 were prepared by blending GTR with different amounts of aromatic oil (33.33 phr, 50 phr and 100 phr) at 140 °C over 4 h. From Fig. 4(A)–(C), it can be seen that, compared to the sample that contains no aromatic oil, the addition of aromatic oil strongly increases the sol fraction content. These data suggest that the high degree of swelling greatly promotes the degradation of GTR. As shown in Fig. 4(C), the cross-link density is lowest for sample 3/1-140-4 which has a low oil content, while 1/1-140-4 possessing a higher oil content has the largest cross-link density. During the reclaiming process, the value gradually increases as the aromatic oil loading increases. As shown in Fig. 4(D), the three experimental points lie above the cross-link scission curve in the Horikx plot due to the heterogenous liquefaction of vulcanized rubber. This trend becomes more obvious as the oil content increases. As mentioned previously, competitive reactions are present in the liquefaction system. The swelling behavior of the rubber is considered as a major factor for the promotion of liquefaction (Saiwari et al., 2013; Xu et al., 2014); the cross-linked network of vulcanized rubber is stretched and loosened in the aromatic oil leading to a weakening of the intermolecular forces and reducing the entanglement of molecular chains. Therefore, damage to the rubber network dominates. However, higher aromatic oil content leads to a more heterogenous degradation of the samples. It can be assumed that the oil not only facilitates the recombination of the molecular chain segments with the rubber network forming new cross-linked points and thus increasing the cross-link density, but also promotes interactions between the active chain segments, resulting in an increase in Mn of the sol fraction (Table 2). As depicted in Fig. 4(E), it can be seen that the sulfur content of the gel fraction of the three samples is similar, indicating that the aromatic oil has little effect on the selective scission between the main chain and the cross-linked bonds. As show in Fig. 4(F), the elasticity of the samples decreases as oil content increases. These results are consistent with measurements carried out for the sol fractions. Remarkably, the storage modulus of aromatic oil is lower than that of GTR that possesses a greater proportion of aromatic oil in the samples, thus giving a lower storage modulus. This is also one of the factors leading to differences in the elasticity of samples with differing oil content. The degradation of GTR under the combined action of heat and oxygen is based on a mechanism of free radical chain autocatalytic oxidation. The effect of the addition of aromatic oil on the liquefaction process is clear. In the first case, the cross-linked network stretches as the GTR swells in the aromatic oil. Concurrently, intermolecular forces that are present between strands over large dis-

tances are weakened and the entanglement of molecular chains is also reduced; this leads to a more facile decomposition of the molecular chain. Secondly, the network is more easily able to come into contact with air, thus allowing the radicals that are produced by chain breakage to more readily react with oxygen and accelerate the thermal aging process of the rubber (Rajan et al., 2006). However, the percentage of oil content is also limited by the amount of swelling that can occur in certain amounts of oil (Flory, 1950). 3.4. Morphology of carbon black separated from GTR during liquefaction During the extraction process, it was observed that the toluene extracts of samples 1/1-140-4, 1/1-140-6, 1/1-140-8, 1/1-140-12, 1/1-140-24, 1/1-160-4 and 1/1-180-4 are black in color, indicating that the carbon black had separated from the GTR. The toluene extract of 1/1-140-4 (1/1-140-4-s) was found to have a weight loss of 3.24 wt% at temperatures between 550 and 700 °C (Fig. 5). Because the toluene extraction contain aromatic oil, the amount of carbon black is estimated to be 2.33 wt% following deduction of the weight loss of the aromatic oil at temperatures between 550 and 700 °C. The TEM and SEM images of the carbon black separated from liquefied rubber are shown in Figs. 6 and 7. From these images it can be seen that carbon black particles are about 30–50 nm in diameter and aggregate with a size of approximately 200 nm. The carbon black’s surface is covered with a layer of rubber, showing a core-shell structure with a thickness between 2 nm and 10 nm. In the liquefaction process, carbon black is separated from the rubber network in two steps, as illustrated in Fig. 8. Firstly, the aromatic oil enters the rubber network leading to an expansion of the network structure. The main chain and cross-linked bonds of the rubber break and the three-dimensional cross-link network is severely damaged by heat and oxygen. It can be assumed that when the free rubber chain breaks, the carbon black that is coated with bound rubber separates from the cross-link network of the GTR during the extraction process. However, no carbon black was present in any of the toluene extracts of samples 2/1-140-4 and 3/1-140-4, which indicates that the separation of carbon black is also limited by the liquefaction conditions. It is reasonable to assume that a larger amount of carbon black can be obtained by liquefaction of GTR at 140 °C over a certain period of time and with greater oil loadings; however, this requires further investigation.

Fig. 5. TGA results for aromatic oil and 1/1-140-4-s.

Please cite this article in press as: Cheng, X., et al. Liquefaction of ground tire rubber at low temperature. Waste Management (2017), https://doi.org/ 10.1016/j.wasman.2017.10.004

X. Cheng et al. / Waste Management xxx (2017) xxx–xxx

9

Fig. 6. SEM micrographs of carbon black separated by liquefaction.

Fig. 7. TEM micrographs of carbon black separated by liquefaction.

Fig. 8. Process of separation of carbon black from GTR by liquefaction.

4. Conclusions Liquefaction has been put forward as a new method provides an alternative method for the recycling of GTR. The process can be

carried out at a low temperature to avoid the formation of secondary pollution pollutants and produce liquefied rubber with a high sol fraction content (68.34 wt%). The efficiency of the liquefaction is significantly influenced by time, temperature and the oil

Please cite this article in press as: Cheng, X., et al. Liquefaction of ground tire rubber at low temperature. Waste Management (2017), https://doi.org/ 10.1016/j.wasman.2017.10.004

10

X. Cheng et al. / Waste Management xxx (2017) xxx–xxx

content. The optimum reaction time for liquefaction at 140 °C with 100 phr aromatic oil is 8 h. Longer liquefaction times result in a dominant recombination reaction and reduced reaction efficiency. Additionally, a significant improvement in liquefaction efficiency is observed at higher temperature; the temperature should be maintained within certain range because it promotes the recombination of the active chain segments that are separated in the system. The addition of aromatic oil also significantly influences the degradation of GTR. This is due to stretching of the cross-link network in the oil which reduces the intermolecular forces and weakens the entanglement of the molecular chains. Moreover, the expansion of the network increases the accessibility of oxygen, leading to an acceleration in thermal-oxidative aging. Furthermore, carbon black that is separated from the reaction system during liquefaction was present in the toluene extracts. The carbon black is aggregated with a size of approximately 200 nm, and the outer surface of the carbon black is coated with a layer of rubber with a size of 2–10 nm, showing a clear core-shell structure. This observed phenomenon will have a significant impact on the application of materials produced by liquefaction. Additionally, competing reactions take place simultaneously in the reaction system, leading to a heterogenous degradation of GTR under different conditions. Based on the presented results, further studies in this research area should focus on the optimization of the liquefaction process and the conditions required for the separation of carbon black. Acknowledgments The authors are grateful for the research foundation provided by the International Cooperation Project (2013DFR50550) and Guangzhou Science and Technology Project (201604020126). References Ameli, A., Babagoli, R., Aghapour, M., 2016. Laboratory evaluation of the effect of reclaimed asphalt pavement on rutting performance of rubberized asphalt mixtures. Petrol. Sci. Technol. 34, 449–453. _ ´ ska, R., Jouhara, H., Spencer, N., 2017. Use of pyrolytic gas Czajczyn´ska, D., Krzyzyn from waste tire as a fuel: a review. Energy. Datta, S., Antos, J., Stocek, R., 2017. Characterisation of ground tyre rubber by using combination of FT-IR numerical parameter and DTG analysis to determine the composition of ternary rubber blend. Polym. Test. 59, 308–315. Dluzneski, P.R., 2001. Peroxide vulcanization of elastomers. Rubber Chem. Technol. 74, 451. Edwards, D.W., Danon, B., van der Gryp, P., Görgens, J.F., 2016. Quantifying and comparing the selectivity for crosslink scission in mechanical and mechanochemical devulcanization processes. J. Appl. Polym. Sci. 133. Flory, P.J., 1950. Statistical Mechanics of Swelling of Network Structures. J. Chem. Phys. 18, 108–111. Formela, K., Cysewska, M., Haponiuk, J., 2014. 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 -Warsaw 59, 170–177. Formela, K., Cysewska, M., Haponiuk, J.T., 2016a. Thermomechanical reclaiming of ground tire rubber via extrusion at low temperature: efficiency and limits. J. Vinyl Add. Technol. 22, 213–221. Formela, K., Klein, M., Colom, X., Saeb, M.R., 2016b. Investigating the combined impact of plasticizer and shear force on the efficiency of low temperature reclaiming of ground tire rubber (GTR). Polym. Degrad. Stab. 125, 1–11. Fukumori, K., Matsushlta, M., 2003. Material recycling technology of crosslinked rubber waste. R&D Rew Toyota, CRDL 38, 39–47. Gabriel, D., Karbach, A., Drechsler, D., Gutmann, J., Graf, K., Kheirandish, S., 2016. Bound rubber morphology and loss tangent properties of carbon-black-filled rubber compounds. Colloid Polym. Sci. 294, 501–511. Garcia, P.S., Sousa, F.D.B.D., Lima, J.A.D., Cruz, S.A., Scuracchio, C.H., 2015. Devulcanization of ground tire rubber: physical and chemical changes after different microwave exposure times. Exp. Polym. Lett. 9, 1015. Gawel, I., Stepkowski, R., Czechowski, F., 2006. Molecular interactions between rubber and asphalt. Indust. Eng. Chem. Res. 45, 3044–3049. Ghavibazoo, A., Abdelrahman, M., 2013. Composition analysis of crumb rubber during interaction with asphalt and effect on properties of binder. Int. J. Pavement Eng. 14, 517–530.

Grigoryeva, O., Fainleib, A., Starostenko, O., Danilenko, I., Kozak, N., Dudarenko, G., 2004. Ground Tire Rubber (GTR) reclamation: virgin rubber/reclaimed GTR (RE) vulcanizates. Rubber Chem. Technol. 77, 131–146. Horikx, M.M., 1956. Chain scission in a polymer network. J. Polym. Sci. 19, 445–454. Jiang, K., Shi, J., Ge, Y., Zou, R., Yao, P., Li, X., Zhang, L., 2013. Complete devulcanization of sulfur-cured butyl rubber by using supercritical carbon dioxide. J. Appl. Polym. Sci. 127, 2397–2406. Karger-Kocsis, J., Mészáros, L., Bárány, T., 2013. Ground tyre rubber (GTR) in thermoplastics, thermosets, and rubbers. J. Mater. Sci. 48, 1–38. Leblanc, J.L., 2002. Rubber–filler interactions and rheological properties in filled compounds. Progr. Polym. Sci. 27, 627–687. Li, S., Lamminmäki, J., Hanhi, K., 2004. Effect of ground rubber powder on properties of natural rubber. Macromol. Symp. 216, 209–216. Li, S., Wan, C., Wang, S., Zhang, Y., 2016a. Separation of core-shell structured carbon black nanoparticles from waste tires by light pyrolysis. Compos. Sci. Technol. 135, 13–20. Li, S., Wan, C., Wu, X., Wang, S., 2016b. Core-shell structured carbon nanoparticles derived from light pyrolysis of waste tires. Polym. Degrad. Stab. 129, 192–198. Litvinov, V.M., Orza, R.A., Klüppel, M., Duin, M.V., Magusin, P.C.M.M., 2011. Rubberfiller interactions and network structure in relation to stress-strain behavior of vulcanized, carbon black filled EPDM. Macromolecules 44, 4887–4900. Lo Presti, D., 2013. Recycled tyre rubber modified bitumens for road asphalt mixtures: a literature review. Constr. Build. Mater. 49, 863–881. Lo Presti, D., Airey, G., Partal, P., 2012. Manufacturing terminal and field bitumentyre rubber blends: the importance of processing conditions. Proc. - Social Behav. Sci. 53, 485–494. Luo, S., Feng, Y., 2017. The production of fuel oil and combustible gas by catalytic pyrolysis of waste tire using waste heat of blast-furnace slag. Energy Convers. Manage. 136, 27–35. Machin, E.B., Pedroso, D.T., Jr, J.A.D.C., 2017. Energetic valorization of waste tires. Renew. Sustain. Energy Rev. 68, 306–315. Madorsky, S.L., 1964. Thermal Degradation of Organic Polymers. Interscience Publishers. Mangili, I., Lasagni, M., Anzano, M., Collina, E., Tatangelo, V., Franzetti, A., Caracino, P., Isayev, A.I., 2015. Mechanical and rheological properties of natural rubber compounds containing devulcanized ground tire rubber from several methods. Polym. Degrad. Stab. 121, 369–377. Martínez, J.D., Murillo, R., García, T., Arauzo, I., 2014. Thermodynamic analysis for syngas production from volatiles released in waste tire pyrolysis. Energy Convers. Manage. 81, 338–353. Nazzal, M.D., Iqbal, M.T., Kim, S.S., Abbas, A.R., Akentuna, M., Quasem, T., 2016. Evaluation of the long-term performance and life cycle costs of GTR asphalt pavements. Constr. Build. Mater. 114, 261–268. Rajan, V.V., Dierkes, W.K., Joseph, R., Noordermeer, J.W.M., 2006. Science and technology of rubber reclamation with special attention to NR-based waste latex products. Progr. Polym. Sci. 31, 811–834. Saiwari, S., Dierkes, W.K., Noodermeer, J.W.M., 2012. Tires back into tires devulcanization of tires: devulcanization of tire rubbers, 10th Fall Rubber Colloquium. Saiwari, S., Dierkes, W.K., Noodermeer, J.W.M., 2014. Comparative investigation of the devulcanization parameters of tire rubber. Rubber Chem. Technol. 87, 31– 42. Saiwari, S., Dierkes, W.K., Noordermeer, J.W.M., 2013. Devulcanization of whole passenger car tire material. Kgk Rubberpoint 66, 20–25. Shi, J., Jiang, K., Ren, D., Zou, H., Wang, Y., Lv, X., Zhang, L., 2013. Structure and performance of reclaimed rubber obtained by different methods. J. Appl. Polym. Sci. 129, 999–1007. Shi, J., Zou, H., Ding, L., Li, X., Jiang, K., Chen, T., Zhang, X., Zhang, L., Ren, D., 2014. Continuous production of liquid reclaimed rubber from ground tire rubber and its application as reactive polymeric plasticizer. Polym. Degrad. Stab. 99, 166– 175. Verbruggen, M.A.L., Does, L.V.D., Noodermeer, J.W.M., 1999. Mechanisms Involved in the Recycling of NR and EPDM. Rubber Chem. Technol. 72, 731–740. Wu, X., Formela, K., Rasool, R.T., Wang, S., 2017. Evaluation of structural change during fast transformation process of cross-linked NR into liquid NR by light pyrolysis. Polym. Degrad. Stab. 136, 48–57. Wu, X., Wang, S., Dong, R., 2016. Lightly pyrolyzed tire rubber used as potential asphalt alternative. Constr. Build. Mater. 112, 623–628. Xiang, H.P., Qian, H.J., Lu, Z.Y., Rong, M.Z., Zhang, M.Q., 2015. Crack healing and reclaiming of vulcanized rubber by triggering the rearrangement of inherent sulfur crosslinked networks. Green Chem. 17, 4315–4325. Xu, J.-Y., Mei, S., Wang, X.-Q., Chen, C.-H., Xin, Z.-X., 2014. The effect of reclaim softener on properties of reclaimed rubber. J. Macromol. Sci., Part B 53, 1182– 1192. Yan, K.Z., He, W., Chen, M., Liu, W., 2016. Laboratory investigation of waste tire rubber and amorphous poly alpha olefin modified asphalt. Constr. Build. Mater. 129, 256–265. Yao, H., Zhou, S., Wang, S., 2015. Structural evolution of recycled tire rubber in asphalt. J. Appl. Polym. 133, 42954. Yazdani, H., Karrabi, M., Ghasmi, I., Azizi, H., Bakhshandeh, G.R., 2011. Devulcanization of waste tires using a twin-screw extruder: the effects of processing conditions. J. Vinyl Addit. Technol. 17, 64–69.

Please cite this article in press as: Cheng, X., et al. Liquefaction of ground tire rubber at low temperature. Waste Management (2017), https://doi.org/ 10.1016/j.wasman.2017.10.004