Char from Pyrolysis of Waste Tires to Increase Bitumen Performances

Abstract

Road pavement characteristics can be improved by modifying bitumens through addition of fine particles. To avoid environmental issues, attention is recently being paid to bio-materials. In this study, a 50/70 bitumen was modified through the addition of char obtained from the pyrolysis of waste tires. Char addition causes an increase in transition (gel to sol) temperature of up to 4 °C and an increase in rigidity under working conditions (50 °C) of up to about one order of magnitude. The effect of the gas type flowing under the pyrolysis process (CO₂, N₂ and CO₂ + N₂) on the mechanical characteristics of the bitumen was also investigated. More marked effects on the rheological performances were found if char coming from pyrolysis under CO₂ was used (the gel-to-sol transition temperature is increased by about 4.5 °C) compared to that coming from pyrolysis under N₂ (gel-to-sol transition temperature is increased by about 3 °C). The effect is at a maximum for char from CO₂ at 3% wt/wt, whereas regarding char from N, a lesser effect is seen with a more uniform (flat) effect as a function of char% and with an increase in gel-to-sol transition temperature of about 3 °C. Coherently, char obtained from pyrolysis under CO₂ was found to possess a higher surface area constituted by smaller particles than that coming from pyrolysis under N₂. In fact, the BET-specific surface passes 79 m²/g for N₂-char to 174 m²/g for CO₂-char and the micropore volume fraction increases from 2 to 9. The observed differences can be attributed to the oxidizing environment (CO₂) that is more aggressive and reactive in the synthesis phase compared to an inert environment (N₂). Char also showed an anti-aging effect, hindering the increase in rigidity typically associated with the aging process. This effect was explained in terms of the compatibility of char with bitumen’s organic nature, and presumably its more effective hosting in a bituminous structure, which is presumably higher when char is produced under CO₂ rather than N₂. This study quantitatively indicates how a residue derived from the pyrolysis of a waste material can be efficiently re-used to increase the mechanical characteristics of bitumen, accomplishing the recent circular-based needs for environmental protection.

1. Introduction

In the last decades, the ever-increasing traffic demand brought about through the transportation of people and merchandise has elucidated the need of high-performance road pavements. Road pavements are made up of 95% wt/wt crushed stones and inorganic aggregates bound together by a viscoelastic organic-based fluid called bitumen. That being said, the most effective way to obtain better road pavements is to improve their fluidic part, i.e., the bitumen binder, since its role in holding up the whole structure gives it a pivotal role in determining the overall properties of the road pavement. Road pavement characteristics depend therefore on the physico-chemical properties of the bitumen used [1].

Through bitumen modification, the characteristics of a bitumen of low quality can be improved, thereby optimizing economic benefit from industrial raw materials, especially in a situation of deficiency of oil resources [2]. A bitumen modification approach, which deserves more attention, is the blending of the binder with fragments of wasted vehicle tires. This brings about important modification of rheological properties such as an increase in resistance to fatigue and fracture, and in rutting resistance [3,4]. A more sophisticated approach implies the treatment of bitumen with the solid residue (char) derived via the pyrolysis of waste tires. Pyrolysis is a fundamental process in the reduction in urban waste, which entails a thermal degradation process in the absence of oxygen, which converts wastes into combustible gases [5], an oil rich in hydrocarbons and oxygenated species (used for several applications [6,7,8,9]) and a solid, powdery residue (char) [10,11], which contains high amounts of carbon. It must also be noted that the absence of oxygen prevents combustion and therefore CO₂ emission.

For the purposes of our work, the last (solid) fraction is the most important as it can be well incorporated into bitumen’s organic-based structure thanks to its carbonaceous nature [12] and the porous/fibrous structure allows, in principle, enhanced interactions with bitumens [13].

Similar studies showed the great potentialities of char, whose applications need, however, adequate characterization [14]. A recent study demonstrated the use of char as an example of how to use waste in a circular economy view [15]. The scientific idea originating our work lies in the possibility that the waste from urban activities can be treated through pyrolysis, becoming added-value input ingredients for the preparation of bitumens. In this way, bitumens can have stronger resistance upon a temperature increase, suggesting lower maintenance costs of road pavements and eventually a reduction in waste materials.

The environmental benefit of this kind of work needs to be emphasized. In fact, efforts to minimize resource input and waste output need to be made [16]. Legislation in fact is leaning toward a transition to a regenerative circular economy [17] and so initiatives fostering raw material re-use and recycling while reducing the production of waste [18] must be encouraged. Production of asphalts with a higher performance coupled with optimal re-use of wastes facilitated through pyrolysis is an ideal match.

The use of pyrolysis residues is not trivial, since its characteristics (chemical and morphological) strongly depend on the pyrolysis parameters: the heating rate, the temperature range and the gas residence time. Therefore, these parameters must first of all be defined for reproducible results and, secondly, their effects need to be clarified. Although these parameters have been explored in the literature, to the best of our knowledge, the effect of the gas flowing on the material under the pyrolysis process has not yet been deeply studied. This is important knowledge, which is lacking, since the type of the flowing gas can affect the chemical reactions taking place in the material under pyrolysis, especially at the surface.

In the present work, the use of chars derived from the pyrolysis of waste tires under different flowing gases is proposed. In order to achieve comprehensive data interpretation, we carried out the following actions: (i) the morphology/porosity was investigated via microscopy and nitrogen absorption experiments, and (ii) bituminous materials modified with various amounts of chars were characterized by applying rheometry as a function of temperature.

2. Experimental Section

The waste-tire-derived powdered rubber was supplied by Ecopneus s.c.p.a. It is a black powder with 0.4–0.8-mm-sized grains obtained in a laboratory through particle sieving from tire fragments of a broader size distribution. The powder was washed with toluene and drying was carried out in an oven with the temperature set at 135 °C. The results of proximate and ultimate analyses on waste tires are reported in

Table 1. WT composition.

C (% w/w)	82.8
H (% w/w)	7.1
N (% w/w)	1.5
S (% w/w)	1.6
O (% w/w)	7
Humidity (% w/w)	0
Volatiles (% w/w)	66
Ashes (% w/w)	5
Fixed carbon (% w/w)	29

. The analysis shows the high carbon content typically expected for such a type of a material. It must be pointed out, in light of environmental attention, that they are classified as waste with CER 16 01 03, so they need to be treated according to the specific legislation.

The bitumen used for the purpose of this study was supplied by Lo Prete (Reggio Calabria, Italy) although the origin of the bitumen is Saudi Arabia. The characteristics of the bitumen have been determined in previous investigations [19]. Briefly, (i) the bitumen has a crossover temperature, i.e., the temperature at which the bitumen passes from a glassy elastic solid (G′ > G″) to a viscoelastic liquid (G′ < G″), of 7.8 °C; (ii) the penetration grade of the bitumen is 50/70 as measured through the standard procedure [20] of a standard needle being loaded with a weight of 100 g and then measuring the length traveled into the bitumen specimen in tenths of a millimeter for a known time, at a fixed temperature. The penetration grade of a bitumen is expected to decrease as an effect of particle addition [21] (see, later, discussion of the results).

Pyrolysis was carried out using an electric furnace with a sample holder in the form of a cylindrical stainless-steel reactor with a length of 15 cm and an internal diameter of 5 cm. The reactor linked to the ¼-inch pipeline allows the influx/efflux of gas through a Mass Flow Controller. The set parameters of the procedure such as temperature, gas flow and heating rate were monitored regularly in real time using homemade software developed via the LabView version 8.2 system (National Instrument data acquisition system) [22].

The samples (about 16 g for each run cycle) were placed in the stainless-steel reactor and heated up to 900 °C. The pyrolysis/activation test lasted approximately 6 h. These conditions were chosen after preliminary thermogravimetric tests and follow the general trend of pyrolysis of different raw materials. All samples (char) were obtained through pyrolysis/activation of waste tires under nitrogen and carbon dioxide constant flow rates to establish an inert/oxidant atmosphere inside the chamber. Different samples were obtained by changing synthesis parameters as reported in

Table 2. Sample’s synthesis parameters (“Nl” stands for normal liters). Final step, involving static temperature condition, is evidenced in green.

Sample	N2-char					N2+CO2-char					CO2-char
Parameters	Time (min)	Temperature (°C)	Gas	Flow (Nl/min)	Rate (°C/min)	Time (min)	Temperature (°C)	Gas	Flow (Nl/min)	Rate (°C/min)	Time (min)	Temperature (°C)	Gas	Flow (Nl/min)	Rate (°C/min)
Value	-	25	-	-	-	-	25	-	-	-	-	25	-	-	-
	80	500	N2	0.5	6.0	80	500	N2	0.5	6.0	80	500	CO2	0.5	6.0
	60	700	N2	0.5	3.3	60	700	N2	0.5	3.3	60	700	CO2	0.5	3.3
	120	900	N2	0.5	1.6	120	900	N2	0.5	1.6	120	900	CO2	0.5	1.6
	60	900	N2	0.5	-	60	900	CO2	0.5	-	60	900	CO2	0.5	-

. The final char (yield between 30 and 40%) was recovered from the reactor at the end of the experimental test. All of them were collected when the reactor reached room temperature. All samples were synthetized by using the same synthesis parameters except for the gas specimen, according to the synthesis parameters reported in Table 2.

Bitumen reinforcement by char was carried out as follows: char was added in three varying proportions (1, 3, 6% wt/wt) to bitumen, which was in its liquid flowing state at a temperature of 150 ± 10 °C. Homogenization of the mixture was carried out at the same temperature using a mechanical stirrer (IKA RW20, Königswinter, Germany) set to 500–700 rpm for 30 min. Rotation rates at a lower rpm are not effective in obtaining homogeneous samples, whereas if rotation rates are higher than 700 rpm, the bitumen could begin to undergo oxidation due to the increase in the bitumen–air interface in the vortex, which forms during stirring, and presumably also due to the obvious increased interfacial kinetics. For these reasons, this method is regarded as a usual procedure, as reported by other authors [23]: different conditions would deteriorate the final bitumen rheological properties, and also would render comparison with existing literature difficult.

After mixing the bitumen with char at the aforementioned temperature of 150 ± 10 °C, the modified bitumen was poured into a sealed can, stored in a dark chamber and allowed to cool down to a temperature of 25 °C. Due to the sensitivity of the samples to the annealing time [24], i.e., the time required for cooling down to room temperature, care was taken to ensure that the temperature cooling rate (5 °C min⁻¹) and annealing time (25 min) of all the investigated samples were the same. As a reference, a bitumen sample devoid of filler, hereafter labeled as “ref”, was used.

So, to evaluate the effect of the filler content during aging, all the above-mentioned samples (filler-modified bitumen samples and reference bitumen) were subjected to a 75 min aging cycle using the RTFOT (Rolling Thin-Film Oven Test according to ASTM D2872-04). However, in order to obtain bitumens with a rigidity, which simulate a prolonged aging process of about 10–12 years, typical of an asphalt pavement lifespan, the simulated aging was also prolonged to 225 min.

Nitrogen (N₂) adsorption isotherms were measured on ASAP 2460 (Micromeritics) at −196 °C so as to evaluate the porosity. Prior to the experiments, outgassing of all samples was carried out at 200 °C, with the exception of the sample “as is” char that was kept at room temperature, until a constant vacuum of 10⁻⁷ mbar was reached. The BET method was used to calculate the specific surface area (S_BET) while the volume of micropores (V_mic), volume of mesopores (V_mes), total pore volume (V_T) and pore size distribution (PSD) were calculated using the Non-local Density Functional Theory (NLDFT) [25], with the assumption of the pore wall heterogeneity.

A thermogravimetric analysis (TGA) was performed using a DSC-TGA STA409C analyzer (Netzsch-Gerätebau GmbH, Selb, Germany) from 40 to 500 °C at a heating rate of 5 °C in static air.

IR spectra were obtained using FTIR spectrometer Bruker Alpha Fourier transform spectrometer (Karlsruhe, Germany) in transmission. The measurements were recorded in the wavelength range 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹. In each run, 100 sample scans were recorded, and the background was also taken before each sample. A KBr disc was placed in the sample holder, and the holder was placed in the FTIR spectrometer. Ten spectra were taken from each disc and the mean of them was used for a further analysis.

The characterization of sample surface morphology was acquired using a Scanning Electron Microscope (SEM) (Field Emission SEM FEI Quanta 200 Kyoto, Japan) and Electron Probe Micro Analyzer (EPMA, JEOL JXA 8230t, Kyoto, Japan), equipped with a wavelength dispersive spectrometer (WDS) in order to determine both the elemental composition and spatial distribution. The surface of the samples was coated with a 5-nm-thick layer of carbon, using a Carbon Coater QUORUM Q150T-ES (Judges House, UK). The operating conditions for the SEM analysis were the following: HV of 15 KeV; Probe Current of 10–20 nA; working distance of 11 mm; image BSE signal; detector image of a solid state detector (SSD). The WDS analytical conditions (Spectrometers WDS XCE type and X type, Carl Zeiss AG, Oberkochen, Germany) were an accelerating voltage of 15 keV and a Faraday cup current of 10 nA, 11 mm being the working distance. All WDS analyses were carried out at room temperature (22 ± 1 °C).

Optical microscopy was carried out through an Olympus BX53 (EVIDENT Corporation, Tokyo, Japan) microscope equipped with a Nikon 50× magnification and 0.35 numerical aperture objective. The camera model was Olympus DP23 EVIDENT Corporation (Tokyo, Japan). A drop of a suspension containing char at about 0.13% wt/wt was placed onto a glass slide and spread onto it with the help of another glass slide. The liquid components were then allowed to evaporate, and the resulting deposit was observed through the microscope taking at least 30 pictures in different places for sufficient statistics. Images were analyzed by opportune software for the recognition of particles and their size determination.

Dynamic Shear Rheological (DSR) tests were performed to measure the complex shear modulus [26] G* = G′ + i G″ in the regime of small-amplitude oscillatory shear [27] at 1 Hz as a function of temperature. A dynamic stress-controlled rheometer (SR5000, Rheometric Scientific, Piscataway, NJ, USA) with a Peltier plate for temperature control (uncertainty ±0.1 °C) was used to perform the rheological measurements. The parallel plate gap was set to 2 mm, in line with existing literature [28], with the plate possessing a diameter of 25 mm. The real and imaginary parts describe the in-phase (storage, measure of the reversible elastic energy) and the out-of-phase (loss, irreversible viscous dissipation of the mechanical energy) moduli, respectively [29]. Time cure tests were carried out with a temperature ramp to investigate the phase transition of the samples: temperature was increased at a rate of 1 °C/min from 25 °C to 95 °C. To guarantee linear viscoelastic conditions at all tested temperatures in these measurements, stress sweep tests were carried out to choose the right conditions and apply the appropriate stress value in the measurements.

3. Results and Discussion

3.1. Char Characterization

All char samples possess millimeter and sub-millimeter particles. See Supporting Information (Figure S1) for a representative photograph. However, such particles are smaller particles that loosely aggregate and easily disassemble during handling. During the mixing and stirring of the char with hot bitumen to form the char–bitumen composite, they disassemble to give smaller particles. In fact, the final char-containing bitumen is quite homogeneous in texture, not containing such big mm-scaled (or sub-mm) particles. To estimate the size distribution of the “individual” particles, the CO₂- and N₂-chars were dispersed in silicone oil, the suspension sonicated in a commercial low-power sonicator and the final solution observed through optical microscopy. The soft sonication in silicone somehow simulates the char dispersion during the preparation of char-containing bitumens, where a suitable amount of char is stirred in hot bitumen. Chemically, silicone oil is apolar and reproduces the basic apolar characteristic of the bitumen [30], and its high viscosity at room temperature reproduces the viscosity of the bitumen at the high temperature (150 °C) employed during the mixing of the char with bitumen. In addition, sonication should somehow speed up char dispersion, in the same way as the stirring with hot bitumen does. Obviously, this treatment helps in disassembling the big char aggregates. It is just a way to simulate the bitumen preparation procedure. Microscopic investigation on the samples treated in this way just helps in understanding the size distribution of the individual char particles. Nevertheless, observing the real state of the char particles within the bitumen would need different and ad hoc techniques.

The size distribution of both char samples after the disassembling of the aggregates in silicone oil is reported in Figure 1. It must first be noted that the particle sizes for both char samples mostly fall below the value range of 1–5 μm with very few particles of a size >5 μm. It can also be noted that CO₂-char is more abundant in particles with a size below 1 μm than N₂-char. Contrarily, N₂-char contains more particles with a size greater than 1.2 μm compared to CO₂-char. The N₂-char size distribution is therefore shifted to slightly larger sizes compared to that of CO₂-char. It is also broader, whereas the size distribution of CO₂-char particles looks sharper. The particle size distributions for N₂+CO₂-char (not shown in the figure for clarity’s sake) are in-between that of the N₂–char and CO₂-char. The general behavior that CO₂-char particles are generally smaller than those of the N₂-char suggests that the latter has a higher natural chemical tendency to form aggregates. The observed differences can be attributed to the oxidizing environment (CO₂) that is more aggressive and reactive in the synthesis phase compared to an inert environment (N₂).

It is important to note that further experiments, performed with different dispersing solutions (see Supporting Information and Figures S2 and S3), gave similar results. These considerations are consistent with the N₂ adsorption experiments on the as-prepared (untreated) chars.

The behavior of the char upon a temperature increase was investigated using Thermogravimetry. The results are shown in Figure S4 of Supporting Information. Basically, the data show that at about 500 °C, degradation takes place for all the chars. Interestingly, the degradation of the CO₂-char is complete within just a hundred degrees, whereas it is much slower for the other chars. This confirms that the more oxidizing/reactive environment in which CO₂-char is prepared leads to a more reactive char particle surface. Moreover, investigating more deeply, a slight weight loss is seen for the N₂- and N₂+CO₂-chars at a temperature just above 100 °C, attributable to some residual humidity tightly bonded to the char. Interestingly, this behavior is not present in CO₂-char, which maintained its initial weight up to the degradation temperature of 500 °C. Further char characterization was carried out using FT-IR. The spectra are reported in Figure S5 of Supporting Information. The spectra are quite similar, due to the common carbonaceous nature of the three materials. The typical C=C and C=O stretching modes are visible around 1600 cm⁻¹. This band also reasonably contains the ring stretch of the benzene ring in aromatic compounds. At higher frequencies, the C-H stretching modes are present in the region 2800–3000 cm⁻¹, as well as the broad band due to O-H and/or N-H stretching around 3400 cm⁻¹. On the other hand, at lower frequencies, a peak centered at 1120 cm⁻¹ can be due to the vibrations of sulfur-containing functional groups, whose presence is expected since rubber sulfur compounds are used in the vulcanization process, making the tire rubbers. This information is confirmed through an elemental analysis performed through Scanning Electron Microscopy–Wavelength Dispersive Spectroscopy (SEM-WDS) on selected areas (see complete report in Supporting Information), which showed that the char contains small domains of ZnS. In fact, to facilitate rubber vulcanization for tire preparation, ZnO is typically used as an additive. Then, during the waste tire pyrolysis process, ZnO retains sulfur forming ZnS as a consequence of the reaction with H₂S evolving from the decomposition of the sulfur rubber compounds resulting from the vulcanization process. Apart from these isolated inorganic domains, the matrix is made up of carbonaceous materials, with an oxygen content generally increasing when passing from N₂-char to CO₂-char as a consequence of the presence of CO₂ during pyrolysis.

Figure 2 shows the nitrogen adsorption/desorption isotherms acquired on the samples labelled as N₂-char, N₂+CO₂-char and CO₂-char at liquid nitrogen temperature. The curve relative to an unpyrolysed waste tire is not shown since, due to its almost zero specific surface area, it was not possible to acquire the isotherm. According to the IUPAC classification [31], the shapes of the isotherms are predominantly of Type II with a very small contribution of Type I (P/P₀ < 0.05) and a more meaningful contribution of Type IV (0.85 ≤ P/P₀ ≤ 1.0), with a presence of a slight H1-type hysteresis. For these reasons, we can assert that synthetized materials have a slightly porous structure, which are composed of a very small part of micropores (<2 nm) and a preponderant part of mesopores (>2 nm).

The Type II isotherm exhibits unrestricted monolayer–multilayer adsorption. The beginning of the knee on the isotherms, which occurs almost immediately (micropores have been filled), indicates the stage at which monolayer coverage is complete and multilayer adsorption is about to begin. The completion of the multilayer followed by capillary condensation with its hysteresis loop, typical of Type IV isotherms, highlights adsorption on the mesoporous part over a range of a high P/P₀.

Table 3. Sample textural properties calculated from the adsorption isotherms.

Sample	SBET (m2/g)	V < 0.7 nm (cm3/g)	0.7 nm < V < 2 nm (cm3/g)	Vmic (cm3/g)	Vmes (cm3/g)	VT (cm3/g)	Vmic/VT (%)
“as is”-char	-	-	-	-	-	-	-
N2-char	79	0.010	0.003	0.013	0.521	0.534	2
N2+CO2-char	123	0.011	0.019	0.030	0.559	0.589	5
CO2-char	174	0.027	0.024	0.051	0.521	0.572	9

summarizes the structural differences. It reports the parameters calculated by analyzing adsorption isotherms. Figure 3 shows the pore size distributions (PSDs, Figure 3A) and cumulative pore volume (Figure 3B) evaluated using NLDFT. The higher cumulative pore volume of the CO₂-char sample is evident.

All the PSD plots (see Figure 3A) showed a significant peak in the microporous region (<1 nm) centered on 0.6 nm for the CO₂-char sample while it is shifted to the lower region (0.5 nm) for the N₂-char and N₂+CO₂-char sample. For all samples, a series of peaks in the mesoporous region are present (20 < x < 50 nm). It is worth noting that the BET-specific surface (S_BET) changes from 79 m²/g for N₂-char to 174 m²/g for CO₂-char. Coherently, the Vmic/V_T ratio, i.e., the micropore volume fraction, also changes from 2 for N₂-char to 9 for CO₂-char. This suggests a better dispersion of CO₂-char within the bitumen and, consequently, its more marked modifying properties toward the bitumen characteristics.

3.2. Bitumen Characterization

Both the real (G′) and imaginary part (G″) of the elastic modulus were measured through temperature-sweep measurements, implying the heating of the sample at a constant heating rate of 1 °C min⁻¹ and subjected to a shear deformation at a frequency of 1 Hz.

A typical result is shown in Figure 4A, where the curve associated with the pristine bitumen is reported. It can be easily seen that G′ is lower than G″ in the considered temperature range and for all samples. This highlights that the samples are characterized by a pseudoplastic fluid behavior. Interestingly, both G′ and G″ monotonously decrease with temperature, suggesting that the samples are at values greater than their glass temperature, where G″ should have a maximum, and usually occurring below 0 °C [32].

The plot in Figure 4A also allows the derivation of the G′ value at 50 °C (G′@50 °C). This parameter is taken to represent the material’s rigidity at working conditions (50 °C), in line with recently published research [33].

Upon a temperature increase, all samples become softer: at temperatures high enough, G′ suddenly drops, indicating the gel-to-sol transition. The graphical derivation of this transition temperature (T*) is shown in Figure 4A. From the physical chemistry point of view, T* marks the situation where the molecular relaxation rate, obviously depending on the molecular thermal agitation, becomes sufficiently high for the system to accommodate the perturbation given through the mechanical distortion. In this situation, purely flowing behavior is shown, with vanishing of any elastic storage of mechanical energy (G′ = 0).

G′@50 °C and T* values are shown in Figure 4B as a function of added char (wt%). The values shown give some interesting information. The addition of char causes an increase in both G′@50 °C and T*. This is quite a well-known effect: the presence of fine particles causes the reinforcement of the structure formed by the molecules in bitumen. This is basically due to interactions between the bituminous molecules and the particles’ surface. In this way, a network formed by alternate …bitumen–particle–bitumen–particle… is made with an overall effect to reinforce the overall structure. In this sense, and this is the scientific idea at the basis of the present work, organic-based particles like those of char can offer an organic-based surface for effective interactions with bituminous molecules. This chemical compatibility is expected to allow (i) better accommodation of the carbon-based particles within the bituminous matrix and (ii) effective interactions with the molecules of bitumen and a more marked effect of the hosted char. Looking at the data, the effect is at a maximum for char from CO₂ at 3% wt/wt, where the gel-to-sol transition temperature is increased by about 4.5 °C in the presence of CO₂-char. For char from N₂, a slightly lesser effect is seen with a more uniform (flat) effect as a function of char% and with an increase in the gel-to-sol transition temperature of about 3 °C. It is seen that char from CO₂ gives an effect that is overall more dependent on the char%. Coherently, the penetration index, expressed in tenths of a millimeter, changes from 66 of the virgin bitumen (rif) to the following values for modified bitumen (3% wt/wt):

-

58 for bitumen modified with N₂-char;

-

57 for bitumen modified with CO₂-char;

-

58 for bitumen modified with N₂+CO₂-char.

This effect can be due to the fact that CO₂, during pyrolysis, can cause the formation of oxidized (polar) chemical groups, which most probably can better interact with the polar parts (asphaltenes) of the bitumens. This would give stronger interactions with the bitumen. It can be argued that this can also give better distribution among the sample.

Comparison with bitumens modified with CaCO₃ is interesting. CaCO₃ is used in many works on bitumen and asphalt concretes as a standard inert filler. Addition of CaCO₃ at 3 wt% to bitumen gives a 4 °C increment in T* and a three-fold increase in G′@50 °C [34]. Interestingly, the addition of our chars gives a not-significant effect in T* (increase of about 3.5–4 °C) but a markedly higher effect in G′@50 °C (about seven times higher). It can be concluded that our char gives an enhanced effect in increasing the rigidity at working conditions.

It is interesting to note that the two quantities are generally correlated, as clearly visible in the correlation plot of Figure 4C. It must be pointed out that G′@50 °C refers to the Y-value of the G′ plot in Figure 4A, and T* refers to the X-value at which the curve drops in the same plot, so they are two independent parameters. The fact that they both give the same information means that they can be safely considered as self-consistent indicators of the effects exerted by char. The data in Figure 4C are compared with literature data (bitumens reinforced with organic molecules [35] shown by brown crosses, cellulose [36] shown by vertical green bars and inorganic fine particles [37] shown by horizontal blue bars). Clearly, all the data share the same behavior, suggesting universal behavior. Upon deeper observation, it can be noted that the data of this manuscript are slightly above the others, confirming that the char under investigation causes a marked effect on G′@50 °C. The fact that the proposed chars cause a bigger effect on the increase in G′@50 °C rather than in T* could be the consequence of the interactions between the char particles and the bituminous matrix being more effective at lower temperature, but they easily become less effective as the temperature is increased, thus ruling out particular effects when temperature approaches the transition temperature T*, where the effect is therefore not higher than that of any inert filler. However, the effect of char addition on the increase in G′@50 °C being quite high is noteworthy.

Interesting information can also be derived from the simulated aging experiments. It turns out that G′@50 °C and T* in each aged bitumen are higher than in the corresponding unaged sample. This is due to the increase in the number of polar functional groups of the molecules in bitumens [38] with their ultimate aggregation and constrained dynamics typically caused by the aging process. Therefore, the chemical reason for this change will not be discussed deeply here, and the reader can see paper [39] for further reading.

However, a quantitative analysis is insightful: to highlight the effect of char addition, an estimation of the change in G′@50 °C and T* with the aging time (i.e., the slopes of the plots G′@50 °C and T* vs. aging time) as a function of char addition was evaluated. Results are reported in Figure 5. It can be seen that the increase in T* with the aging time is almost constant for all the amounts of chars added to the bitumen and it does not differ from the values for the reference (undoped) bitumen. In addition, no significant effect is seen if the flowing gas is changed.

In synthesis, it can be concluded that the addition of char does not affect the increase in T* upon aging, but it almost halves the aging effect as reflected by the increase in G′@50 °C upon aging. This highlights a special effect exerted by the chars on the rigidity at working conditions of the bitumen, rather than on its transition temperature. This is also in line with the indications of the data in Figure 4, where the comparison with the inert CaCO₃ highlighted a more pronounced effect on G′@50 °C than on T*. The anti-aging effect could be ascribed to the effect of char particles, reasonably slowing down the kinetic processes of the molecules in bitumens by presenting to them a solid obstacle (the interfacial barrier). In this way, some of the bituminous molecule degrees of freedom are hindered. Since char can be classified among substances under the definition of activated carbon, the char carbon surface reactivity can also give a further chemical effect, in our opinion. For example, the presence of small amounts of zinc sulfide can give rise to further interactions between the metal and the asphalthene or resins present in the bitumen, as already suggested in recent works [37]. After all, the peculiar physico-chemical properties of activated carbon have been of added value in many applications, filtration, cleaning and adsorption of gases, liquids and contaminants [40,41], so that some anti-oxidation effects [42], and some increased thermal storage stability [43], have also been recently found in bituminous materials.

Rheological data allowed the derivation of the rutting parameter. It is defined as G*/sin δ. It was found that under usage conditions [44] (50 °C), the observed values are always higher than 1 kPa for unaged samples, meeting the limits imposed by the Superior Performing Asphalt Pavements method (Superpave SHRP, Strategic Highway Research Program) [45].

It must be underlined that a recent study [46] highlighted that pyrolysis of materials rich in carbon, like waste tires, can give carbon-rich char, which can be well accommodated within the bituminous matrix. In this regard, the indications of the present work corroborate that study well, by suggesting the use of CO₂ as the flowing gas to be used during pyrolysis for even more marked effects. The chemical explanation for this effect, i.e., the higher reactivity of CO₂ with respect to N₂, allows for suggesting that other reactive gases could be explored for future experiments in the piloted design of char for bitumen modification.

4. Conclusions

In this study, chars derived from the pyrolysis of waste tires were tested as green additives to improve bitumens’ characteristics. This study explores the effect of pyrolysis flowing gases on the char characteristics and on the final bitumen performances. The motivation lies in the consideration that the type of the flowing gas can affect the chemical reactions taking place in the material under pyrolysis, especially at the surface, affecting the final char characteristics, and consequently the performances of the bitumen modified with it.

Char addition brought about a significant increase in the rigidity under working conditions (i.e., G′ at 50 °C), especially at the char amount of 3 wt%, whereas the effect on the gel-to-sol transition temperature was found to be practically equivalent to that of a standard inert filler (CaCO₃). In particular, the char derived from pyrolysis under CO₂, which was found to possess a greater surface and be constituted by smaller particles than that produced under N₂, showed a higher effect. The observed differences are attributed to the oxidizing environment (CO₂) that is more aggressive and reactive in the synthesis phase compared to an inert environment (N₂). This means that the modified bitumen has a stronger resistance against temperature, which may be advantageous for uses in certain (hot) conditions.

Char also showed an anti-aging effect due to its high carbon content, which brings about greater compatibility of char with the organic nature of bitumen, and presumably its more effective hosting within the bituminous matrix. Self-consistently, the anti-aging effect was seen especially in the slowing down of the increase in rigidity upon aging, rather than in changes of transition temperatures. The anti-aging effect of the char is important because it prolongs the life-time of a bitumen and reduces maintenance costs.

The work fills a void of knowledge related to the rationalization of the type of gas used during the process. Apart from the comparison of all the data, which will give important information for future char production and its application [15], the environmental benefit of this work needs to be emphasized, by considering that (i) pyrolysis can reduce the amount of urban waste and that (ii) longer-lasting road pavements reduce maintenance costs, and waste production, whereas improved bitumen can clearly bring about higher performing asphalt, which will be durable and conducive for road users.